Amsterdam Business School
The effect of cultural values on the relation between performance
measure participation and performance measure quality: A
cross-cultural analysis of Chinese and Dutch firms’ practices
Name: Yang Yuxi
Student number: 11082054
Thesis supervisor: Bianca A.C. Groen Date: 10 January 2017
Word count: 12182
MSc Accountancy & Control, specialization [Accountancy] Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student [YUXI, YANG] who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
Existing literature indicates that employees’ participation in performance measure system is positively related to performance measure quality. This study investigated whether, under the effect of various cultural values, the intensity of positive association between employees’ performance measure participation and performance measure quality will vary between Dutch and Chinese employees. Based on survey responses from 80 pairs of managers and low-level employees, results of this study offered findings that differ from prior research and suggested that cultural values had no effect on the relationship between performance measure participation and performance measure quality. Finally, the study explained the unexpected results from different perspectives and put forward insights for further studies.
Contents
1 Introduction ... 5
2 Theory and hypotheses development ... 8
2.1 Performance measure participation ... 8
2.2 Performance measure quality ... 10
2.3 Performance measure participation and performance measure quality ... 11
2.4 Cultural dimensions theory ... 13
3 Methodology ... 18
3.1 Data collection and sample size ... 18
3.2 Measurement of variables ... 19
4 Result ... 23
4.1 Preliminary Analysis ... 23
4.2 Assumption testing ... 24
4.3 Hypotheses testing ... 24
5 Conclusion and Discussion ... 27
6 Directions for future research ... 32
References ... 33
1 Introduction
The blooming globalization during past several decades has turned the whole world into an integrated system. There are different components in this system, such as business, politics, and culture, and those components are indispensable and closely linked with each other. Meanwhile, multinational firms play a central role in accelerating the economic globalization and regional economic integration (Cooper and Ezzamel 2013). A growing number of firms competes and cooperate in areas around the world that are quite distant from their traditional strongholds, both geographically and culturally (Franke and Nadler, 2008) Because of the constantly changing business environment, multinational firms now rely more on management control system to set and implement organizational strategies. That creates a pressing need to understand how management practices work across different nations and cultures when companies want to develop competitive advantages in the global market, especially one of the most attracting and important aspects of management practices: performance measurement. Among various control mechanisms, performance measures are used to monitor the strategy implementation and detect the deviation of employees’ performance (Simon, 1995). Many studies have already responded to the call via investigating whether there are cross-cultural differences in management practices and, if so, what’s the source of the differences. (Jansen et al., 2009, Hexter and Woetzel, 2007). Those studies document that culture has distinctive effects on how people work, what they prioritize and value from various perspectives (Berman et al., 2013), e.g. adoption of key performance measure indicators, reliance on financial incentive compensation and evaluation of stakeholder’s needs.
Existing literature indicates that the quality of performance measurement is better when employees take part in the performance measure development process, in other words when employees have substantive effect on performance measures, and those investigations also identify a list of factors that might increase or decrease employees’ performance measure participation, such as attitude, social pressure and capacity (Abernethy and Bouwens, 2005, Hunton and Gibson, 1999, James et al., 2006). At the same time, low-level staff’s influence are often neglected in those studies, for researchers are more inclined to investigate the association between senior management and performance measure quality. In addition to this, most of those findings are based on western background. Evidence that could draw the same conclusion under Asian setting is insufficient and needed. Last but not least, researchers who conduct cross-cultural studies usually focus on comparing specific performance measures between different
nations, little is known about whether cultural characteristics have an influence on the relationship between performance measure participation and performance measure quality.
In order to fill in the blank, the study first examines whether the extent to which employees can affect performance measures is related to the performance measure quality in both Chinese and Dutch firms, then explains the influence of cultural values in the relation. This paper aims to contribute to the research area of performance measurement by shedding light on the moderate effect of cultural values on the relationship between employees’ performance measure participation and performance measure quality in different nations.
There are three main reasons that I choose Chinese and Dutch firms as the resource for my study. First of all, the sharp contrast of performance measures between China and the Netherlands. One the one hand, due to the fact the China has initiated open and reform policy for less than 40 years, the performance measure system in Chinese firms is relatively simple and more subjective, while the Netherlands, one of the most representative European countries that have opened for long history, has more sophisticated, metric-based performance measures. (Merchant et al., 2009). On the other hand, the systematic performance measurement has been introduced to Chinese firms at an accelerating pace because of the booming economy (Chow et al., 2007, Fleming et al., 2009). Hence, it will be meaningful to reduplicate the investigation focusing on the relationship of performance measure participation and performance measure quality in China, to see if findings will further strengthen the above argument or provide different opinions.
Secondly, the national cultural characteristics of China are found to be distinctively different from those of the Netherlands (Hofstede, 2001). As a representative maritime country, the Netherlands, locating in the North Sea, has plenty of good harbors. Given the uniquely advantageous geographic conditions, people conducted commercial activity through the sea early. Meanwhile, since the Netherlands is a tiny country, which is known as the “low countries” because much of the land is below sea level, people seek for overseas development more actively. As a result of that, Dutch people’s culture psychology is more extroverted, and the culture system is always in a state that is dynamic and fluid. As the culture is based on the marine competition, the Netherlands’s culture identity is full of the spirit of free competition and speculation (Emmeline, 2010).
By contrast, even though China also has long coastlines, the vastness of the land, the rich natural resources, and the agreeable climate make the mainland a better place to live, providing a good production and living condition. Hence the mainland is the birthplace of Chinese civilization and
where Chinese culture prospers. The core of China's agriculture centers on the closed Yellow River basin and then evolves into the core of China’s business and traditional culture (Liu, 2011). Because of the relatively geographically isolated environment and autarkic, economy, Chinese people have been complacent since ancient times.
Last but not least, China and the Netherlands, one is the biggest developing country, and another is one of the most important developed countries in the world, have even more reasons to cooperate with each other. Considering the overall influence that the Chinese and Dutch firms have on the global market, it would be interesting to find the relationship between cultural values and the performance measure practices of powerful local or multinational organizations. Business cooperation can usually be found between Chinese firms and Dutch firms and many famous Dutch firms now have branch companies in China while Chinese firms start an investment boom in Europe recently. Therefore it’s necessary to investigate the effect of cultural values on their management practices, a better understanding of how different culture structure facilitate employees’ participation can be indicative for multinational corporations and those organizations that plan to join the foreign market, since the result of this investigation will provide them with useful information about constraints and characteristics of different cultures. Apart from the potential suggestions for organizations, this research also tries to make a contribution to the development of Hofstede’s (2001) cultural dimensions theory in the area of performance measurement. Due to the inherent difficulties of conducting the cross-national studies, for examples: find suitable sample and research method, research progress in this area has been slow, most existing research turn to search other determinants and influences of performances measures or investigate a single value’s effect in relative small geographic area, very few researchers use cross-national study to investigate the relationship between cultural values and employees’ participation in performance measures. Hofstede’s theory provides researchers a direction and framework to give inferences, however verifying the theory and testing the hypothesis need the supplement of empirical practices.
