The mediating role of innovation in the relationship between high-performance human resource management practices and firm performance

Abstract

The study aims to evaluate the impacts of high-performance human resource management practices on firm performance through the mediating variable of innovation. The research model was developed and integrated from high-performance human resource management theory, innovation, and firm performance. The analysis results from 312 enterprises in industrial and service sectors indicated that high–performance human resource management practices had an indirect impact on firm performance and a direct effect on innovation. Furthermore, innovation plays a mediating role in the relationship between high-performance human resource management practices and firm performance. In addition, organizational size is also a characteristic variable that impacts firm performance. Finally, the study provides some theoretical and practical implications for managers to improve enterprises' human resource management systems.

1. Introduction

Human resources are considered one of the most important resources of an enterprise [1,2]. High-performance human resources are an essential asset to create competitive advantages in the market [1,2]. Many studies have found an apparent influence of human resource quality on firm performance in different fields such as manufacturing [[3], [4], [5], [6]] and services [[7], [8], [9], [10]]. Thus, improving human resource management practices is proposed to solve new problems of enterprises such as quality management systems, lean production, technology improvement, or business process change [11,12]. Human resources is also seen as a form of human capital that generates productivity and helps improve firm performance [[13], [14], [15]]. Consequently, by improving the high-performance human resource management system, enterprises can enhance their position in the market and improve their competitiveness [[16], [17], [18]].

Due to the vital role of human resource management in enterprises, this topic has attracted many studies from different scholars in recent decades [8,[19], [20], [21], [22]]. Recently, studying the relationship between high-performance human resource management and firm performance has emerged as a trend in research on human resource management in enterprises [12,22]. Because a high-performance human resource management system is a set of separate human resource practices, functions, and processes that are interrelated to attract, retain, and develop an enterprise's human resources [23]. This system aims to sustainable business development in many business activities, not only towards financial efficiency as the traditional approach [16,24,25]. A high-performance human resource management system is a flexible system that improves firm performance [6,10,12,22,26,27] and a trend for human resource management activities in modern enterprises [3,12].

Studies also confirm the influence of high-performance human resource management on the innovation process in enterprises [12,28,29]. Because practicing a high-performance human resource management system helps enterprises motivate employees to promote the sharing and application of knowledge in the workplace [[30], [31], [32]]. This leads to promoting changes in the management system to improve productivity employee performance and create a competitive position for enterprises [2,19,22]. As a result, employees' new ideas, knowledge, and experiences are transformed into new products, processes, or services to meet the rapidly changing needs of customers and create competitive advantages for enterprises [33,34].

Innovation activities have long been considered an important factor affecting firm performance [12,[35], [36], [37], [38], [39]]. However, research on the relationship between high-performance human resource management systems, innovation, and firm performance is not popular, especially in developing countries except research of Lu et al. [12]. According to Lu et al. [12,40,41] innovation have mediator role in the relationship between the high-performance human resource management system and firm performance. The argument to explain this relationship is that a high-performance human resource management system promotes organizational change from within, encourages employees to change behavior, and applies new knowledge and ideas to their work [12,40,41]. As a result, firms are more open to new ideas from their employees, more open to outside knowledge, and can engage in more innovative activities. Applications of innovation lead to organizational change, product, process, or marketing improvements that lead to significant advances that help enterprises improve performance to increase profits. In other words, practicing a high-performance human resource management system promotes innovation in enterprises and indirectly affects firm performance.

In Vietnam, the number of operating enterprises has rapidly increased from a few thousand to nearly one million enterprises after over 30 years of transitioning from a planned economy in the Soviet model to a market economy [42]. New enterprises are being established more frequently, with over 40,000 new enterprises being founded each year [42]. With the goal of rapid international integration, the Vietnamese government encourages the development of the private economy, opening doors for free business activities for all economic components, not only the state economy as in the Soviet model during the centralized planning period [43]. The government also implements various policies to attract foreign investment and promote the transfer of not only technology but also managerial knowledge from developed countries, particularly Western countries. Observations by researchers indicate that the trend of enterprises is moving towards greater openness, with dimensions of adhocracy culture and market culture emerging and reducing the hierarchical and clan cultures in enterprises as their level of integration with the global economy increases [44]. As a result, Vietnamese enterprises are increasingly adopting modern management systems and tools such as KPI, BSC, OKR, LEAN, and TQM [45]. Alongside the growing application of management systems and tools in areas of production and operations, high-performance human resource management systems are also being introduced and widely accepted in many enterprises recently. Despite receiving strong support from many enterprises, research on the impact of high-performance human resource management system practices in Vietnam is still relatively limited [46,47]. Questions still persist about whether the application of high-performance human resource management systems truly stimulates innovation and improves firm performance for enterprises. Therefore, it remains essential to conduct studies reevaluating the influence of implementing high-performance human resource management systems on innovation and firm performance in the context of Vietnamese enterprises – an evolving economy with distinctions from developed economies.

