“Human resources practices as a mechanism for improving performance within public institutions and state-owned enterprises in Morocco”

In an economic environment marked by rapid change, characterized by factors such as globalization and increasing demands from stakeholders and clients, public institutions and state-owned enterprises have become significant players through their multiple interventions in providing public services to citizens and businesses, in implementing structuring projects for economic and social development, and in promoting investment. This paper aims to examine the correlation between human resource management and employee performance within Moroccan public institutions and state-owned enterprises to understand how these organizations should manage their human capital to enhance their performance. The sample consists of 67 top executives of Moroccan entities operating in various sectors. Data were collected through a self-administered questionnaire completed by the participants. Partial least squares (PLS) was used to estimate structural equation models and analyze causal relationships between variables. SmartPLS 4 software was employed for model analysis. The findings reveal a positive and significant impact of training, selective recruitment, digital transformation, and performance-based compensation on employee performance improvement. The results indicate that the T-values are 3.126, 2.870, 2.178, and 2.406, respectively. Regarding the Q² value, it stands at 0.899, confirming the model’s predictive capability. The GoF coefficient is 0.851, affirming the overall validity of the model. However, it was observed that there is no significant link between job security and performance, as the T-values did not exceed the threshold of 1.64. This study suggests adopting changes in HRM practices to enhance the organizational performance of public institutions and state-owned enterprises.


INTRODUCTION
Morocco has achieved numerous accomplishments over the past two decades with a focus on establishing good governance and transparency.The new concept of authority has marked a significant turning point in the evolution of state administrations, making them more capable of serving the public interest and providing high-quality services to citizens.Despite these significant achievements, which have addressed major challenges by improving public services through its political institutions, the development model adopted in 2021, charting the path for the country's development by 2035, faces difficulties hindering the desired human and social development.Public institutions and state-owned enterprises are among these challenges due to structural dysfunctions that make it difficult to keep pace with the changes in Moroccan society.These dysfunctions were highlighted by the Parliament and the Court of Auditors in 2016 and by the Special Commission on the Development Model in 2021.
In this context, His Majesty King MOHAMMED VI has issued high-level instructions for launching a comprehensive public sector reform to address the structural dysfunctions of public institutions and state-owned enterprises.Managerial transformation is a crucial aspect of this reform.This transformation aims to revise human resource management (HRM) methods and approaches to reconsider the role of human capital within public institutions and state-owned enterprises.While the human factor was traditionally seen as a support activity to other organizational functions, HRM now represents the function that enables a structure to provide high-quality services to users.Consequently, HRM encompasses many practices that must be considered to achieve the desired outcomes (Tiwari & Saxena, 2012).Al-Lawama et al. ( 2021) have emphasized the importance of better understanding the factors influencing employee performance while considering various HRM practices.Therefore, there is a need to assess the extent to which HRM practices (training, selective recruitment, digital transformation, performance-based compensation, and job security) can influence the performance of Moroccan employees of public institutions and state-owned enterprises.