The rest of this paper is organized as follows: section 2 first provides a theory about performance measure participation, performance measure quality and cultural values, then illustrates the hypothesis development, section 3 describes the methodology, sample selection and variable measurement and section 4 represents the empirical result. Finally, section 5 discusses the theoretical implication and research limitation of this study.
2 Theory and hypotheses development
2.1 Performance measure participation
As one of the most common and influential management techniques around the world, performance measurement systems, information systems that bond employees’ activities with reward and punishment are used to help firms and employees to achieve their organizational and individual objectives. Employees’ participation in the performance measurement system is positively associated with the desirable performance measurement outcomes (Roberts, 2003.). Researchers’ effort on investigating the cognitive mechanism and utility of performance measure participation has presented a mixed outcome. On the one hand, fruitful findings of the influence and functions of participative performance measurement are provided. Performance measure participation has always been a key topic in management accounting studies and information system (Burney and Matherly, 2007; Cawley et al., 1998; Hunton, 2002). Among information system studies that focus on Performance measure participation, Banker and Kauffman (2004) found a link between managers’ involvement in the performance measurement system and their subsequent user satisfaction. Hunton (2002) explored the importance of participation of performance measurement in improving system acceptance. While management accounting studies concentrate on the association between participation and performance, Chenhall and Brownell (1988) found that budgetary participation can improve job performance by providing job-related information, Kren (1992) proved that managerial participation positively interacted with job-related information.
On the other hand, several features indicate that a consensus on the inherent meaning and mechanism of Performance measure participation is absent. First of all, although there is a growing number of studies that focus on Performance measure participation, most of those studies adopt managers’ perspective to do the conceptual and operational analysis. As a result of that, even though senior employees’ role in performance measurement has been acknowledged by most management accounting scholars, seldom people try to link junior employees with Performance measure participation. However, the fact that low-level employees’ effect on performance measurement has not received adequate attention does not mean that low-level employees should be excluded from performance measure participation. Kaplan and Norton (1992, 1996) argued that only when low-level employees and managers reach the consensus on corporate strategies and performance measures, those strategies and performance measures can be implemented successfully. Beside this, Kren (1992) found that job-related information can
improve employees’ apprehension and use of performance measurement system and Groen (2012) believed that low-level employees’ participation could gain them more job-related information. In order to gain employees’ understanding of the performance measurement system and ensure the implementation of corporate strategies, companies should encourage low-level employees to take part in performance measurement so that employees and managers are more likely to reach consensus.
What’s more, recent studies pay more attention on the employees’ participation in goal setting and budget making (Kren, 1992; Eker, 2009), two areas with significant importance in performance measurement systems. However, despite of budgetary situations, environmental factors, employee skills, and style of leadership also have a major impact on performance (Chalos and Haka 1990). Hence, actually performance measure participation has a much more extending scale rather than limits in financial performance measures. It’s important to point out that the whole performance measurement system is neither static nor unitary. Performance measurement systems represent a logical bridge that connects management accounting and information systems since these functions are closely bound up with performance measurement systems design, implementation, use, and maintenance (Burney and Matherly, 2007). Actually, to serve as translators that transform employees’ behavioral data into assessments of organizational and individual performance, performance measurement systems have their own operating cycles, which consist of four main stages: a) PM indicators design phase, b) testing and implementation phase, c) self-assessments and modification phase, d) ongoing feedback and revising phase (Zuriekat et al., 2011).
Based on the argument above, employee’s performance measure participation can be reflected in any stage within the cycles. Furthermore, Groen (2012) recognizes that performance measure participation involves substantial influence (rather than only involvement) of employees on the content of the performance measures used to measure their performance. This article will follow the definition of Groen (2012).
Early research identified two factors that contribute to performance measure participation: ownership and contribution (Greller, 1978). Ownership refers to the extent that employees’ opinion was welcomed and appreciated, while contribution measures the extent of influence to which employees feel themselves have in the performance measurement system. According to the two factors, employee participation can be observed in several ways, for example: whether employees obtain right to voice their opinions on specific issue, whether employees can set goals
for themselves and whether employees believe that they can impact the performance measure’s design or implementation in any way (Cawley et al., 1998).
The motivation of introducing employee’s influence on performance measurement systems consists of four components. First of all, knowledge spread. According to the agency theory, information asymmetry inherent in the subordinate-supervisor relationship impairs the organizational operation efficiency. Given the opportunity to voice their own idea in the performance measurement system, employees are capable of sharing specific knowledge regarding their working experience with higher level managers. As a result of that, empowerment moderates the adverse effect of information asymmetry and improve working efficiency. Secondly, the accumulation of knowledge will not only benefit the performance measurement system but also enable employees to learn from peers and gain growth during the process, which leads to the establishment of self-growth culture. Thirdly, the implementation of a new performance measure will face strong resistance when employees lack sufficient understanding of it. In other words the more participation of employees within the PM cycle, the more they understand the function of this cycle and the fewer resistance employees will express with regard to those measures. Furthermore performance measure participation is believed to inspire employees’ working initiation; employees are expected to pour more passion in work and tend to set higher performance goals than management when they possess the certain autonomy and resource support (Robert, 2003).
2.2 Performance measure quality
Performance measurement is seen as an essential and evitable part in comprehensive management control system, which driven by firm’s need to mitigate information asymmetry and interest confliction between principles and agents.
According to the framework of Merchant&Van der Stede (2012), except for the information asymmetry, employees’ actions are usually not perfectly aligned with firm (managers) objectives when they: a) lack of direction, employees don’t know what is expected of them; ; b) lack of ability, employees may be unable to perform well because of any of a number of other limitations, some of them person-specific; c) lack of motivation, individual and organizational objectives do not naturally coincide since individuals are born to be self-interested and will sometimes act in their own personal interest at the expense of their organization’s interest.
Therefore, in order to address the above problems, an effective performance measurement that can response to the deviation of employees’ action and incentivize employees to contribute to firm value at the same time becomes the pillar of firm’s management control system.