In this light, although previous research findings have demonstrated the influence of practicing high-performance human resource management on innovation and firm performance [2,3,12,20,26,29], such results remain to be thoroughly validated in transitioning economies like Vietnam. Furthermore, there are variations in the measurement of high-performance human resource management practices across studies. For instance, initial studies considered job security as a crucial component of this system [2]. In contrast, research by Lu et al. [12] in China, a country undergoing economic transition, did not regard job security as a significant element within this system. The rationale put forth by the Chinese authors is that assuring job security for businesses is increasingly challenging due to the rapidly changing market environment [6]. While this rationale may be suitable for China's business and labor market context, it may not be as applicable to other emerging markets, smaller in scale, such as Vietnam. In Vietnam, given a delayed economic opening and a stronger influence of the traditions established during the Soviet-style economic period (pre-1986), employees tend to prioritize job security from employers. Surveys indeed indicate that laborers in many areas tend to favor working for state-owned enterprises or larger corporations over the private sector or smaller businesses [48]. The reason is that Vietnamese workers seek stable employment and income, which state-owned enterprises and larger companies are better positioned to provide compared to private sector entities or smaller businesses [49,50]. Consequently, the argument against considering job security in the high-performance human resource management system, as seen in studies in China, may not be applicable. Job security could remain an important component for enterprises applying high-performance human resource management systems in Vietnam. Hence, our study continues to regard job security as a vital component of the high-performance human resource management system. The research aims to assess the impact of high-performance human resource management practices on firm performance by fostering innovation for enterprises in transitioning economies like Vietnam. The findings of this study may offer valuable insights both academically and practically for managers.

2. Theoretical background and hypotheses

2.1. High-performance human resource management system

To use a high-performance human resource management system to create a competitive advantage based on a resource perspective (RBV) [51,52]. From the perspective of resources, there are three types of resources (i) human resources, (ii) physical resources, and (iii) organizational resources [51,52]. Furthermore, human resources are considered as one of the resources that create core competitiveness because human resources are considered to satisfy VRIN criteria: value, rarity, inimitability, and non-substitutability [2,[16], [17], [18],51]. Therefore, a high-performance human resource management system is considered an important factor affecting many aspects of the organization, such as improving innovation capacity [29,40], promoting improved firm performance [2,6,12].

Despite being a popular research topic, a high-performance human resource management system is controversial without a unified definition [2,6,19,22]. A high-performance human resource management system is defined as a model of human resource deployment and planned activities aimed at helping an organization achieve its goals [15]. In addition, a high-performance human resource management system is defined as a set of distinct, interrelated activities, functions, and processes aimed at attracting, developing, and retaining the human resources of enterprises [23]. In addition, a high-performance human resource management system is also a set of human resource management activities that, by improving the skills of employees, create opportunities for them to participate in the workforce decision-making processes and foster discretionary efforts by employees that can provide a solid competitive advantage reflected in better financial results [2,53]. In this study, a high-performance human resource management system is considered a coherent and logical organic combination of human resource practices. This system may influence the behavior and attitudes of employees, thereby affecting work results and employee enthusiasm leading to better operational performance for the enterprises [12].

Evaluating a high-performance human resource management system is quite complex and is often assessed by multidimensional scale [6,12,22]. For example, Delery & Doty [8] introduced the evaluation of a high-performance human resource management system through components: (1) training, (2) job analysis, (3) employee participation, (4) employee performance evaluation, (5) employee development, (6) job security (Table 1). This scale has been widely accepted and used in subsequent studies [6,22].

Table 1.

The components of the high-performance human resource management system.

Components	Description	Source
Training	It is how enterprises provide employees with comprehensive working knowledge and skills to complete their jobs well and advance in the enterprise.	[8,12]
Employee Participation	The enterprises motivate employees by allowing them to express their views to the manager to offer the best working methods.	[8,12]
Job Analysis	It is the process by which an enterprise analyzes and gives specific responsibilities for each job position to employees based on the job description.	[8,12]
Performance Evaluation	It is the fact that the enterprise recognizes and evaluates the results of the employees in an objective, fair and accurate manner.	[8,12]
Employee development	It is the enterprise's construction of a career promotion plan/route for employees in the enterprise.	[6,12]
Job security	The enterprise's construction of a stable working environment creates a sense of security for employees who are not easily fired at work.	[6,8]

2.2. Firm performance

Firm performance is often described as the achievement of outputs through business activities and reflects the business's success [54,55]. Firm performance is also often defined as the firm's performance and its ability to meet shareholder expectations [39,56,57]. There exist two different ways or viewpoints of measuring the firm performance of enterprises: measuring objective and measuring subjective [58]. In the objective aspect, firm performance is evaluated by financial indicators such as revenue, profit, ROA, ROE according to the accounting approach [2,8]. However, an objective scale may be inappropriate for businesses at different market stages or from different business sectors. For example, in the market penetration phase, a firm can endure expected loss to achieve its market entry objective and still be considered a good firm performance. Therefore, firm performance can also be measured subjectively based on achieving organizational goals [39,55,59]. In this way, firm performance measurement of the enterprise is determined by assessing whether the enterprise has achieved the organizational objectives in both financial and non-financial aspects [36,55,57]. This shows that it is possible to use firm performance measurement approaches subjectively in studies thanks to comprehensiveness and convenience in accessing and collecting data. Therefore, we use the subjective measurement approach with the firm performance variable in this study. According to the approach from Lu et al. [12], and Zhang & Li [6], firm performance is reflected through two components:

Operational performance: Enterprise's outputs are related to the employees' working morale, and the overall results of the enterprise are reflected through fundamental financial indicators.

Market performance: The market performance reflects outputs of the business related to the industry's sales, market share, and growth rate.

2.3. Effect of high-performance human resource management system on innovation

A high-performance human resource management system focuses on job design, improving employee problem-solving skills, reducing individualism, and increasing cooperation among members to develop innovative ideas that help teams complete tasks more efficiently [40]. In addition, a high-performance human resource management system also helps enterprises to accurately train and evaluate the employee performance of well-qualified employees, helping them to enthusiastically come up with innovative ideas and creatively apply these ideas to teamwork to increase innovation outcomes for the entire enterprise [12,29]. In addition, by participating in the decision-making process through open and honest discussions, a high-performance HRM system also helps to strengthen the relationship between employees and managers, which helps employees' new ideas easily be accepted and implemented by managers at work, thereby improving innovation capacity of enterprises [12]. And when the relationship between employees and managers becomes cohesive, employees will have a sense of security as well as positive attitudes to innovate behavior through the formation and implementation of new ideas at work. Therefore, it can be concluded that if enterprises have a solid, high-performance human resource management system, it will create a foundation to promote innovative and creative ideas of employees [29,40,41].

