Remuneration consequences due to national culture
Rijksuniversiteit Groningen Faculty of Economics and Businesses Master of Science in Business Administration Specialization Organizational & Management Control
Author MT van der Velde Student number: 1424432
Thesis supervisors Rijksuniversiteit Groningen Supervisor: C.P.A. Heijes
Preface
The paper in front of you is the result of a research on the influence of national culture on the design of remuneration systems. This thesis is written at PricewaterhouseCoopers Amsterdam, in order to bring the Master ‘Organizational & Management Control to an end. I would like to thank this company for the support, data and contacts that made writing this thesis possible during my internship.
The interest in this topic has its origin in the course ‘Field Course Organizational & Management Control’ at the Rijksuniversiteit Groningen. During this course, much attention is paid to remuneration and national culture as separate topics. Curiosity in connecting these concepts and the actual discussions about the rewarding of directors convinced me to write my thesis about the impact of national culture on the rewarding of the board of directors in major organizations in different countries. Next to PricewaterhouseCoopers, I would like to thank a number of other persons. First of all, I want to thank my parents, who gave me the opportunity to study at the University and for their support and patience while bringing this study to a good end.
Further, I took advantage of the comments of Coen Heijes, who was my supervisor on behalf of the University. In making clear appointments from the beginning to the end, he made it possible for me to make progress in writing my thesis.
At last I would like to thank the friends I met during my time as a student in Groningen. They provided me unforgettable experiences, which developed me to the person who I am today.
I wish you much pleasure in reading this thesis!
Amsterdam, November 2009,
CONTENT PREFACE………2 ABSTRACT……….4 1. INTRODUCTION………5 2. LITERATURE REVIEW……….8 2.1. National culture……….8 2.1.1. Power distance………...9 2.1.2. Individualism………11 2.1.3. Masculinity………...11 2.1.4. Uncertainty avoidance………..12 2.1.5. Long-term orientation………...12 2.2. Remuneration systems………..13
2.3. Relation remuneration systems and national culture………....15
2.3.1. Hypothesis development………...15 3. RESEARCH METHOD………...20 4. RESULTS……….26 4.1. General findings………....…27 4.1.1. Hong Kong………27 4.1.2. The Netherlands………29 4.1.3. South Africa………..30 4.1.4. Sweden………..31 4.1.5. United Kingdom………....32 4.2. Empirical results………...34 4.2.1. Regression analysis………...34
5. CRITICS HOFSTEDE DIMENSIONS………39
6. CONCLUSION AND DISCUSSION ………..42
7. REFERENCES……….48
Abstract
Despite of the continuous debate about remuneration for executive directors, little is known about the relation of these compensations and national culture. This study tries to broaden the current dialogue by researching the influence of national culture on the distribution of remuneration elements in five selected countries scattered over the world.
The cultural dimensions evoked by Geert Hofstede (power distance, uncertainty avoidance, individualism/collectivism, and long-term orientation) are related to two dimensions of remuneration (short-term/long-term, fixed/variable) by applying regression analysis. The most important findings of this research were that 1) organizations operating in countries with a high score on masculinity in general apply more variable compensation; 2) organizations operating in countries with a high score on individualism in general apply more variable compensation. The remaining dimensions of national culture don’t significantly influence the type of director’s remuneration.
It could be concluded that although some significant relations between national culture and remuneration policy were found, a number of other variables also should have a certain impact.
1. Introduction
Hence the ENGLISH, of any people in the universe, have the least of a national character; unless this very singularity may pass for such.
David Hume, essay XXI, 1742.
Last year’s news was full of negative publications about excessive remunerations for executives. The bigger part of the population abominates high payments to directors for loosing organizations money. High bonuses rouse directors to take exceptionally risks in trying to achieve certain performance criteria. Some state that these bonuses are the main cause of the current credit crunch. Correctly or not, the fact is that since the start of this credit crunch in the summer of 2007 a lot of attention is paid to the rewards directors received. Due to globalization, this debate is stretched out all over the world.