Moers (2006) puts forward three properties to evaluate performance measure quality: (1) sensitivity, the extent that the performance measure is influenced by employee’s actions, performance measure with high sensitivity means that the measurement outcome is closely associated with employees’ behaviors so that the appropriate reward and punishment can be linked with those behaviors. Employees are better informed with manager’s expectation for them via the outcome of a measurement. (2) Precision, the extent that performance measure is influenced by factors out of employee’s control, if the excessive effect of outside factors is found on the performance measures, the outcome will lose credibility since employees are measured by noisy criteria. An outside factor can magnify or lessen employees’ job performance, for example, employees might achieve abnormal excellent performance with a strong market and unusual bad performance with a recession. (3) Verifiability, the extent that the process of performance measures evaluation is in a transparent and fair environment. This requires managers disclose the detail of measurement process to every employee instead of just announcing the result without any explanation, and employees should have the right to question the result if they think it is biased or unacceptable.
Furthermore, based on the perspective of managers, performance measures quality can be defined as the extent to which managers find the performance measures to be sensitive to the action of their employees, precise in measuring relevant aspects of their employees’ performance, and verifiable (Groen, 2012).
2.3 Performance measure participation and performance measure quality
As noted before, the information advantage enables employees to bring benefits to organization’s performance measures when employees cooperate in the stages of performance measure cycle. Former research believes that those benefits can enhance performance measure quality since they can transfer to the three properties of performance measure quality in certain ways. Moreover, in order to best reflect user requirements, employees who gain a better understanding of the system through participation will enhance the performance measure quality by selecting a range of performance measures that will maximize employees’ demand (Zuriekat et al., 2011). With regard to performance measure quality measurement, existing literature witness two different kinds of methods, one stands for the appraisement of employees, and the other listens to the perspective of managers. This article will employ managers’ evaluation of current
performance measure quality in respective firms since I consider the fairness of managers’ answers will be more credible.
Before I illustrate how performance measure participation leads to higher performance measure quality, it’s important to articulate that the three properties of performance measure quality are not isolated but interact with each other. For example, a transparent process can prevent the overuse of manager’s subjectivity, such as favoritism, and enhance the precision. Meanwhile, precision is an indispensable precondition of sensitivity. When performance measures receive the excessive effect of outside factors, such as macroeconomic condition and specific criteria, employee’s behaviors have such an insufficient impact that performance measures fail to respond timely and fairly. Finally, such performance measures will turn out to be useless because they can neither measure employee’s performance nor stimulate employees to achieve organizational objectives.
However, introducing performance measure participation to performance measure cycle can weaken the noisy resulted from outside factors and further strength the relationship between precision and sensitivity. Obviously, employees have sound understanding of the factors that might influence their performance measure outcome, in order to maximum their performance measure outcome, employees will try to minimize the uncertainty of performance measures as possible as they can, in other words, employees will devote more effort into ruling out the outside factors if they get the opportunity to affect the material part of performance measures. Nevertheless, the performance measures can also lose the precision if employees’ influence is overstated, which means employees will abuse their influence to mitigate or even eliminate some measures that are essential to ensure the completion of organization objectives but might impair their job performance. For example, they might set the goals low enough to get incentives easily and because of conflict interest they might favor the measures that reflect personal goals instead of organizational goals. Based on this situation, managers’ role is vital because managers have to perform as impartial judges who evaluate the extent of performance measure quality with the primary criteria that whether the performance measure motivates employees to work for organizational goals.
Given the joint effect of employees’ performance measure participation and managers’ fair evaluation, the sensitivity of performance measure is expected to be enhanced. Employee’s information advantage can also be viewed as a reason that increases the sensitivity. In addition to that, employees will grasp a deeper understanding of what their firms is expected on them following the continuous participation and are more likely to make self-growth. Along with
manager’s instruction, employees are suggested to perform better and hence improve performance measure quality.
As for verifiability, a measure without transparent information and procedures is a measure without credibility, reasonable performance measures should provide unquestionable evaluation metrics and undoubtedly information processing, which is asked by employees throughout the whole performance measure cycle. Since employees’ insistence towards the verifiability, they are expected to maximize the fairness and transparency in every detail of performance measures once they can participate in designing and implementing performance measures. Finally, their insistence will translate into the improvement of performance measure quality.
Considering the reason listed above, I put forward the first hypothesis:
H1: Performance measure participation is positively associated with performance measure quality.
2.4 Cultural dimensions theory
National culture has the significant influence on the design and enforcement of management control system (Chow et al., 1999; Harrison, 1992; Tsui, 2001, Newman & Nollen, 1996, Schneider, 1988). Such influence is usually represented in the performance measurement systems, incentive compensation practices and organization strategies (Jansen et al., 2009; Merchant et al., 2011; Van der Stede, 2003; O’ Connor, 2004).
Chow et al. (1999) find that both Japanese and American companies modify their original management control system (MCS) design to fit local culture characteristics when they operate in Taiwan, Merchant, et al. (2011) argue Chinese firms are more likely to provide incentive compensation to their managers than American and Dutch firms, furthermore Li and Tang (2009) demonstrate that although there is an increasing number of Chinese companies adopting western management accounting system, the existence of pervasive political constraints, to some extent, offset the expected effect of these management control techniques. According to those findings, two inferences can be concluded: first of all, the use of management control systems has extended to China with an accelerated pace, the next, management control system is not a “one size fits all” technique and some distinctive cultural characteristics of Asian-block may render an effective western control practice less attractive to the local companies. As one of the most widespread management control systems in China since the opening-up, performance measurement system is expected to be a focal point that reflects the cultural differences between Asian-block and Western block.
Although a number of studies have been conducted to prove the positive relationship between performance measure participation and performance measure quality, most of those studies are based on western context, which raises the question that whether the result is transferable in the cross-cultural analysis, especially the analysis focused on Asian block. At the same time, as one of the most representative Asian countries, China has always been the breakthrough point when researchers try to conduct a cross-nation field study comparing the difference between Asian-block and Western Asian-block. Tsui (2001) questions the role of budgetary participation in non-western environment and argue that participation in Chinese firms may be less functional or even dysfunctional in China. Based on the former discussion, I still suggest a positive relationship between performance measure participation and performance measure quality in Chinese firms. However, the strength of the upward association may differ between Chinese firms and western firms because of the influence of national values.
Before testing and verifying the influence of culture values in different cultural backgrounds, it’s necessary to compare how the influence is introduced into management practice in China and a typical western country; here I choose the Netherland as the contrast. Therefore in this part, I will focus on demonstrating how national cultures moderate the relationship between performance measure participation and performance measure quality in the two countries. The conceptual framework of this part derives from cultural dimension theory, which is put forward by Hofstede, a theory that is most commonly used in the cross-culture study.