H1

High-performance human resource management system has a positive impact on innovation in the enterprise.

2.4. Effect of high-performance human resource management system on firm performance

A high-performance HRM system affects the ability to innovate and directly affects firm performance [6,12,22,27]. A high-performance human resource management system will provide employees with comprehensive training programs, helping them improve their qualifications and problem-solving skills, thereby increasing their work results as well as employee and firm performance [2,6,27,60]. In addition, with job analysis, a high-performance human resource management system will help to clearly define the scope of the employee's work, helping the employee to do the assigned work well, without being distracted by other tasks, thereby helping enterprises improve operational performance [6,12]. At the same time, if enterprises allow employees to participate in the decision-making process, it will motivate employees to come up with creative ideas, contributing to improving the quality of management decisions, which increases the performance of the enterprise's goals [12,53].

Furthermore, when the enterprise guarantees jobs for employees, it will send a positive message that the enterprise wants employees to stick with the organization for a long time, thereby promoting the capacity and motivation of employees to work so that they work hard and bring better operational performance to the enterprises [8,49,50]. In addition, a high-performance HRM system allows employees to develop within the enterprise [8,19]. And if they receive a clear promotion path, employees will try their best to achieve the best employee performance, thereby bringing better firm performance [6,12]. Finally, a high-performance HRM system is also concerned with profit-sharing for employees [8]. This has created a motivation to help employees achieve higher performance at work, thereby assisting firms to achieve better performance [6,8]. These results have been verified in many studies on the impact of high-performance human resource management systems on firm performance in areas ranging from manufacturing [3,5,6] to services [7,8,10,61]. Therefore, in this study, we hypothesized.

H2

High-performance human resource management system has a positive impact on firm performance.

2.5. The effect of innovation on firm performance

Innovation capacity is considered the ability to create and search for new ideas, opportunities, knowledge, or internal and external [36,59,62,63] resources to apply them to the operating system. It will help enterprises create added values improve the competitive advantage [59,63]. Innovation activity depends on the ideas and creativity of employees in the enterprise. The process of perceiving and creating ideas is the first stage of innovation activities [64]. Innovative activities in enterprises are derived from employee behavioral innovation [12,65,66]. Therefore, in this study, we define innovation as the process of supporting the necessary resources to help employees form ideas and transform ideas into new products, services, or technologies [12]. Innovative behavior of employees has three stages: (i) Production of innovation cogitation (idea generation); (ii) promotion of innovation cogitation (idea promotion), and (iii) implementation of innovative behavior (idea implementation), including:

Production of innovation cogitation (idea generation): Refers to whether employees in the enterprise can come up with creative ideas to improve their work as well as come up with new solutions to solve problems, or simply find new ways of doing their works [29,38,67].

Promotion of innovation cogitation (idea promotion): Refers to the process of managers in the enterprise encouraging employees to enthusiastically come up with new ideas, and at the same time, managers with innovative thinking will accept ideas of employees easier [12].

Implementation of innovative behavior (idea implementation): Enterprises introduce and implement new ideas of employees into useful working methods and systems throughout the enterprise [12].

The innovation process of enterprises starts from changing the behavior of employees [12,65]. Employees' innovative behavior will help enterprises develop new ideas for change, thereby creating new processes, products, and services [29,68]. Consequently, when firms innovate, they increase profits compared to non-innovative firms [46,69]. Firms that innovate better are more successful in meeting customer needs and developing new capabilities to achieve better operational performance or even generate excess profits [70,71]. Economic development theory holds that when firms introduce new products to the market for the first time, it will limit the level of competition from competitors and thus can achieve relatively high profits [72]. However, this profit can be diminished over time due to imitation and competition in the market. But if enterprises continue to innovate and introduce new products to the market, high profits can be achieved over a sustainable period [73]. Many studies have shown that innovation has a positive impact on firm performance [31,41,46,74]. Therefore, in this study, we put forward the third hypothesis as follows.

H3

Innovation has a positive impact on firm performance.

2.6. Firm characteristics

There are different characteristics between enterprises that may affect firm performance. Therefore, in this study, we consider some characteristics of enterprises as control variables in our model. Three characteristics of enterprises include organizational size, (ii) organizational age, and (iii) the business sector, which considered entered in the model referring from previous studies [10,22,60].

• •Organizational size: In Vietnam, according to Decree No.39/ND-CP/2018 of the Vietnamese Government, small and medium enterprises have 200 or fewer employees [75]. Therefore, in this study, we considered dividing the enterprises into two groups: (i) small and medium enterprises group (Group I), and (ii) large enterprise (Group II) as dummy variables in the research model.
• •Organizational age: Some studies showed that the number of years in operation might affect firm performance. The arguments of the studies explaining this phenomenon are based on the assumption that the enterprises that have been in operation for longer may have more knowledge of the market customers, and as a result, they are able to take advantage of the market than other firms. Consequently, the enterprises with a higher number of years of operation may have better performance than others.
• •Business sectors: Enterprises from different business sectors exhibit distinct characteristics, and their profit margins also vary. Therefore, enterprises operating in different sectors may have differing performances. In this study, we examine two groups of enterprises: those belonging to the manufacturing sector and those in the service sector. To evaluate the impact of business domains on the business outcomes of these enterprises, we utilize analysis with dummy variables. In this approach, the group of enterprises within the manufacturing sector is coded as 0, while the group of enterprises in the service sector is coded as 1.