Today’s bankers face great resistance of shareholders who see their money condense. In reaction governments of different countries are trying to reduce the payouts to directors and emphasize the importance of corporate governance regulations. In a number of countries different ‘codes of conduct’ are developed in order to define in which way organizations should act to be efficient, transparent and responsible. ‘Code Tabaksblat’ (the Netherlands), Sarbanes-Oxley (United States of America), Cadbury Code (United Kingdom) are examples of corporate governance codes in use. These regulations demand amongst other things for disclosure of remuneration of directors in order to inform stakeholders. Applying International Financial Reporting Standards (IFRS) is even obligatory for all stock market listed organizations in the European Union1.
The commotion about the high payments for the higher management of organizations doesn’t leave the board of director’s cold. Former Dutch minister of finance and current CEO of ABN Amro Gerrit Zalm announced the higher management of the bank to turn in their remuneration. Consequently, refuse of this announcement could lead to termination of employment (Financieel Dagblad). ABN Amro is not the only company that reacts to the increasing pressure of stakeholders for reducing or even releasing remunerations. Bank-insurance company ING is another example (Financieel Dagblad). The number supporters for adjusting the remuneration policies are rapidly growing. In the United Kingdom CEO’s and former politicians argued last February for revising the system of bonuses. Rewards were also a big issue during one of the latest G-20 summits. These instances show the intensity of this topic today.
Globalization started in the late 19th century2. An increasing amount of organizations is moving all over the world in trying to obtain a competitive position in the worldwide market. These international
1
http://www.ifrs.com 2
activities have converted numerous organizations into multinational corporations acting in several countries. Employees from different countries are admitted to the business. In that way, it is of great importance for managers to know how to deal with upcoming cultural differences and whether cultures of countries have an impact on the generally accepted methods of managing human resources. One comment should be added on the still increasing globalization. Pauly and Reich (1997) determined that cultures of smaller countries tend to be more influenced by globalizations than bigger countries, because of their weaker cultures.
Taking into account the actuality and the increasing globalization, it is interesting to look at the subject of remuneration and international differences between organizations.
In this thesis I will research the influence of cultural differences on the fulfilment of remuneration systems in organizations. The purpose of this study is to broaden the understanding of the relation between national culture and the design of remuneration systems in organizations. This relation will be researched by using theory.
The use of compensation practices that are geared towards the cultural expectations of the host countries’ employees of multinational firms may help firms in creating competitive advantages (Schuler and Rogovsky, 1998). One major part of managing organizations and HR management is the remuneration policy. By providing attractive remuneration packages, organizations are able to appoint the best employees in order to trump the competition. Especially in this fast moving world, where people are all working over the world it is important to stay competitive by attracting the best personnel.
Is it useful to apply performance rewards, and are these bonus systems in different countries designed in the same way? The actual subject of bonuses and the increasing globalization resulted in the following research question:
To which extent does national culture have influence on the design of remuneration systems in organizations?
In order to answer this question, the following sub questions are developed: - What is national culture?
- What is a remuneration system?
National culture is defined following the dimensions evolved by Geert Hofstede (1980); - Power distance (PDI)
- Individualism (IND) - Masculinity (MAS)
- Uncertainty avoidance (UAI) - Long-term orientation (LTO)
These dimensions demonstrate cultural differences between countries by applying scores. In the literature review section, the dimensions will be further explained. Following Chiang (2005), a remuneration system consists of
• financial / non-financial elements • short term / long-term elements • variable / fixed elements
The emphasis in this research is on the last two elements of the remuneration system.
2. Literature review
In this section the theoretical basis for this study will be discussed. Understanding this knowledge is necessary for answering the research question formulated in the introduction section. First of all, the five dimensions of national culture developed by Geert Hofstede will be considered. Following these dimensions, elements of the remuneration systems will be expounded. In doing this, the two variables for this research are explained. After these parts, previous research about the combination of both variables will be analysed.
Current theories and research about remuneration for directors in organizations are mainly focused on the United States (Tosi and Greckhamer, 2004), while the debate on excessive remuneration for the higher management is a global one. Some authors have paid attention to cultural differences and remuneration systems, but the extent in which these variables are related is still not clear. Therefore, in the rapid changing world of globalization, this study could be useful to broaden the dialogue in trying to understand the role and influence of national culture on the design of remuneration systems. This could raise possibilities for management in multinational organizations to adjust their remuneration system to the cultural values of the country in which it operates. The human resource function greatly affects a firm’s organizational success and therefore can be used to gain competitive advantages (Schuler and MacMillan, 1984).