As we all know that economic base, ideological belief system, moral heritage, history and other diverse factors together shape a nation’s distinct culture (House and Javidan 2004, p. 15). Undoubtedly, specific national culture can significantly affect employees’ recognition and attitudes toward performance measure participation and further the performance measure quality (Chong and Jerome, 2010) and Hofstede’s cultural dimensions theory provide researchers a way to track, classify and measure the extent of such effect.
Based on the cross-nation survey, Hofstede (1980) identified five culture dimensions that embed in every society and can be used to test the cultural differences. The five dimensions are Power distance (PDI), individualism (IDV), uncertainty avoidance (UAI), masculinity (MAS) and long-term orientation (LTO). Recently a new dimension is added to the original classification: Indulgence. (Hofstede 2011; Minkov and Hofstede, 2011) With the order of relevance, power distance and individualism are most suitable to explain the presumptive differences between China and the Netherland because of the inherent meaning of this two dimensions.
Power distance (PD) reflects the extent to which society members accept an unequally distributed power system, where hierarchy that not only relates to administrative management but also personal status is strictly defined and operated in social organizations. Power distance has a profound effect on determining the degree of centralization, the acceptance of inequality and the amount of organizational participation in society (Nollen, 1996). An organization with high power distance tends to allocate power based on people’s administrative positions, a top-down hierarchy. Under such circumstance, subordinates are supposed to listen to their supervisors automatically, rely on supervisors’ instruction passively and react to participation reluctantly (Harrison, 1992),
they hardly participate in decision-making process actively and when managers ask them for opinion, they are likely to response in a mild and incomplete way because they are afraid of making mistakes, furthermore if employee’s opinion is against manager’s, subordinates might choose to keep silence or even distort their true idea to adhere to manager’s view. Such feeling of fear and distrust decrease not only the amount of performance measure participation but also the quality and usefulness. What’s more, it’s still managers who have the absolute decision right to adopt the opinion or not in this context; they are likely to pick those favorable opinions and ignore the rest. Meanwhile, Hofstede (1984) argued that superiors who encourage subordinates participate actively and consult employees’ advice in an equal way will be regarded as lack work capability in a high power distance society. Therefore, in a high power distance oriented society, employees tend to express themselves reluctantly, mild, incomplete and even distorted, while managers are less likely to treat subordinates’ opinion fairly, those limitations restrict performance measure participation’s effectiveness in promoting performance measure quality. On the contrast, subordinates in low power distance companies are seen as dialogists equal with their superiors; they are promoted to voice their opinion without feeling restrained by their current positions. In this situation, the hierarchy is only used for administrative convenience while the personal relationship is not constructed by position title, and people are treated as equal individuals in essence. Interactive communication is common between managers and employees and usually be seen as an efficient and productive way to work, where cooperation among organizational members are enhanced, and performance is promoted. Managers are more likely to ask for employees’ opinion during the decision-making progress and employees take their right to participate seriously. Hence, low power distance empowers employees to participate efficiently and actively, which is beneficial for performance measure quality.
According to the cultural scores measured by Hofstede (2001), China is high in power distance (80), which stands in sharp contrast to the figures of the Netherlands, low power distance (38). Given the argument above, it’s reasonable to assume that compared with Dutch employees, Chinese employees’ performance measure participation has less utility in enhancing performance measure quality, since the high power distance restrains their ability to state their own opinion, impairs their willingness to participate and undermine their participation’s usefulness .
Apart from the power distance, individualism also reveals how culture impacts performance measure participation’s efficiency in improving performance measure quality. The degree of individualism is reflected by individuals’ notion regarding the relationship between themselves and other social members. (Harrison, 1992). This dimension is used to measure the interdependence among individuals in society, a high score in this dimension suggests that the society’s social framework is loosely knitted and people place themselves and direct families members above others (Hofstede, 2001). While a low score, called collectivism, stands for a tight-knit framework where people keep integrating into specific groups and emphasis groups’ interest. The fundamental unit of society in this situation is the group instead of the individual. (Hui, 1984)
Performance measure participation’s usefulness is likely to be improved by collectivism rather than individualism. Employees in a collectivism-based environment are deeply interdependent with their group members and motivated by group interest. So, for the sake of organization strategy and development, employees driven by this same goal are willing to participate in the performance measure decision-making process and maximize their ability in improving the performance measure quality. Harrison (1992) notes that employees in high collectivism society tend to participate favorably. Hofstede (2001) also states there is a common believe that group decisions are better than individual decision among people in a collectivist society, such conception promotes group discussion since group members have to communicate with others more frequent. To the contrary, a society high in individualism encourages people to achieve their own goals and regards individual decision better. Hence, employees might gain personal benefits at the expense of organization interest when they expand their influence on performance measurement system, which will diminish the performance measure quality. What's more, it would cost-inefficient to arrive a consensus when employees are motivated by their own, diversified interests. Meanwhile, Lau and Caby (2010) demonstrates that unlike people in an interdependent group where everyone’s opinion is seen as equally important, managers are treated as superiors and their views get more attention in an individualist society, which decrease
lessen the performance measure quality at last. Furthermore, similar support concerning the positive impact of participation in a high collectivism society is found in Chow (1995), a research focus on performance participation in Japan and U.S, that Japanese’s emphasis on “consensual” implies high decision-making participation and more long-term decisions.
Hofstede (2001) identifies that China is a high collectivism country with a score of 20 for individualism, while the mark of 80 hints that the Netherlands is an individualism-oriented country. Hence, Dutch employees who prefer to act in their own interests instead of group interests are likely to show more divergence of views during the process of performance measure participation, which lead to the participation more time-consuming and less productivity.
Based on the argument above, power distance and collectivism have opposite moderate effect on the relation between performance measure participation and performance measure quality and lead to adverse expectations; however, it’s unknown that which one has a dominant effect, hence I put forward a null hypothesis here:
H2: Dutch and Chinese firms do not differ in the intensity of association between performance measure participation and performance measure quality
Table 1 Culture sores of China and the Netherlands
Nations Culture Scores Moderate effect on the relationship between PM participation and PM quality
China high PD (80) negative
low individualism(66) positive
The Netherlands low PD (38) positive
3 Methodology
3.1 Data collection and sample size
A survey approach was employed to collect data. Firstly, an online questionnaire was administered to a sample of pairs of managers and employees who mainly came from the Netherlands and China. Then seminar members selected respondents through snowball sampling, which means seminar members started searching potential respondents from their own network and then extended searches with the contacted people’s network. Snowball sampling method is suited for sampling design and inference when populations are hard-to-reach or hidden. (Heckathorn, 2011). As the respondents for this study have to meet certain specific requirements and they are geographically dispersed, so the sample population is indeed a hidden population, relatively small and difficult to access via household survey and the snowball sampling is preferred concerning this precondition.