Based on the aforementioned arguments, it can be inferred that organizational size, organizational age, and business sectors potentially affect firm performance. Therefore, in this study, we formulate the following hypotheses.

H4

There is a difference in business performance between small and medium-sized enterprises (SMEs) (≤200 employees) and large enterprises (>200 employees).

H5

There is a difference in business performance between businesses operating for less than 10 years and those operating for more than 10 years.

H6

There is a difference in business performance between enterprises in the service sector and those in the manufacturing sector.

3. Methodology

3.1. Survey instrument

We used a structured questionnaire as a data collection instrument to analyze and test proposed hypotheses. The constructs in the study were adapted from literature and previous studies [6,8,12]. Specifically, high performance-human resource management system construct was the second-order construct with six components: training, job analysis, employee development, employee participation, performance evaluation, and job security [6,8,12]. In which, we used four items to measure training construct, which was referenced from stuty of Delery & Doty, Lu et al. and Zhang & Li [6,8,12]. Job analysis was evaluated by four items based on research of Delery & Doty, and Lu et al. [8,12]. Employee development was measured by four items that adapted from Lu et al. and Zhang & Li [6,12]. Employee development was measured by four items which adapted from Lu et al. and Zhang & Li [6,8,12]. Performance evaluation was assessed by two items adapted from Lu et al. and Zhang & Li [6,12]. Job security construct was adapted from Delery & Doty and Zhang & Li [6,8]. Innovation construct is also a multidimensional scale with three components: promotion of innovation cogitation, production of innovation cogitation, and implementation of innovative behavior [6]. Promotion of innovation cogitation was evaluated using three items, production of innovation cogitation involved three items, and implementation of innovative behavior was measured with three items adapted from Lu et al. [12]. Firm performance was measured by two components: operational performance and market performance. Operating performance consutruct was measure by three items based on from Lu et al. [12], and market performance was measured by five items adapted from Lu et al. and Zhang & Li [6,12]. All items were measured by a five-point Likert scale anchored by 1: strongly disagree and 5: strongly agree.

We translate the questions in the questionnaire from English to Vietnamese and use the back-translation method to ensure the equivalence of meaning in all items. A pilot survey was conducted with 30 enterprises to assess the relevancy of survey questions before the official survey. The analysis result showed that the Cronbach's Alpha coefficients were greater than 0.7, and three items did not reach reliability which has been removed from the questionnaire for the official survey. As a result, we kept 38 items described in Table 2.

Table 2.

The characteristics of the surveyed enterprises.

Categories		Organizational size
		50–200 employees	>200 employees
Business sectors	Service	138 (44.2 %)	65 (20.8 %)
	Manufacturing	78 (0.25 %)	31 (0.1 %)
Organizational age	<10 years	166 (53.2 %)	51 (16.3 %)
	>10 years	50 (16.0 %)	45 (14.4 %)

3.2. Sample and data collection

The sample size was determined to be 300, which reached a good level according to the rules of Comrey & Lee [76] for quantitative research. The sample size also satisfies various common sampling rules, for example, the rule of a minimum 5:1 ratio between sample size and the number of items in the questionnaire [77]. To ensure research ethics, the study conducted anonymous surveys with representatives of the enterprises, and the enterprise information was encrypted without disclosing the exact names of the companies in the survey. Survey respondents were informed about the research purpose, potential risks, and voluntary nature of participation. Additionally, participants were asked to sign an informed consent form, indicating their commitment to adhere to the research ethics principles of the research project. The research project was also approved by the Ethics Research Committee at QA Global JSC. The respondents were identified as senior managers or enterprise owners who have a comprehensive understanding of the firm's business activities. With the support of the General Department of Taxation, we dispatched over 2300 questionnaires to the addresses of CEOs of enterprises across all three regions: North, Central, and South. The survey was conducted from October 2019 until the end of January 2020. The questionnaire was designed with an initial YES-NO screening question regarding the implementation of a high-performance human resource management system in the enterprise. Enterprises responding “YES” proceeded to answer further survey questions, while the survey would conclude once an enterprise responded with “NO. Consequently, after four months of conducting the survey, we received 339 responses from the enterprises, resulting in a response rate of nearly 20 %. These questionnaires were carefully reviewed, and 312 valid questionnaires were selected for the official analysis. The characteristics of the surveyed enterprises were described in Table 2 as follows.

3.3. Data analysis method

We used multivariate data analysis methods to analyze surveyed data and test hypotheses. Because the items were adapted from previous studies, we used confirmatory factor analysis to evaluate reliability and validity [78,79]. All constructs in the model were multidimensional scales. Therefore, we used the measurement model and saturated model in CFA analysis. In the measurement, we assessed the component's reliability unidimensional, model fit indexes, convergent validity, and discriminant validity within constructs in multidimensional scales. In the saturated model, we checked model fit indexes, the convergent validity of each second-order construct, and discriminant validity between any multidimensional constructs. The criteria of confirmatory analysis, including Chi-square/df, is less than 3, CFI, TLI, IFI all are greater than 0.9, and RMSEA is less than 0.08 [[78], [79], [80]]. If the factor loading of items in the unidimensional scales is larger than 0.7, the scale achieves convergent validity [78]. To test discriminant validity, we used the criteria comparing between the square root of average variance extracted and any the correlations of constructs in the model or comparing 95 % interval confidence of correlation of constructs with one value. If the square root of AVEs is larger than any correlation between constructs or the 95 % interval confidence of correlation does not contain one value indicating that the constructs in the model reach discriminant validity [81]. We used structural equation modeling with statistical significance at a 5 % level to test the proposed hypotheses. To test the mediating role of construct in the model, we used the analysis by bootstrapping method with indirect effect coefficients [82,83].