LITERATURE REVIEW AND HYPOTHESES
Human resource management (HRM) is of strategic importance for organizations.Numerous theories aim to elucidate the causal relationship between HRM practices and performance results within various business entities.Bourguignon (1995) defined performance as a system of complementary parameters that determine the results produced by the studied individual and the process through which these results are achieved.In the modern era, HRM has become an urgent necessity.Several studies have shown that effective HRM enhances workers' productivity and an organization's competitive advantages (Quresh et al., 2016;Wright et al., 2003;Balochi et al., 2010).Guest (2002) demonstrated that the interrelationship between employees and HRM practices is a crucial factor in determining the impact of HRM on organizational performance.In other words, when employees perceive these approaches positively, it can lead to good organizational performance, while a negative perception can result in mediocre performance.Quresh et al. (2016) found that all the tested variables (selection, training, performance evaluation system, compensation system, and employee participation) correlate with banks' financial performance.This underscores the importance of the HR department in optimizing employee performance by providing qualified professionals whose impact is felt throughout the organization (Amir et al., 2022).Therefore, it is imperative for companies to continually improve their HR strategies to remain competitive in the market, especially in highly competitive industries (Hee & Jing, 2018).According to Collins and Smith (2006), performance-based compensation encompasses all the financial benefits an executive, manager, or employee can receive in addition to their base salary.As an HRM practice, compensation is crucial in achieving a company's success.Islami (2015) specified that performance-based individual compensation encourages employees to consistently exert effort to improve their results, enhancing their motivation and engagement.Takeuchi et al. (2003) note that employees' professional development enables them to enhance their skills to strategically achieve the organization's objectives.
Godard and Khemir (2023) examine the link between incorporating corporate social responsibility (CSR) criteria into executive compensation and companies' social and financial performance in the SBF 120.The results highlight that considering CSR approaches to compensate executives has a positive impact on the social and financial performance of companies.Ali (2016) described employee performance evaluation as a process aimed at assessing their work performance by examining their task execution and behavior.The goal is to measure their ability to perform tasks successfully and efficiently while contributing to their personal development and that of the organization (Sweis et al., 2020).
According to the World Bank report ( Job security is workers' confidence in maintaining their employment over the medium and long term without fear of arbitrary or frequent layoffs.Previous research on the impact of job insecurity on worker performance has yielded mixed conclusions.James (2012) considers job security to be crucial for overall performance because when job security is lacking, personnel will lack confidence in their professional future, negatively affecting performance.It is noted that the greater an employee's job security, the more likely they are to perform their tasks effectively, which positively influences overall organizational results and performance (Society for Human Resource Management, 2011).
Without stable employment, employees will likely experience workplace stress and negative emotional reactions that can affect their professional efforts (Jordan et al., 2002).Cheng and Chan (2008) reported that job insecurity, one of the most significant stressors, is negatively associated with productivity, work engagement, and employee health.Furthermore, gender does not significantly affect the relationship between job insecurity and the studied variables.
However, other empirical studies have shown that job security and organizational performance are not necessarily linked.Vlachos (2009) indicated that job security is not essential in predicting organizational performance.Sverke et al. (2002) found that job security had no significant or measurable effect on performance.In other words, there was no clear and statistically significant link between the degree of job security and the quality of work performed by employees.The explanation for the weak negative correlation between performance and job security offered to employees is that rational employees experiencing job insecurity can make efforts and maintain their professional performance to be perceived as applicable to the organization (Greenhalgh & Rosenblatt, 1984;Sverke et al., 2002).
The first definition of the digital transformation concept appeared in a book by Stolterman and Fors (2004) 1).The research hypotheses are formulated as follows: H1: Employee training is positively linked to organizational performance.
H2: Performance-based compensation is positively linked to organizational performance.H4: Job security is positively linked to organizational performance.
H5: Digital transformation is positively linked to organizational performance.

Data collection
An investigation was conducted between January and March 2023 using a randomly selected sample from 10 public institutions and state-owned enterprises providing public services in the capital of the kingdom.Seven of them agreed to participate in the survey.The target population consisted of personnel working in the HR departments of these organizations, totaling 67 individuals.This approach aimed to shed light on the research questions of the survey.Data collection was carried out through a questionnaire completed by the respondents themselves.This questionnaire includes items inspired by the literature review and empirical studies, where respondents assess the degree of implementation of HR practices using a Likert scale ranging from 1 to 7. Table A1 (Appendix A) defines the measurement items.