In the following subsections, the components of the research question, and answers to the sub questions evoked in the introduction section will be explored.
2.1 National culture
National culture in this study will be defined as ‘the values, beliefs, and assumptions learned in early childhood that differentiates one group of people from another (Beck and Moore, 1985; Hofstede, 1991). One of the most accepted and comprehensive cultural typologies are those proposed by Hofstede (1980). On the books in humanities list, published by Times Higher Education Hofstede is ranked even higher than well known authors like Karl Marx and Friedrich Nietzsche3.
In this model, national culture is delineated into a set of measurable constructs. From 1967 through 1973, Hofstede did research at subsidiaries of IBM in 66 countries by interviewing 117.000 employees. Despite of the development of new cultural dimensions, which will be explained in the
final part of this research, this almost thirty years old research is still the base for a lot of studies related to national culture.
After executing a statistical analysis of the results obtained from the IBM research, the following was concluded; in every country, there are similar problems, but the solutions for these problems differ from each other in relation to the subsequent matters:
• social disparity
• relation between individual and groups
• desired distribution of roles between men and women • ways to overcome uncertainty
On account of these results, Hofstede came up with four dimensions for national culture (Hofstede, 1980). After another study of Hofstede, in cooperation with Bond (1988), a fifth dimension is added. The next subdivision will describe these five dimensions of national culture.
Despite the popularity, there is a growing critique on the methodology used by Hofstede (McSweeney, 2002). New dimensions are developed to overcome these shortcomings (Hampden-Turner and Trompenaars, 1993). In a later section in this thesis, attention will be paid to the critics on the Hofstede dimensions. Nevertheless, today Hofstede is still the most cited non-American in the field of management in the US Social Science Citation Index (Emerald). Because of the widely applied model of national culture, the dimensions build up by Hofstede (1980) will be the foundation for this thesis.
2.1.1. Power distance
‘All societies are unequal but some are more unequal than others’ (George Orwell)
The power distance index (hereafter PDI) provides information about dependency in relations. The foundation for this dimension is human inequality, which is mainly formalized in boss-subordinate relationships. PDI is a measure of the interpersonal power or influence between boss and subordinate as perceived by the least powerful, the subordinate (Mulder, 1976). One could expect that an employee wants to reduce the power distance between himself and his subordinate. On the other hand, the boss wants to enlarge this distance. Despite of influences from both sides in the relation, the study of Hofstede suggests that the level of power distance depends on the society.
The definition used for the power distance dimension in this thesis comes from Hofstede (2005): the extent to which the less powerful members of organizations and institutions (like family) accept and expect that power is distributed unequally.
The scores for PDI, together with the scores on the other dimensions are given in table 1 of the appendix. The computation of the PDI scores for the different countries used mean scores on a five-point scale (1 = very frequently, 5 = very seldom) for question a, which is presented below; these mean scores have been multiplied by 25 to make their range, and therefore their contribution to the PDI, roughly equal to the range in percentage values of question b and c, demonstrated below. The questions are asked to subordinates, who can be supposed to be better judges, because it could occur that their bosses will conceal the actual hierarchy.
(a) Nonmanagerial employees’ perception that employees are afraid to disagree with their managers (B46). (b) Subordinates’ perception that their boss tends to take decisions in an autocratic (1) or
persuasive/paternalistic (2) way (A55).
(c) Subordinates’ preference for anything but a consultative (3) style of decision-making in their boss: that is, for an autocratic (1), a persuasive/paternalistic (2), or a democratic (4) style (A54).
This could be translated into formula (1):
PDI = 135 – 25 (mean score employee afraid)
(% perceived mgr. 1 + 2) + (1)
(% preferred mgr. 3) -
The number of 135 is used as a correction in this formula in order to give the PDI a range between 0 and 100.
The answers on these questions are translated into relative scores for different countries.