4. Findings

4.1. Measurement model assessment

4.1.1. High-performance human resource management system

The result analysis from official data indicated that the model reached overall fit with the actual data: Chi-square/df = 2.452 was less than 3, CFI = 0.946, TLI = 0.934, IFI = 0.946 all were bigger than 0.9, and RMSEA = 0.068 was less than 0.08. The factor loading of CFA analysis was larger than 0.7, showing that the sub-constructs achieved convergent validity. The Cronbach's alpha and composite reliability coefficients were larger than 0.7, and AVEs were larger than 0.5, presenting that the components in the multidimensional construct are reliable (Table 3).

Table 3.

The convergent validity, and reliability of each construct in the model.

Constructs/Items	Statements	Factor loadings (CFA)	Cronbach's Alpha	Composite reliability	Average variance extracted
High–performance human resource management system
Training: Mean = 3.785, SD = 0.714
TRA1	The company's employees are trained every period	0.813	0.894	0.897	0.686
TRA 2	The company provides our employees with a formal training program to them for promotion	0.871
TRA 3	The company provides comprehensive training for employees	0.888
TRA 4	The company has formal training programs to teach new employees work skills	0.731
Job analysis: Mean = 3.961, SD = 0.657
JA1	There is a clear definition of employees work responsibilities	0.769	0.892	0.895	0.853
JA2	The responsibilities manual of employees includes all the responsibilities of employees	0.862
JA3	The employees can be in strict compliance with the specification of responsibilities manual in practical work	0.870
JA4	The company will promptly revise the responsibilities manual when necessary	0.795
Employee development: Mean = 3.759, SD = 0.698
ED1	The employees have a clear career path within the company	0.863	0.912	0.905	0.704
ED2	Direct superiors learn the occupation development intention of the employees	0.852
ED3	Employees have more than one suitable position to achieve promotion	0.853
ED4	There is career development for employees working in this company	0.787
Employee participation: Mean = 3.972, SD = 0.631
EP1	Managers and employees constantly make open and honest communication	0.827	0.791	0.800	0.574
EP2	Company employees have the opportunity to put forward the recommendation on improving the working methods	0.769
EP3	The managers of the company often make decisions referring to the views of staff	0.668
Performance evaluation: Mean = 3.948. SD = 0.684
PE1	The performance is usually measured as the objective and quantifiable results in the company	0.927	–	0.908	0.832
PE2	The employees' performance appraisal is based on the objective and quantifiable results	0.897
Job security: Mean = 3.759, SD = 0.698
JS1	It is challenging to dismiss an employee in this job	0.805	0.839	0.850	0.587
JS2	Job security is almost guaranteed to employees in this job	0.710
JS3	If the enterprise were facing economic problems, employees in this job would be the last to get cut	0.740
JS4	Employees in this job can expect to stay in the organization for as long as they wish	0.804
Innovation
Production of innovation cogitation: Mean = 3.969, SD = 0.634
PD1	The employees can create new ideas for improving the work	0.860	0.850	0.857	0.668
PD2	The employees are constantly searching out new working methods, techniques, or instruments	0.876
PD3	The employees can generate original solutions to problems	0.705
Promotion of Innovation cogitation: Mean = 4.033, SD = 0.635
PM1	The company mobilizes support for innovative ideas	0.864	0.881	0.885	0.720
PM2	Innovative thinking can acquire approval for creative ideas	0.875
PM3	The company makes important organizational members enthusiastic for innovative ideas	0.805
Implementation of Innovative behavior: Mean = 3.860, SD = 0.675
IM1	Innovative thinking can be turned into useful methods.	0.882	0.910	0.911	0.774
IM2	Innovative thinking can be introduced into work with the systematic method	0.897
IM3	The company will evaluate the effectiveness of innovative thinking	0.860
Firm performance
Market performance: Mean = 3.680, SD = 0.709
MAR1	The company's sales margins maintain a high level in the same industry	0.879	0.953	0.954	0.805
MAR 2	The company's overall performance is very good	0.914
MAR 3	The company's market share maintains a high level in the same industry	0.925
MAR 4	The company's competitive position is very favorable	0.885
MAR 5	The company's sales growth rate maintains a high level in the same industry	0.883
Operating performance: Mean = 3.655, SD = 0.701
OP1	Over the past three years, the company's return on assets has been higher	0.859	0.873	0.910	0.773
OP2	The company's sales margins maintain a high level in the same industry	0.839
OP3	The company's profit is higher	0.936

4.1.2. Innovation

The analysis with three innovation-construction components showed that the model reached overall fit with actual data: CFI = 0.972, TLI = 0.959, IFI = 0.972. The factor loading of each item in the sub-constructs was greater than 0.7 showing the sub-constructs achieved convergent validity. The Cronbach's alpha and composite reliability coefficients were larger than 0.7, average variance extracted is larger than 0.5, indicating that the components of innovation reached reliability (Table 3).

4.1.3. Firm performance

With firm performance, the findings presented that the model also achieved overall fit with the actual data: CFI = 0.972, TLI = 0.959, IFI = 0.972 all were greater than 0.9. The factor loading of items in each component ranged from 0.839 to 0.936 was larger than 0.7, showing the sub-constructs of firm performance reached convergent validity. The Cronbach's Alpha and composite coefficient of each component were greater than 0.7, and the average variance extracted was larger than 0.5, presenting that the firm performance construct was reliable (Table 3).