Data analysis techniques and model measurement
The study adopted the partial least squares structural equation modeling (PLS-SEM) for the analysis.Using structural equation models, this statistical analysis technique examines complex relationships between a set of latent and manifest variables.This technique can estimate models with small samples since its various algorithms produce results with a high degree of statistical power (Rigdon, 2014) First, Cronbach's alpha coefficient was used to measure the reliability of the questions posed to the respondents.This coefficient is the mathematical equivalent of estimating the average of all correlations between two equal parts of the scale and ranges from 0 to 1.The closer the alpha value is to 1, the more homogeneous the set of items is considered.According to Cronbach (1951), values exceeding 0.7 are considered appropriate (Hair et al., 2017).This threshold indicates that the measurement error is equal to more than half of the standard deviation of the total score distribution.The calculation formula is: where α represents Cronbach's alpha; N is the total number of items associated with the latent variable; k is the sum of variances of the items; σ i 2 is the variance of each item or manifest variable (i), and σ i 2 is the total variance of the scale.
Next, the study examined convergent validity to analyze if the survey components that are theoretically expected to be interconnected demonstrate this interconnection in reality (Hair et al., 2011).
To this end, the study checked three important criteria: composite reliability, individual indicator reliabilities (loadings), and average variance extracted (AVE).
Composite reliability (CR) assesses the internal consistency of a latent concept.Adequate internal consistency in a model should have a value greater than 0.70 (Nunnally, 1978).
The reliability of individual indicators, measured by their factor loadings, is a crucial measure of the quality and robustness of the relationship between the questions in a questionnaire and the latent fac-tor they are supposed to measure.In other words, it determines how much each indicator reliably contributes to measuring the underlying concept.
High factor loading indicates a strong correlation between the indicator and the underlying concept, ensuring the indicator's reliability.In contrast, low factor loading questions its relevance in measuring the concept, which is essential in confirmatory factor analyses and the validation of measurement scales.
The Average Variance Extracted index evaluates the proportion of variance in a latent concept explained by its indicators compared to the variance attributable to measurement errors.An AVE value of at least 0.50 indicates that the underlying concept explains over half of the variance in the indicators (Hair et al., 2011).The formula for AVE is as follows: where k is the number of items in the measurement scale; λ i represents the factor contribution of item i, meaning the correlation between item i and the latent construct.In other words, it measures the strength of the relationship between the item and the concept to be measured, and e i is the residual variance of item i, which represents the portion of the item's variance not explained by the latent construct.This is the measurement error or noise associated with the item.
Finally, discriminant validity was assessed to measure the degree of differentiation of a construct from others through two aspects: the Fornell and Larcker criterion and cross-loadings of different items.Regarding the first aspect, the square root of the AVE (Average Variance Extracted) of each latent construct should exceed the correlation coefficients with other latent constructs (Chin, 1998; Abraouz & Chakir, 2020); as for the second aspect, the loading of an item should not be higher for another construct than the one to which it is attached.
As for the structural validity, which is essential to test the hypotheses of the model, the evaluation criteria adopted in this study are the Bootstrap procedure, determination coefficients, Cohen's ƒ effect size index, the Stone-Geisser coefficient, and the GoF index (Chin, 1998;Hair et al., 2017).
The study used standardized correlation coefficients (path coefficients) to measure the direction of links between different latent or observed variables in the model and the impact of one variable on another (Bennaceur & Chafik, 2020).The closer the coefficient is to 1 (or -1 in the case of a negative relationship), the stronger the relationship.A coefficient close to zero indicates a weak relationship (Hair et The determination coefficient assesses the quality of structural models by quantifying the proportion of variance in a dependent variable explained by the associated independent variables.This choice is because this coefficient is considered the most appropriate criterion (Henseler et al., 2009;Hair et al., 2017).It ranges from 0 to 1, and the closer it is to 1, the higher the model's predictive capability.Researchers generally aim for high values like 0.75 or above, but Chin (1998) suggests that values exceeding 0.67 indicate strong consistency; those between 0.19 and 0.33 are considered moderate, while values below 0.19 are weak.
After assessing the determination coefficient for each endogenous latent variable, it is crucial to study the change in this coefficient following the removal of exogenous latent variables, which amounts to determining the intensity of the impact of the latter.Values exceeding 0.35 are associated with a large effect size; those ranging from 0.15 to 0.35 indicate a moderate effect size, while values below 0.02 indicate a small effect size.The formula for f² is as follows: where R 2 included is the determination coefficient R² of the model that includes the explanatory variable in question.It represents the proportion of variance in the dependent variable explained by the model that includes the variable.R 2 excluded is the R² of the model that excludes the explanatory variable.It represents the proportion of variance in the dependent variable explained by the reduced model without the variable.
The Stone-Geisser coefficient (q²) is used to measure the impact of each variable.If Q k 2 > 0, it is said that the model exhibits predictive validity (Fernandes, 2012).The formula is as follows: with N being the number of observations, Y i representing the observed values of the dependent variable, Ŷ i being the values predicted by the regression model, and Ŷ being the average of the observed values of the dependent variable.
It is important to note that the PLS-SEM method cannot optimize a global scalar function.As a result, the study must use the Goodness of Fit (GoF) as an additional indicator to assess the robustness of this method (Tenenhaus et al., 2004).The GoF represents an operational solution to this problem as it can be conceptualized as a global validation index for the PLS model.This criterion varies between 0 and 1, where values greater than 0.36 are large, those between 0.36 and 0.25 are medium, and those below 0.19 are small (Tenenhaus et al., 2005).The GoF is calculated as follows:

Respondents' demographics
The demographic results of this survey are presented in Table 1.Demographically, most participants were women, accounting for 57%, while men constituted 43% of the sample.Most participants were under 40 years old, representing 73% of the total.