This survey adopted Likert scale to collect data. As a scaling method, Likert scale is one of the most extensively used types in survey research, mainly because Likert scale has a simple structure and is easy to operate. With regard to the length of Likert scale, Russell and Bobko (1992) pointed out that 5-point Likert scales were only suitable for measuring discrete value and they were too coarse to reflect participants’ intended responses on moderator effects, while Lewis (1993) found 7-point scales resulted in stronger correlations with t-test results. According to Cummins and Gullone (2000), the expansion from 5-point Likert scales to higher-valued Likert scales will not adversely affect reliability and validity. Therefore, data of performance measure participation and performance measure quality were collected by 7-point Likert scale anchored with 1= Totally disagree, and 7 = Totally agree.
At the first step of survey, both respondents from China and the Netherlands are required to meet following requirements: (1) subordinates have to be professionals or members of staff who carry out the work; (2) they must have worked in their current function for at least one year; (3) the superiors have to use performance measures to evaluate their subordinates’ performance. Next, given the relative low English literacy degree among Chinese employees, another 4 Chinese research members and I translated the English questionnaire into Chinese and then we made a back translation see if the Chinese version and the English version actually expressed the same meaning. Furthermore, in order to guarantee the confidentiality and anonymity, all respondents received a cover letter when the survey was distributed, the letter claimed clearly
that their responses would not be accessed by any third parties and also made an explicit explanation regarding the objective of the research.
At last, a total of 106 employee responses and 103 manager responses were received, after deleting the incomplete responses that can’t translate into pairs, I got 99 complete pairs. However some of the remaining responders are neither Dutch nor Chinese, their responses were also excluded. Eventually, there are 80 usable pairs, 49 from the Netherlands and 31 from China. Table 1 gives an overview of the characteristics of the respondents.
3.2 Measurement of variables
The survey employed in this study is a revised version of the one used in Groen et al. (2015). The online questionnaire has 3 version so that participants can switch the language between Dutch, Chinese, and English since they have different linguistic background. In order to ensure the reliability of the result, the notice at the beginning of the questionnaire advised participants to complete the questions with their intuition. When the data was available, Cronbach’s alpha and principal analysis were used to test the validity of the results and effectiveness of the measure items. Survey for employees was different from that of managers; employees were asked Table 2
Characteristics employees managers
All
the
Netherlands China All
the Netherlands China (N=80) (N=49) (N=31) (N=80) (N=49) (N=31) Sex (%) male 47.50 55.10 35.48 66.25 77.55 48.39 female 52.50 44.90 64.52 33.75 22.45 51.61 Education (%) lower 1.25 2.04 0.00 0.00 0.00 0.00 intermediate 12.50 20.41 0.00 10.00 16.33 0.00 higher 38.75 46.94 25.81 22.50 30.61 9.68 scientific 47.50 30.61 74.19 67.50 53.06 90.32 Age mean 33.31 32.94 33.90 39.36 39.37 39.35 (SD) 9.79 10.84 7.99 7.18 6.31 8.55 Department tenure mean 5.13 4.78 5.68 7.18 6.31 8.55 (SD) 5.09 4.88 5.45 5.90 5.09 6.86 PMP mean 3.59 2.85 4.75 (SD) 1.75 1.59 1.31 PMQ mean 5.14 4.96 5.42 (SD) 1.00 1.09 0.785
to rate on items that measured their own perceived influence on performance measure (performance measure participation), while managers used the same rating system to reflect their concerns about performance measure quality.
Performance measure participation (completed by employees). Consistent with Abernethy and Bouwens (2005), employee’s performance measure participation is measured by a five-item instrument. Each item measures employees’ influence on performance measurement system’s design, implementation, data input, ongoing modification, and maintenance respectively and finally the five items reflect employees’ overall participation during the performance measure development process. A Cronbach’s alpha of 0.949 indicated that all those items measured a single construct, and the principal component analysis confirmed that the item scores were explained by one factor.
Performance measure quality (completed by managers). The items used to evaluate performance measure quality are designed based on Moers’s definition of the quality of contractible performance measures. This article provides 3 factors to examine performance measure quality according to the extent to which a performance measurement system is precise, sensitive and verifiable. Managers’ results show how they view the performance measure is influenced by their employees (sensitivity), uncontrolled factors (precision) and measurement process (verifiability), and a result with high sensitivity, precision and verifiability indicates good quality.
Originally, in total 11 items were used to measure performance measure quality and all the items were listed in the appendix. After I had tested the final data with the principle component analysis, 3 factors relating to the performance measure quality were found, which matched the expectation. However, not all the results had sufficient validity and some items measured things that deviated from the definition of performance measure quality. In order to ensure and enhance the relevance of the results, I deleted the same items that Groen, B.A.C deleted in her study. The first 4 items listed in the appendix failed to give a precise description of the external factors that out of employees’ control, so they were unable to provide convincing results and I deleted them. What’s more, the last three items that measured the link between employees’ input and performance were also deleted because theoretically they didn’t fit the definition of performance measure quality. Performance measure quality should reflect employees’ output instead of input. At last, I use the results from the remaining 4 items. Reliability of the result is confirmed by the Cronbach’s alpha of 0.762.
Control variables. In case that manager respondents’ evaluation of performance measure quality varies by personal or industry characteristics, I add “age,” “tenure” and “operation model” as
control variables to explain the potential deviation, data of those variables were collected from managers’ response.
Existing research suggests that management tenure has a complex impact on management. Luo et al. (2014) find that manager tenure is positively related to firm-employee relationship strength. Hambrick and Fukutomi (1991)suggests that during the early period of career, managers who are new to the industry or company will try to make up for their experience gap by expanding their initially low knowledge of management practices, labor relation, and business environment. To accumulate the operational-related knowledge, managers usually turn to their subordinates. Meanwhile, managers who serve their company for a longer period tend to search internal information rather than the external source. Hence, they emphasize more the first-hand information collecting from their employees (Baum et al., 2000). According to Hitt et al. (2001), employees can have more empowerment and enhance self-efficacy when managers care to learn from them. With the delegation, employees can view their own opinion and thus enhance the precision, sensitivity, and verifiability of performance measurement system. Therefore, manager’s tenure are supposed to have a positive effect on the performance measure quality. Normally, a manager who has to serve for the same company for many years are older than the new managers, so the effect of age should also be taken into consideration.