The result analysis presented that the square root of AVEs of each component in the high-performance human resource management system construct was greater than the correlation between sub-construct in the scale (Table 4). This noted that the sub-constructs in the high-performance human resource management system achieved discriminant validity.

Table 4.

Discriminant validity within in high-performance human resource management system.

Constructs	TRA	JA	ED	EP	PE	JS
TRA	0.828
JA	0.658	0.924
ED	0.754	0.660	0.839
EP	0.682	0.659	0.711	0.758
PE	0.615	0.604	0.706	0.602	0.912
JS	0.562	0.477	0.573	0.621	0.538	0.766

Notes: TRA is training, JA is job analysis, ED is employee development, EP is employee participation, PE is performance evaluation, JS is job security.

For the innovation construct, the analysis result indicated that the components of the innovation scale reached discriminant validity. Furthermore, the square root of AVEs of sub-constructs was larger than the correlation between any sub-construct (Table 5).

Table 5.

Discriminant validity within components in innovation.

Construct	PD	PM	IM
PD	0.817
PM	0.792	0.849
IM	0.798	0.811	0.880

Notes: PD is Production of innovation cogitation, PM is Promotion of Innovation cogitation, IM is Implementation of Innovative behavior.

For firm performance, the square root of two components was greater than the correlation between operation performance and market performance, presenting that the sub-construct in the firm performance reached discriminant validity (Table 6).

Table 6.

Discriminant validity within the component in firm performance.

Constructs	MAR	OP
MAR	0.897
OP	0.815	0.879

Notes: MAR is Market performance, OP is Operating performance.

4.2. Saturated model

The result analysis with saturated model (final model) also indicated the model achieved with actual data: Chi-square/df = 2.306 was less than 3, the indexes CFI = 0.917; TLI = 0.910, IFI = 0.917 all were larger than 0.9, and RMSEA = 0.065 was less than 0.08. The factor loading of each component in the multidimensional scale was greater than 0.7, showing that the model's second-order construct reached convergent validity. The composite reliability coefficients were larger than 0.7, average variance extracted was larger than 0.5, indicating that the second-order construct in the model was reliable (Table 7).

Table 7.

The reliability of the multidimensional scale.

Constructs/components	Factor loadings	Composite reliability	Average variance extracted
High performance – Human resource management system
Training	0.814	0.912	0.635
Job analysis	0.759
Employee development	0.875
Employee participation	0.836
Performance evaluation	0.783
Job security	0.701
Innovation
Production of innovation cogitation	0.891	0.925	0.804
Promotion of Innovation cogitation	0.898
Implementation of Innovative behavior	0.901
Firm performance
Market performance	0.959	0.901	0.821
Operating performance	0.850

The result analysis showed that almost the square root of AVEs of second-order constructs was greater than the correlation between any constructs in the model except the high-performance human resource management system construct. However, the 95 % confidence interval of the correlation did not contain value 1 (Table 8). Therefore, it concluded that the model's second-order constructs reached discriminant validity.

Table 8.

The discriminant validity of second-order constructs in the model.

Construct	HRM	INO	FP
HRM	0.797
INO	0.840 (0.752–0.908)	0.897
FP	0.644 (0.549–0.726)	0.584 (0.486–0.683)	0.906

Notes: HRM is the high-performance human resource management system, INO is innovation, FP is firm performance.

4.3. Structured model and hypothesis testing

We used structural equation modeling analysis with five models to test the proposed hypotheses (Table 9). The finding indicated that the Chi-square/df of each model all were less than 3, the indexes as CFI, TLI, IFI all larger than 0.9, and RMSEAs were less than 0.08 showing the model achieved with actual data.

Table 9.

The result of SEM analysis.

Relationships			Model 1	Model 2	Model 3	Model 4	Model 5
HRM	→	INO	0.858*(11.107)	0.858*(11.108)	0.858*(11.109)	0.858*(11.109)	0.858*(11.112)
INO	→	FP	0.628*(9.659)	0.628*(9.691)	0.628*(9.707)	0.627*(9.701)	0.625*(9.810)
Business sectors	→	FP	–	−0.051 (-1.021)	–	–	−0.081 (-1.656)
Organizational age	→	FP	–	–	0.066 (1.325)	–	0.076 (1.548)
Organizational size	→	FP	–	–	–	0.111*(2.232)	0.115*(2.343)
Model fit indexes			Chi-square/df = 2.330, CFI = 0.915, TLI = 0.908, IFI = 0.916, RMSEA = 0.065	Chi-square/df = 2.272, CFI = 0.914, TLI = 0.908, IFI = 0.915, RMSEA = 0.064	Chi-square/df = 2.280, CFI = 0.914, TLI = 0.907, IFI = 0.914, RMSEA = 0.064	Chi-square/df = 2.273, CFI = 0.914, TLI = 0.908, IFI = 0.915, RMSEA = 0.064	Chi-square/df = 2.203, CFI = 0.910, TLI = 0.904, IFI = 0.911, RMSEA = 0.062

Notes: Business sectors (Manufacturing = 0, Service = 1), Organizational age (<10 years = , >10 years = 1), Organizational size (<200 labors = 0, ≥ 200 labors = 1), *p < 0.05.

The estimation results indicated that innovation was positively impacted by high-performance human resource management practices (β > 0, p < 0.05). Firm performance was positively affected by innovation (β > 0, p < 0.05). Meanwhile, firm performance was only indirectly influenced by high-performance human resource management practices. The results also revealed that variables such as business sector and organizational age had no impact on firm performance. However, firm performance was positively influenced by organizational size. Specifically, firms with a size of over 200 employees demonstrated higher firm performance than those with fewer than 200 employees (β > 0, p < 0.05). In other words, hypotheses H1, H3, and H4 were supported, while hypothesis H2, H5, and H6 were rejected (Table 9). The relationship between high–performance human resource management systems, innovation, enterprise characteristics, and firm performance can be described in Fig. 1.