Hypotheses testing
Using the smartPLS software for data analysis, the study developed the measurement and structural models.Figure 1 visualizes the AVE values and loadings, while Table A2 in Appendix A succinctly summarizes the values of Cronbach's alpha, CR, AVE, and loadings.
According to Table A2 in Appendix A, the exploratory factor analysis revealed that the Cronbach's alpha values for the variables of training, selective recruitment, digital transformation, performance-based compensation, and job security are all above 0.7, representing 0.973, 0.882, 0.906, 0.857, 0.906, and 0.898, respectively.Furthermore, the composite reliability values for these variables exceed 0.70, representing 0.980, 0.919, 0.934, 0.903, 0.935, and 0.929, confirming the presence of satisfactory internal consistency among the questions of each variable.The estimates of factor contributions (loadings) are significant at a 1% level and range from 0.752 to 0.932, suggesting the reliability of the measures.Finally, all AVE values exceed 0.50, indicating convergent validity.Since the study confirmed both internal consistency reliability and convergent validity, it proceeds to assess the discriminant validity of the constructs.
Discriminant validity for all constructs has been confirmed in this study as both conditions have been met.Firstly, referring to Table 2, it can be ob-served that the square root of the AVE values exceeds the correlations with other latent concepts (Chin, 1998), indicating that the latent constructs exhibit substantial variance with the measures that compose them compared to other latent constructs.Thus, it can be affirmed that the Fornell-Larcker criterion is met.Secondly, as shown in Table A3 in Appendix A, the indicators in the model exhibit significant cross-loadings with the latent constructs they belong to compared to other latent constructs.Therefore, the confirmation of discriminant validity for the measurement instruments is achieved.
The psychometric results of the model are satisfactory, allowing the analysis of the structural relation-ships between constructs and testing five research hypotheses.The results from the bootstrapping analysis are summarized in The Q² value obtained from the software is which means that the model developed in this study is predictive.As for the GOF, by applying the previously mentioned formula, the study obtains a value of 0.851, which is considered significant according to Wetzels et al. (2009).The results conclude that the reliability and validity of the proposed model are confirmed and widely acceptable.

DISCUSSION
The present study examines the effect of HRM practices (training, selective recruitment, digital transformation, performance-based compensation, and job security) on employee performance in public institutions and state-owned enterprises.After examining the results, it was found that train-ing, selective recruitment, digital transformation, and performance-based compensation have a significant and positive effect on the performance of employees, thus confirming H1, H2, H3, and H4.Notably, training and selective recruitment were the top factors contributing to the modernization of public institution and state-owned enterprise services, followed by performance-based compensation.However, job security had no impact on the model, potentially because employees facing job insecurity may be more motivated to maintain good performance to demonstrate their value to the organization.
This study complements and reinforces other studies that have demonstrated the effectiveness of HRM policies in enhancing organizational performance (Vlachos, 2009 In light of these mixed results, it is crucial to consider that employee reactions to job security are not solely determined by individual factors but also depend on how organizations treat their employees (Sverke & Hellgren, 2002).In summary, the way HRM is implemented has a direct impact on employee performance in various organizations.These findings will help address questions about the role of HRM as a lever for performance within public institutions and state-owned enterprises.
As a recommendation, it is advisable to implement tailored training programs for staff and involve them in decisions regarding their professional development to enhance the performance of public institutions and state-owned enterprises.
Compensation should be linked to performance, moving from a seniority-based to experience-based logic, with performance evaluations based on clear and transparent criteria.In terms of recruitment, it is recommended to select candidates based on their skills, knowledge, attitude, and qualifications, using appropriate selection methods.
Although job security has no direct effect on organizational performance, public institutions and state-owned enterprises are advised to act on three key aspects: developing employee employability, providing better support during organizational changes, and establishing a clear agreement between top management and employees.Job security and performance within an organization are interdependent, but their influence may vary depending on the context and management.
Concerning digital transformation in the public sector, while public administrations do not seek a competitive advantage over each other due to the state's monopoly, they aim to improve services to citizens while maintaining sustainable efficiency despite the costs involved.To succeed in this transformation, it is necessary to modernize administrative services by digitizing procedures, enhancing cybersecurity, ensuring digital sovereignty, expanding access to high-speed and very high-speed Internet for the entire population, and guaranteeing service quality.However, it is essential to note that digital transformation presents challenges regarding cybersecurity, data protection, and equity in access to digital services, requiring careful planning, effective change management, and stakeholder collaboration.
Despite the significant contributions made by this study, it has several limitations.Firstly, the sample size is relatively small, limiting the generalizability of the results.Future research should consider including more public institutions and stateowned enterprises for more robust conclusions.Additionally, it would be beneficial to explore moderating variables such as organizational climate, labor market conditions, and the legal and regulatory framework for a more in-depth understanding.Lastly, the use of self-reported data, rather than verifiable and objective data, may limit the reliability of the reported figures and, consequently, the conclusions drawn from these data.