In addition to the potential effect of tenure, manager’s evaluation of performance measure quality might also vary by the companies’ operation model, For instance, Chenhall et al. (2010) find that the expressive role of performance measurement system is more important for the non-profit organization than for-non-profit organizations. Hence for the non-non-profit organization where employees can have more opportunities to express themselves and are more willing to exchange their experience, the precision of performance measure is likely to enhance and so for the performance measure quality.
Table 3 Varimax rotated component matrix
Q.no. Item no. Factor 1 Factor 2
PM Participation(α=.949) 1 PM participation 1 .944 2 PM participation 2 .910 3 PM participation 3 .929 4 PM participation 4 .892 5 PM participation 5 .879 PM Quality(α=.762) 6 PM quality 1 .547 7 PM quality 2 .901 8 PM quality 3 .812 9 PM quality 4 .795
4 Result
4.1 Preliminary Analysis
Before the hypotheses were tested, a preliminary analysis was made to examine the overall characteristic of the research sample, the demographic that portrays the respondent’s general characteristics is provided in section 3. Table 2 also represents the means and standard deviations of two countries’ respondents’ responses of performance measure participation and performance measure quality. According to the descriptive statistics, it’s noticeable that Chinese employees averagely perceived more performance measure participation than Dutch employees. What’s more, Chinese managers generally rated higher on the performance measure quality items than the Dutch managers, but the score gap of performance measure quality between the two national samples is smaller than that of performance measure participation. At the same time, based on the standard deviation, the scores graded by Chinese respondents have a higher concentration when compared with the data gathered from Dutch respondents. These data were put to use to get a formal statistical analysis later when I tested the hypotheses.
Table 3 represents the bivariate relation between variables. Except for the positive association between control variables, the data also indicates a significant correlation between performance measure quality and performance measure participation, which matches my expectation.
Table 4 Correlations (p values between brackets)
PMQ PMP Age Tenure Operation model
PMQ 0.205 -0.077 -0.023 0.036 0.068 0.497 0.843 0.752 PMP 0.143 0.097 0.146 0.067 0.207 0.393 0.196 0.552 Age -0.065 0.091 0.418** 0.15 0.565 0.424 0 0.185 Tenure -0.055 0.081 0.331** 0.084 0.63 0.478 0.003 0.46 Operation mode 0.04 0.068 0.175 0.075 0.727 0.551 0.122 0.507
Pearson correlations appear below the diagonal, non-parametric Spearman correlations appear above the diagonal
4.2 Assumption testing
In order to test whether the responses of performance measure quality are normally distributed, I looked at the histograms and ran a Kolmogorov-Smirnov test simultaneously. The negative values of skewness (-0.775) indicated that the responses tended to center on the right side of the distribution, while the result of kurtosis, which has a positive value of 0.532, represented that the distribution is pointy. For a variable which is normal distributed, its distribution should have a skewness equal to 0 and a kurtosis equal to 3. The z-values of the dependent variable’s skewness and kurtosis are 2.88 and 1.45 respectively, and the dependent variable’s distribution is proved to be significantly skewed by checking the statistic in z-table. What’s more, the results of following Kolmogorov-Smirnov test and Shapiro-Wilk test turn out to be similar, significant scores on both test suggests that there is a deviation from normality in the distribution of the dependent variable.
In case that some extreme cases, which are significantly from the main responses, prevent the data distributed formally, I checked the existence of outliers through frequency table provided by the histogram. According to descriptive statistics, the mean and standard deviation for performance measure quality are 5.14 and 1 respectively. Based on common practices, the case outside the 3 standard deviations should be deleted to ensure the data is concentrated, which means the scores lower than 2.14 and higher than 8.14 are useless. In this case, only the response that has a score lower than 2.14 should be regarded as an extreme case since the score of every item ranges from 1 to 7. Given the output of frequency table, there is no extreme case in the statistic of performance measure quality. Therefore, the normality of dependent variable’s distribution is free from the noise of outliers.
In line with those tests of normality, the significant skewness of distribution violates the assumption that the dependent variable should approximately normally distribute in all groups. Such violation might influence the t value of the result when I test whether the hypotheses are significant.
.
4.3 Hypotheses testing
Hypothesis 1 states that there is a positive relation between performance measure participation and performance measure quality, hypothesis 2 expects the intensity of the relation is the same in Dutch and Chinese firms, a null hypothesis that denies the moderation effect of the country. To test the two hypotheses simultaneously, I performs moderated regressions with
performance measure quality, a 4-item measure of performance measure quality based on the managers’ response, as the dependent variable, the model is:
Yi = b0+b1PMP+b2Country+b3PMP*Country+b4Age+b5Tenure+b6Operationmode+ei.
Where Y= performance measure quality, PMP= performance measure participation
The model includes a dummy variable, country, and a two-way (b3) interaction effect of the independent variable. In order to give the interaction term a realistic interpretation, the original variables are centered on having a zero mean score and the interaction term is calculated with performance measure participation and country’s centered scores.
The adjusted R square in table 4 is 0.007, and F value is 1.097, indicating that the model fit is extremely unsatisfactory and independent variables and the whole model basically explain no variance of the performance measure quality. The reason for the poor model fit might because that other factors that can influence performance measure quality are not included in this model, for example, industry type, firm size, task difficulty, however since the sample size is small, if I classified firms by industry type and firm size, the result might lose its validity.
The t value of performance measure participation’s unstandardized coefficients (0.122) is 1.324, which means there is an insignificant positive relation between performance measure participation and performance measure quality. Hence hypothesis 1 is rejected. The results are opposite to Groen (2012), a possible explanation for the inconsistent result could be the selection of analysis methods. While this study adopts moderated regression model, Groen (2012) uses structural equation model. For the interaction term, the insignificant result can’t reject hypothesis 2, which means that the relations between performance measure participation and performance measure quality in Chinese firms and Dutch are the same. The results of hypotheses testing are unexpected and opposite to most management accounting literature.