Fig. 1.

Result of structure model.

4.4. Mediation analysis

To analyze the mediating role of innovation in the relationship between high–performance human resource management practices and firm performance, we used an indirect effect coefficient using the bootstrapping method. The analysis results indicated that the indirect effect coefficient γ = 0.536 > 0 and p-value = 0.002 < 0.05 (Table 10). This proves that the high-performance human resource management system had an indirect impact on firm performance, or innovation is a mediator variable between high-performance human resource management system practice and firm performance.

Table 10.

Direct, indirect, and total effect coefficients.

Dependent variables	Effect	Size		HRM		INO
		Effect	p-value	Effect	p-value	Effect	p-value
INO	Direct	–	–	0.858	0.002	–	–
	Indirect	–	–	–	–	–	–
	Total	–	–	0.858	0.002	–	–
FP	Direct	0.115	0.028	–	–	0.625	0.003
	Indirect	–	–	0.536	0.002	–	–
	Total	0.115	0.028	0.536	0.002	0.625	0.003

5. Discussion

This study examines the relationship between high-performance human resource management systems, innovation, and firm performance. The research model is developed and expanded from the theory of a high-performance human resource management system integrated with job security factors and business characteristics [8,12]. The analysis results by CFA and SEM show that the constructs in the development model have both reliability and validity. The study also confirms the influence of high-performance human resource management systems on innovation and firm performance. Finally, the study also offers implications for the academy and practitioners.

The study recognizes the direct effect of a high-performance human resource management system on innovation and the indirect impact on firm performance. This shows that the practice of a high-performance human resource management system plays a vital role for enterprises. Enterprises that practice a high-performance human resource management system often focus on improving training activities and improving employees' skills and working capacity to help them improve work performance and improve firm performance. In addition, practicing a high-performance human resource management system also requires enterprises to perform job analysis and job evaluation to organize and set up effective labor arrangements towards promoting job security and employee development. Ensuring job security helps employees feel more at ease with their work, boosting their satisfaction with the organization and potentially leading them to put more effort into their tasks to enhance individual performance. Some studies indicate that job insecurity can decrease job satisfaction and demotivate work engagement [84]. The implementation of high-performance human resource management systems aims to enhance outcomes by increasing employees' work motivation. Meanwhile, numerous studies affirm that fostering positive work motivation can improve individual work performance [60,85]. This indicates that the practice of high-performance human resource management within enterprises can amplify skills, motivation, innovation, and overall enterprise performance. As a result, when enterprises practice a high-performance human resource management system, employees will be willing to come up with new ideas, apply ideas, and improve skills and work performance regularly, which improves firm performance. This is also confirmed in previous studies on the relationship between high-performance human resource management and innovation and firm performance [8,12,19,27]. For example, research by Delery & Doty [8] shows that high-performance human resource management practices directly impact the performance of banks in the US. Several subsequent studies also confirmed this result in the Chinese market [6,12,22], Israel [27]. This result also implies that to promote innovation and improve firm performance, enterprises can build and operate a high-performance human resource management system as a preferred solution in the market context, which is increasingly under competitive pressure as it is today.

Unlike some prior studies [6,8,12,27], our research does not uncover a direct impact of high-performance human resource management practices on firm performance. This discrepancy necessitates a more comprehensive discussion to be adequately elucidated. The disparity in our research outcomes may arise from the contextual differences between our study and others. Previous investigations were conducted in developed nations like the United States [8] and Israel [27], countries with longstanding traditions of prolonged free-market economic practices. Conversely, China embraced openness and economic integration prior to Vietnam, a transformation that began more than a decade ago under Deng Xiaoping. Consequently, enterprises operating in these markets were more extensively introduced to, applied, and adept with modern management systems. Some insights from studies suggest that societies functioning under free-market systems earlier tend to exhibit more market-oriented and adhocracy (innovation) in their corporate culture. In contrast, the later economic opening and delayed implementation of free-market regulations in newly emerging economies like Vietnam resulted in these enterprises having a delayed timeframe for adopting modern management systems. Indeed, Vietnam truly integrated into the global economy upon becoming a member of the World Trade Organization (WTO) in 2007. Although there have been shifts in business management practices from the Soviet-style planned system to modern management, observations indicate that Vietnamese enterprises still embody hierarchical and clan culture attributes in their corporate culture, rather than adhocracy or innovation [86]. As a result, decision-making in these businesses tends to be sluggish, leading to reduced efficiency in implementing high-performance human resource management practices that contribute to firm performance.

We also recognize the precise impact of innovation on firm performance. This is explained by firms oriented to innovation activities that often promote the generation of creative ideas from employees [[64], [65], [66]]. Innovation-oriented enterprises often create a creative atmosphere for employees and lead them to change their behavior by business requirements [87]. Innovation orientation also helps enterprises create positive emotions among employees [88], leading to motivation for employees to adopt innovative behaviors [89]. When enterprises create a positive atmosphere with new ideas and initiatives, it will motivate employees to develop ideas and initiatives that can be applied at work. As a result, enterprises receive more ideas and initiatives that could be put into operation [12,29,38]. These valuable ideas and initiatives can be put into operation to be implemented into new processes and products [29,68]. These new products and processes can create more competition than less innovative firms [72]. Therefore, innovative products will help firms achieve better firm performance and generate superior profits by better meeting customer needs [46,73]. In addition, innovation also creates significant improvements to help enterprises survive and grow quickly as well as increase efficiency compared to non-innovative firms. As a result, innovative enterprises will create a competitive advantage and better firm performance than non-innovative firms [69,72]. This also implies that innovation will create a competitive advantage and improve firm performance. In other words, enterprises that want to make their own advantages need to pay attention to innovation activities in a competitive environment. At the same time, to create an innovation-oriented enterprise, it is also necessary to build a high-performance human resource management system.