2.1.2. Individualism
An individualistic society is defined as a society in which the ties between individuals are loose: everyone is expected to look after him/herself and his/her immediate family. It deals with the way people live together. This is related to families and organizations as well.
A collectivistic society is defined as a society in which people from birth onwards are integrated into strong, cohesive in-groups, often extended families that continue protecting them in exchange for unquestioning loyalty.
The scores for this dimension are based on a factor analysis on the answers given on the questions presented in the appendix.
These factor scores were multiplied by twenty-five and increased by fifty. This makes it easier to create a scale varying form zero for collectivistic countries and hundred for individualist countries. An interesting point here is the positive relationship between prosperity and the value on individualism.
2.1.3. Masculinity
2.1.4. Uncertainty avoidance
Uncertainty is a key element in each organization, because all organizations today are acting in an increasingly uncertain environment characterized by a growing number of competitors from all over the world. Hofstede defined this dimension as ‘the extent in which a culture programs its members to feel either uncomfortable or comfortable in unstructured situations’. The three indicators involved are rule orientation, employment stability and stress. The questions used by Hofstede in order to develop scores for this dimension are as follows:
(a) Rule orientation: Agreement with the statement ‘Company rules should not be broken – even when the employee thinks it is in the company’s best interest’ (B60).
(b) Employment stability: Employees’ statement that they intend to continue with the company (1) for two years at the most, or (2) from two to five years; this, of course, taken with a negative sign (A43). (c) Stress, as expressed in the mean answer to the question ‘How often do you feel nervous or tense at
work?’ (A37).
Mean scores on a five-point scale are used for questions a, c, and mean percentage scores are used for question b. In order to create a range of scores from about zero to hundred, the scores of question a, are multiplied by 30, and the scores of question c is multiplied by 40. These steps created formula (2)
UAI= 300
30 x mean score rule orientation (a) -
% intending to stay less than 5 years (b) - (2)
40 x mean stress score (c) -
The highest score on this dimension is 112, while the lowest is 8. High scores on this dimension are found in Latin and East European countries. Lower scores are typical for North European and Asian countries.
2.1.5. Long-term orientation
Long-term orientation means striving for remuneration in the future by persistence and thrift. High focus on long-term orientation is related to a high score. The opposite, short-term orientation, stands for striving towards behaviour focused on the past and the future, respect for traditions, preventing sight loss and meet with social responsibilities.
2.2. Remuneration systems
As mentioned before, compensation practices that are geared towards the cultural expectations of the host countries’ employees of organizations, may help these firms create competitive advantages (Schuler and Rogovsky, 1998).
Swagerman (2007) mentioned four functions of remuneration for directors in organizations; - to attract
- to retain - to motivate - to reward
Despite the fact that each organization uses different remuneration systems, it is possible to divide these systems in certain components in order to satisfy one of more of the above described functions. For this thesis, adjusted remuneration elements of the remuneration components mentioned by Murphy (1999) will be used. According to Murphy, there are six forms of remuneration, which could be divided by a number of sub elements:
1) Base salaries 2) Annual bonus plans 3) Stock options 4) Pension
5) Long-term cash payments 6) Other
Ad 1)
This remuneration element is a so called ‘benchmarking’ component and therefore based on - in general - competitive companies (Murphy, 1999). Consequently, one could expect that the height – corrected for firm size (Baker and Hall, 2004) - of this remuneration element is almost the same in organizations in the same type of industry. Especially risk averse directors will prefer an increase in base salary to an increase in variable compensation.
Ad 3)
This compensation gives an employee the right to buy a share of the company at a pre-specified strike price at a pre-specified moment. This accounts for a European option. American options could also be exercised before the expiration date (Grimblat and Titman, 2002). As with annual bonuses, motivation of directors is the main driver for stock options. For this research, this component contains the following elements;
- shares - stock options - obligations
Ad 4)
Executive directors routinely participate in supplemental executive retirement plans. These plans can take different forms. Examples of these forms are payments based on years of service, variable benefits based on inflation or benefits determined by company performance. Because it is hard to measure and find amounts of this compensation element, only companies that fully disclose this element are selected for this research.