Table 5 Results or regression of Performance measure quality
variables
Unstandardized
Coefficients Standardized Coefficients
t Sig. VIF B Error Std. Beta (Constant) 5.292 0.598 8.847 0 PMP (centered) 0.122 0.092 0.212 1.324 0.19 2.041 Country 0.434 0.295 0.211 1.468 0.147 1.645 PMP*Country (centered) -0.162 0.171 -0.159 -0.946 0.348 2.243 Age -0.01 0.014 -0.089 -0.709 0.481 1.267 Tenure -0.008 0.021 -0.045 -0.359 0.721 1.267 Operation mode 0.167 0.325 0.06 0.515 0.608 1.067 Adjusted R Square 0.007 Durbin-Watson 1.919 F value 1.097 P 0.372
5 Conclusion and Discussion
The unsupported expectation can be explained from the perspectives of methodology and theory simultaneously. To begin with the methodology part: The reason why the results of two national samples are inconsistent with that of the whole sample might partly lie in the fact that the national sample size is too small to generate a precise outcome. Dolnicar et al. (2016) suggested that survey data are prone to suffer a range of biases and increasing the sample size can compensate for some biases. Due to the limited resource, the sizes for the Dutch sample and Chinese samples are 49 and 31 respectively. It’s probable that the information collected with the relatively small national samples is insufficient to provide a credible or desired precision of the sample estimates since the magnitude of the sample size is positively related to the uncertainty in the sample estimates. Meanwhile, although the study made some basic requirements for the respondent’s characteristics, there is no specific classification for the industry type that participants work in. Former cross- cultural research that investigated the similar topic of management control practices usually collect the data from a single industry, for example Chow et al (1999) uses data derived from the computer (electronic) firms, while Merchant et al. (2011), Gibbs et al. (2004, 2009) and Jansen et al. (2009) collect data in the retail automobile sector. Respondents’ different industry backgrounds might have a moderate effect on their judgment of the same questionnaire items.
In addition to the small sample size problem, the inferences can also be challenged by several biases that are common in the cross-cultural studies. (1) Method bias caused by the response styles. Usually, response styles that are shaped by various cultural values stand for different characteristics (Davidov et al., 2010). According to Harzing (2006), consistent cross-cultural differences in acquiescence and extremity responding are found across 26 countries. Meanwhile, the acquiescence and extremity responding is prevalent among interviewers who live in countries with high scores on collectivism and power distance. Since half of respondents of this study come from China, a country with significantly high scores on collectivism and power distance, their response might have such bias. (2) Item bias, a bias arises since respondents with dissimilar cultural backgrounds fail to reach consensus on the mean score of the same item. Ryan et al. (1999), a study focus on examining source of differential item functioning in global employee opinion survey, confirms that the item bias is associated with the country characteristics.
In the meantime, as is always in any cross-cultural survey, the data source of this research is under the attack of the potential translation problem (Arce-Ferrer and Ketherer, 2003). For the sake of keeping scale equivalence between the source language and target language, the
translation should ensure items and scales of different versions have similar linguistic and measurement characteristics as much as possible (Hambleton, 1994). However, numerous studies have already documented that linguistic translation cannot guarantee that a test’s different versions can be identical in content, complexity, reliability, and validity (Allalouf et al., 1999; Hulin, 1987). As I mentioned in section 3, before the Chinese version, also the final and third version, was available, the survey was initially designed in Dutch and then translated into English by native speakers. Although translators did their best to keep scale balance between the three versions, linguistic challenges can still exist for the different cultural context.
Then, from the point of view of theory, there are also several factors that might diminish the effectiveness of performance measure participation in enhancing the performance measure quality.
First of all, the way managers view the information asymmetry between themselves and employees. Hypothesis 1 is based on the premise that employees obtain the unique, valuable information that is unachievable for managers, e.g. some professional skills and specialized knowledge that can only be accumulated from front-line staff’ routine work. And such information can improve the former performance measurement system and correct the dysfunctions of traditional performance measurement system when front-line employees are willing to share their knowledge with supervisory managers and peers during the process of performance measurement participation. Given the benefits of employees’ participation, performance measure indicators are expected to be more comprehensive are more sensitive to employees’ output during the daily work. At the same time, information can be freely transferred from employees to managers, which reduce information asymmetry. However if managers do not recognize employees’ information advantage of the performance measures or managers do not believe the credibility of employees’ specialized information, they might deny that employees’ performance measure participation can enhance the performance measure quality. (Gravesteijn, Evers, Wilderom, & Molenveld, 2011) According to Vries et al. (2012), managers who support an autocratic leadership style tend to ignore employees’ voice during the participation process; they simply do not believe that employees have better information than themselves. The reason why they offer an opportunity for employees to express their views lies in the organization’s ideology and requirements. If their organizations enforce participative performance measurement procedures, autocratic managers have to comply with such requirements, even though they will not really care about employees’ views.
Secondly, employees’ abuse of their information advantages, if there is an excessive information asymmetry in the firm and managers rely heavily on their employees to collect essential information, employees might manipulate the information asymmetry to seek personal gain when they take part in the performance management system. Such behaviors might render employees’ performance measure participation useless in enhancing performance measure quality.
Baiman & Evans (1983) suggest that participation is a negotiation process. Participative performance measurement system enables employees to negotiate with their supervisory managers over the design of specific performance measure indicators, budget-setting and compensation schedule. During the participative process, specific negotiation strategies adopted by employees will vary with employees’ information. An employee who is better informed about the working environment and growth potential of production might reveal his private information untruthfully. For example, instead of disclosing all the private information, the employee can develop a cherry-picking behavior when he participates in the performance measurement system. By means of manipulating and misrepresenting information, employees can gain themselves “slacks resource” and pursue additional payment. (Christensen, 1982). For example, during the budget setting process, employees who have the opportunity to express themselves might bargain for the slack budget by overstating the task difficulty and understating the revenue (Schiff & Lewin, 1970). Employees might also advocate an objective that is easy to accomplish so that they can get a desirable job performance without devoting too much effort to the work. If employees participate in this way, their participation can have an adverse impact on the allocation of firm resource and hamper the performance measure indicators to evaluate their job performance precisely.
Thirdly, the lack of trust and open communication between employees and managers. Trust plays a fundamental role in increasing managers’ willingness of involving low-level employees in the decision-making process, while open communication is the core of developing and maintaining trust (Thomas, Zolin & Hartman. 2009, Speritzer & Mirshra, 1999). When there is high-level of interpersonal trust among the group, the group will experience excellent cooperation and higher performance, when the trust is relatively low, the group cooperation and performance tend to be less effective (Dirks, 1999). Managers who afraid that employees’ objectives conflict with the collective interest of organization do not trust employees’ participation, and they will suspect employees’ motivation when employees put forward a
different or opposite opinion. What’s more, managers might give negative feedback or punish employees when employees disagree with them. As for employees, after being punished for the disagreement, they believe that managers only accept views that are consistent with organizational or managerial demands and they will quickly participate in a passive way, instead of seeking interactive communication. The ineffective communication resulting from the distrust between managers and employees prohibits employees from contributing valuable information and suggestion to the performance measurement system, hence employees’ participation adds little value to the performance measure quality.