We observed differences in business performance among firms practicing high-performance human resource management, categorized by labor size. Particularly, larger enterprises (>200 employees) had better business performance than smaller enterprises. This result closely resembles certain previous studies indicating that larger firms possess more advantages and achieve higher business performance [10,60]. This result suggests that high-performance human resource management practices can yield distinct business performance with different groups of enterprises of varying sizes. Larger enterprises typically exhibited better results, potentially due to their possession of greater resources and more efficiently decentralized organizational structures, which facilitated the smoother implementation of the high-performance human resource management system. Consequently, these enterprises enjoyed a competitive edge in the market, leading to improved business performance. In contrast, smaller enterprises, though equipped with a human resource system, often faced resource limitations that might hinder their ability to execute activities associated with a high-performance human resource management system, leading to diminished results. This situation could explain the superior business performance of larger enterprises over smaller ones in our study. However, this also implies that effectively implementing a high-performance human resource management system requires enterprises to possess appropriate resources, organizational systems, and corporate culture.

Contrary to initial expectations, we found no difference in firm performance in terms of operational age in this survey. This result is different from some previous research results, such as the study of Sun et al. [10] found the influence of operational age on firm performance in hotels in China, but similar to the results of Zhang & Morris [22] in many fields also in China. This implies that firm performance depends more on the operating systems of the enterprises than the number of years they operate, especially in the context of increasingly rapid market fluctuations as today. Firms that have established a well-functioning system can generate business advantages that surpass the operational experience of firms with little innovation.

6. Implication

6.1. Theoretical significance

Our study also provides important theoretical implications on the mechanism of influence of high-performance human resource management systems on enterprises' innovation and firm performance. Our main contribution is to develop a model that evaluates high-performance human resource management systems on innovation and firm performance. The model is extended from research model of Lu et al. [12] and incorporate the component of job security [8], an important component for workers in developing countries. Our model explains spreading the influence of a high-performance human resource management system on innovation in employee behavior and affecting firm performance. Specifically, our main theoretical contributions are as follows.

• •Our study extends the assessment model of the impact of high-performance human resource management practices on innovation and firm performance. We have integrated and confirmed that job security is a vital component in enacting high-performance human resource management systems within enterprises in developing countries like Vietnam.
• •The study also expands the examination of the influence of a high-performance human resource management system on innovation and firm performance in terms of controlling for factors such as size, number of years activities, and business sectors.
• •The results of analysis by CFA, SEM showed that the scales calibrated in this study are reliable and appropriate. Furthermore, it shows that extending the theory of high-performance human resource management systems in the context of businesses of countries like Vietnam can be applied to emerging markets and other developing countries.

6.2. Practical significance

The study also brings useful, practical implications for business organizations in developing human resources or applying high-performance human resource management systems in enterprises. This study provides a frame of reference on the influence of high-performance human resource management systems on innovation and firm performance. Human capital is always considered one of the most important assets of a firm [90]. Therefore, in this study, we suggest that for enterprises to apply a high-performance human resource management system successfully, it is necessary to pay attention to a few points as follows.

• •Firstly, enterprises need to establish a comprehensive set of high-performance human resource management systems before deploying them on an enterprise-wide scale. The established system needs to ensure uniformity in how policies allow employee participation in decisions, development of capacity and skills training programs, etc.
• •Secondly, enterprises need to set up human resource development programs to form a market-oriented and innovative corporate culture. Research results show that innovation has a powerful influence on firm performance. Meanwhile, innovation is only possible in firms with more market and innovative characteristics than hierarchical or clan culture. Therefore, the development of human resources must be associated with cultural development with many market and creative orientations to increase the innovation capacity inside the enterprise.

7. Limitation and future study

Like any other studies, our study also has some limitations. First, the study utilized convenience sampling, which may affect the representativeness of the study. Therefore, in the future, other studies can use random sampling methods such as stratified sampling and cluster sampling to increase the representativeness of the study further. Second, our study was conducted at a point in time, so we could not track the effect of high-performance HRM systems on innovation and firm performance over time. Therefore, we encourage further studies designed as time-based studies to evaluate the impact of high-performance human resource management systems on innovation and firm performance over time.

8. Conclusions

Our study provides evidence that the practice of high-performance human resource management has an influence on innovation and creativity within enterprises, indirectly contributing to improved firm performance. This research outcome also confirms job security as a crucial component within high-performance human resource management systems in developing countries. The research findings also hold significant implications for promoting the adoption of high-performance human resource management practices within enterprises, aimed at enhancing creativity and improving firm performance.

Data availability statement

Data will be made available on request.

CRediT authorship contribution statement

Duy-Thanh Nguyen: Writing - review & editing, Writing - original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Trung Kien Dao: Writing - review & editing, Writing - original draft, Validation, Software, Methodology, Formal analysis.

Declaration of competing interest

To: The Editorial Board of Heliyon,

On behalf of the authors, I confirm that.

• •We have no conflicts of interest (direct and indirect) with individuals or organizations acting as research participants in this research project.
• •Our research project has been reviewed and approved by the scientific ethics board at QA Global before conducting formal surveys.
• •The authors also have no conflicts of interest with the Editorial Board of the journal that could bias the consideration and evaluation of research results for publication.