Ad 5)
Long-term cash payments are payments based on achievements on the long-term (>1year). These payments could be related to share based payments, pension plans and long-term variable bonus plans.
Ad 6)
Other benefits are benefits mainly related to travel allowances and healthcare insurances.
2.3. Relation remuneration systems and national culture
The variables national culture and remuneration systems being described and having a deeper understanding of these variables used for this research, we will have a closer look on literature which relates these variables to each other.
At the moment of writing this thesis, there are a few studies that directly relate national culture and remuneration (Tosi and Greckhamer, 2004). The study of Pennings (1993) found evidence that national culture contributes to differences in remuneration systems in the United States, France and the Netherlands. In this study, data was collected during 1985-1988 with executives in a set of firms from each of the three countries. Most of the selected firms belong to Fortune 500 firms and were operating in a variety of industries. The study shows that only the US set used long-term executive compensation plans. Further, the payout on those plans in Dutch and French companies was less than the bonus payments at US firms which often exceed the salary. Most French companies only paid out a fixed compensation. Although the results are limited in their generalizebility, the results clearly indicate that remuneration systems are more explicit and prominently demonstrated in US companies. The study of Penning presented a first attempt to explain differences in executive reward systems as at least partially a function of cross-cultural differences.
Another previously executed research is more in line with this thesis in the way it relates national culture to the remuneration system. Schuler and Rogovsky (1998) use the in meantime described dimensions of Hofstede in order to draw conclusions about the way national culture influences directors rewarding. The relations between the cultural dimensions suggested by Hofstede and specific compensation practices were examined quantitatively across twenty-four countries and three large data sets. The results of this study stated that national culture provides a fundamental explanation for the variation in the utilization of different reward practices in different countries.
2.3.1. Hypothesises development
supervisors. Gomez-Mejia and Welbourne (1991) stated that individuals in countries that score high on the UA index are characterized by: fear of random events and the unknown; high value placed on stability and routine; low tolerance for ambiguity; low risk propensity; and comfort in security; lack of tension, and lack of contradictions. This results in managers who are trying to avoid making risky decisions. Therefore, fixed reward packages should play a more central role than any type of variable remuneration.
Following this orientation, it is hypothesized that:
Hypothesis 1: Fixed compensation is more likely to be found in countries with higher
levels of Uncertainty Avoidance.
Hierarchical structures in the workplace indicate that employees accept reward differentials/inequalities based on ‘predetermined’ non-performance criteria (Lincoln and Kalleberg, 1990). This respect for authority and loyalty in high PDI cultures could be translated into a willingness to accept non-performance-based reward systems (Child and Markoczy, 1993). Hofstede mentioned that in countries that score high on PDI, managers are not only entitled to privilege, but also should be independent. This suggests that powerful managers can reduce the dependence of their remuneration on the firm’s achievements (Werner and Tosi, 1995). So when superiors have much power over their subordinates, they are less dependent on firm performance. Therefore these superiors don’t have to be motivated by highly variable payments. Consequently one could expect that fixed compensation is preferred by individuals in high PD countries.
An individual-based performance reward system that is believed to narrow the reward gap/disparity between a superior and subordinate is unlikely to be valued in a high PD culture. Since in high PD countries, rewards are linked more to employee position and seniority than contribution and performance it is proposed that (Chiang, 2005):
Hypothesis 2: Fixed compensation is more preferred by individuals in high Power
Distance than by individuals in low Power Distance countries.
Characteristics of employees in individualistic countries are; personal accomplishment, selfishness, independence, individual attribution, internal locus of control, and control over their own destiny. To promote this individual accomplishment, firms should make use of extrinsic rewards to recognize employee success comparing to colleagues (Gomez-Mejiia and Walbournce, 1991). The acceptance of large differences in reward between individuals makes this cultural orientation highly compatible with individual-based performance schemes (Beer and Katz, 1998).
In contrast to individualistic countries, collectivists should prefer reward systems that are non-competitive in nature and that support group harmony and cohesion (Cable and Judge, 1994). These collectivists favour high levels of interaction have a high degree on reliance on others and are more willing to cooperate (Bretz et al, 1989). Cable and Judge further suggest that individuals with more confidence in their ability may perceive greater expectancy in their actions, and consequently be more attracted to pay systems that link to their individual behaviour to rewards, which could be represented by variable payment over fixed compensation. The arguments that are mentioned above lead to the following hypothesis:
Hypothesis 3: Variable compensation is positively related to Individualism.