CONCLUSION
This study aims to examine the impact of HRM practices on the performance of employees in public institutions and state-owned enterprises in Morocco.The results have confirmed that training, selective recruitment, digital transformation, and performance-based compensation positively impact organizational performance within Moroccan public institutions and state-owned enterprises.This supports the Moroccan government's vision to introduce HRM as an essential pillar for reforming public institutions and state-owned enterprises.However, this study could not confirm a positive relationship between job security and organizational performance.
The results provide general recommendations for the public sector in general and public institutions and state-owned enterprises to enhance employee performance.Therefore, HRM practices should be reviewed and rethought to be more effective and influential in achieving favorable organizational outcomes.
The recommendations formulated in this study aim to strengthen the performance of public institutions and state-owned enterprises in Morocco.Firstly, it is suggested that skills in training and organizational learning be developed.Secondly, these entities should adopt performance-linked compensation, rewarding employees based on their achievements and goals attained.Thirdly, a skills-based recruitment process should be established to select the best candidates objectively, thus avoiding nepotism and favoritism.Fourthly, it is recommended to balance job security and performance, as excessive job security can lead to complacency and decreased performance.Finally, new technologies should be enhanced to optimize processes, improve operational efficiency, and facilitate citizens' access to public services.
However, the study acknowledges limitations related to the sample size and composition.Future research could include larger and more diverse samples.Furthermore, it is recommended that the moderating effect of certain organizational variables in the relationship between innovation and organizational learning be analyzed, as well as that the antecedents and consequences of organizational learning be identified more comprehensively.

Furthermore,
Osman et al. (2011) argue that deficient HRM can lead to multiple adverse consequences, including employee demotivation, reduced effectiveness, and disinterest in the organization, ultimately resulting in a decline in overall performance.DeNisi and Griffin (2001) defined training as the set of approaches implemented to equip workers with all the necessary skills to perform their tasks effectively.Lyons (2009a) emphasized that employee training improves performance by influencing the prevailing culture and employee behavior within the organization.The study byBarzegar and Farjad (2011) on the influence of worker training on the Organization of Martyrs' Affairs revealed a positive impact.However, the results fell below expectations, emphasizing the need to tailor training to individual employee needs and the importance of prior course information to encourage participation.Lyons (2009b) highlights the positive impact of team training in constructing and maintaining templates for individual and team performance in the field.It underscores that employee training is crucial for improving organizational efficiency and aligning employee culture and behavior with the company's objectives.

Figure 1 .
Figure 1.Research framework (Lahchame & Djilali, 2021)gital technologies helps companies accomplish tasks more quickly, cost-effectively, and potentially with better quality(Setia et al., 2013) and provides better service to customers and employees, as well as risk management and cost control(Lahchame & Djilali, 2021).
(Devi, 2018)tko, 2019)"changes that digital technology brings or influences in all aspects of human life."Digitaltransformationrefers to the integration of digital technologies into all areas of an organization or society, resulting in fundamental changes in how activities are conducted, services are provided, and people interact(Verina & Titko, 2019).It has been shown that digital technologies increase productivity, reduce costs, and stimulate innovation.In the context of the health crisis, digital transformation has played a crucial role in enabling economic development and growth (Soto-Acosta, 2020).Digital transformation improves customer understanding to deliver tailored products and services (Lahchame & Djilali, 2021) by using data collection for enhanced customer interaction.It enhances business management and optimizes internal processes, improving performance and simplifying decision-making (Cabinet Étude Mille-Alliance, 2015).It is worth noting that several authors have sought to address emerging issues in this field, encouraging new studies to provide innovative perspectives and research implications while filling gaps in existing research(Devi, 2018).As a result, this work aims to analyze human resource management's influence on employee performance, focusing on Moroccan companies and public establishments.Given the diversity of HRM practices, this study focuses on training, selective recruitment, digital transformation, performancebased compensation, and job security (Figure al., 2017).

Table 4 .
Evaluation of R² and F²

Table 3 .
Results of model validation