The fourth factor that might play a role in weakening the effectiveness of performance measure participation is the favoritism in an organization. Managers’ favoritism can be represented in a variety of ways, and it is common to see that managers consciously classify their employees as “in-group” and “out-group” (Robert, 2003). During the whole process of participation, even though employees have the right to express themselves and share private information, it is still managers to make the decision to accept or reject employees’ opinion. If a manager who is responsible for evaluating employees’ participation shows a preference for individual employees, the fairness of the participation is under the attack of the manager’s favoritism (Ponzo & Scoppa, 2010). For example, when managers and employees cooperate in the development of a performance measure indicator, favoritism might eventually result in not listening to the employees who have the best suggestion but those who are closely related to the managers. Meanwhile, employees who are not preferred by managers have already anticipated their managers’ bias, so they will not treat the opportunity of participation seriously and are less likely to give a meaningful suggestion. In the end, a biased development process will lead to a biased performance measure indicator that fails to represent most employees' job performance fully but tends to benefit managers’ preferred subordinates (Berger, Herbertz & Sliwka, 2011)
A fifth impediment would be the insufficient and unqualified feedback of employees’ participation. Clear and timely feedback lies in the center of improving the effectiveness of performance measure participation. During the participation process, managers are expected to use feedback to deliver nonjudgmental and non-defensive information regarding employees’ suggestion and performance (Stephen, 2002). On the one hand, when employees put forward valuable information, qualified feedback is able to provide an effective and ongoing communication between managers and employees (Latham & Wexley, 1981). Employees can have a better understanding that how their specific skills and professional knowledge are utilized
in the performance measurement system via manager’s feedback, which enhances employees’ acceptance of the performance measures. On the other hand, it is noticeable that not all the employees can provide useful information, some of them can make mistakes or provide less informative information. Qualified feedback can also serve as a means of providing essential instruction when employees are walking in the wrong direction. With the corrective guidance, employees can distinguish useful information from unnecessary and redundant information and find a more efficient way to participate (Longenecker, Scazzero & Stansfield, 1994).
Therefore, when there is no timely and indicative feedback for employees, employees are unable to evaluate their participation fairly, and they might give pointless information, which increases the cost of involving employees in the development of performance measures and fail to increase the performance measure quality.
Although findings of this study suggest the intensity of the relation between performance measure participation and performance measure quality is the same for Dutch firms and Chinese firms, which means the mediating role of cultural values is insignificant. However, the result can also be interpreted from another perspective that the positive moderating effect of low individualism is offset by the negative moderating effect of high power distance. Hence, both Chinese and Dutch employees have a neutral reaction toward performance measure participation.
In contrast to the previous empirical studies, this study failed to find a significant positive relationship between performance measure participation and performance measure quality and indicated that cultural values had little effect on this relationship. Through explaining the unexpected results, this study, from a side, confirmed that the relationship between performance measure participation and performance measure quality was under the attack of various impediments and the operation of performance measurement systems in different cultural environment is more comprehensive than the expectation.
6 Directions for future research
As mentioned above, the finding of this study has several scientific limitations and this topic still needs to be deeply studied. Researchers who are interested in this area are advised to enlarge the sample size and concentrate on a single industry. At the same time, people can also do the empirical studies about the impediments that reduce the effectiveness of the performance measure participation and the outcome of those impediments. Meanwhile, further research can compare the specific design and operation of performance measure systems in different countries and industries to find out the influence of cultural values and industry type on those respects.
References
Abernethy, M. A., and Bouwens, J. (2005). Determinants of accounting innovation implementation, Abacus, 41 (3), pp.217-240.
Allalouf, A., Hambleton, R., and Sireci, S. (1999). Identifying the causes of DIF in translated verbal items. Journal of Educational Measurement, 36, 185-198.
Arce-Ferrer, A. and Ketherer, J. (2003). The effect of scale tailoring for cross-cultural application on scale reliability and construct validity. Educational and Psychological Measurement, Vol. 63, pp. 484-501.
Banker, R. D. and R. J. Kauffman. (2004). The evolution of research on information systems: A fiftieth-year survey of the literature in Management Science. Management Science 50 (March), pp.281–298.
Baiman, S., and Evans, J.H. (1983). Pre-Decision Information and Participative Management Control Systems. Journal of Accounting Research, Vol.21 (2), pp.371-395
Burney, L., and Matherly, M. (2007). Examining performance measurement from an integrated perspective. Journal of Information Systems, Fall, Vol.21 (2), pp.49 (20)
Baum J.A.C, Li, X.S., Usher and J.M. (2000). Making the next move: how experiential and vicarious learning shape the locations of chains’ acquisitions. Administrative Science Quarterly 45(4), pp. 766–801.
Berman et al. (2013). The Impact of Societal Culture on the Use of Performance Strategies in East Asia: Evidence from a comparative survey, Public Management Review, 15 (8), pp.1065-1089
Braun, H. W. and Warner. M. (2002). Strategic human resource management in western multinationals in China; The differentiation of practices across different ownership
forms, Personnel Review, 31 (5), pp.553-579.
Cawley, Brian D., Keeping, Lisa M.; Levy, Paul E. M. and Kevin R. (1998). Participation in the Performance Appraisal Process and Employee Reactions: A Meta-Analytic Review of Field Investigations, Journal of Applied Psychology, 83 (4), pp.615-633.
Chalos, P., and S. Haka. (1990). Participative budgeting and managerial performance. Decision Sciences, 20 (Summer), pp. 334-347
Chenhall, R. H., and P. Brownell. (1988). The effect of participative budgeting on job satisfaction performance: Role ambiguity as an intervening variable. Accounting, Organizations and Society 13 (3), pp. 225-33.
Chenhall, R. H., et al. (2014). The expressive role of performance measurement systems: A field study of a mental health development project. Accounting, Organizations and Society.
Chiles, A.M. and Zorn, T.E. (1995). Empowerment in organizations: employees’ perceptions of the influence on empowerment, Journal of Applied Communication Research, 23 (1), pp.1-25.
Christensen, J. (1982). The Determination of Performance Standards and Participation. Journal of Accounting Research, 1 October 1982, Vol.20 (2), pp.589-603
Chong M. L. and Caby, J. (2010). The effects of national culture on the role of participation in different task situations, Advances in Accounting, incorporating Advances in International Accounting, 26 (1), pp.128-133.
Chow, C.W., Duh, R.R., Xiao, J.Z. (2007).Management accounting practices in the People's Republic of China, C. Champman, A. Hopwood, M. Shields (Eds.), Handbook of management accounting research, Vol. 2, Oxford, Elsevier, pp. 923–967