In societies with high Femininity, quality of life and caring for other people are reflected in compensation practises that take into consideration employees’ personal and social needs, and their lives outside work (Kluckholn and Strodtbeck, 1961; Jackson and Schuler, 1995). Therefore there is less emphasis on achievement, growth and challenge in jobs (Hofstede, 1980). People in these societies are less assertive and show more concern for the quality of the working environment instead for job performance. Working conditions, job satisfaction, and employee participations being
emphasized.
Masculine societies emphasize assertiveness, achievement and material success.
Gomez-Mejia and Welbourne (1991) stated that interest in the financial arrangement, acquiring money and material gain, should be favoured by individuals in masculine countries. Those exhibiting more masculine traits tend to be rewarded with promotions and other favourable personnel decisions. Empirical research of Newman and Nollen (1996), found that a high level of masculinity is compatible with a system that links monetary reward to performance. They say that in the workplace,
masculinity is reflected in merit-based opportunities for high earnings, recognition,
advancement, and rewards. Based on the above described arguments, the following hypothesis can be developed:
Hypothesis 4: Fixed compensation will be more likely to be found in countries with low
levels of Masculinity.
higher the value of the share-based compensation. In other words, one could relate a low score on the PDI with share-based payments (employee participation).
Denison and Mishra (1995) suggest that a high level of involvement and participation create a sense of ownership and responsibility. This high level of involvement and participation could be translated into a smaller power distance in an organization. Because in this new situation the responsibility of each employee is increased, there is more focus on performance. Since the value of a stock option is related to performance of an organization, one could expect power distance to be related to share-based payments. This relation is the basis for the following hypothesis:
Hypothesis 5: Share options ownership will be more likely in countries with low levels of
Power Distance.
Because the long-term orientation dimension of national culture is a relatively new one, few previous studies have taken into account this fifth dimension. Therefore the sixth hypothesis is based on the work of Hofstede. One part of the definition of the long-term orientation dimension stated that this dimension stands for working towards remuneration in the future. This is related to one of the characteristics of countries that score high on this index; that is focus on gradual results. Hofstede argued that little effort in order to achieve results on the long-term is a feature of countries with a focus on the short-term. He also suggests that in cultures with a short-term focus, the main focus should be on results from the last months, or year; managers are judged on these results. This results into the following hypothesis.
Hypothesis 6: Long-term compensation will be more likely to be found in countries with a
high level of Long-term orientation.
Table 2
Overview expected relations national culture - remuneration
Hypothesis Cultural dimension Score Remuneration Movement
1 Uncertainty Avoidance ↑ Variable ↓
2 Power Distance ↑ Variable ↓
3 Individualism ↑ Variable ↑
4 Masculinity ↑ Variable ↑
5 Power distance ↑ Stock options ↓
6 Long-term Orientation ↑ Short-term ↓
The relations between the different dimensions of national culture and the expected influence on the remuneration elements are figurative described in figure 1 below.
Figure 1
3. Research Method
In this section, the research method used for answering the research question and different choices made during this study will be explained.
The form of research applied in this thesis is an empirical one. Quantitative research is used in order to gather statistical input and output.
First of all, as described in the literature review, this thesis started with a literature study in which literature is researched in order to collect information relevant for this thesis. Concepts used in internet search engines were: remuneration, culture, national culture, reward, Hofstede, compensation and incentives. After gathering sufficient information about the concepts, the multifarious literature is analysed. The literature is classified in one of the following groups; 1) remuneration, 2) national culture, 3) remuneration relating to national culture, 4) critics on Hofstede dimensions. This classification created an overview of the relevant literature that could be used for the different sections of this study. After this arrangement, hypothesises from the information obtained by the literature were evolved, which is demonstrated in the previous section.
The next step in the process of collecting information was bringing data of organizations with relation to remuneration together. These data were obtained from a database4 accessible at PricewaterhouseCoopers. This database provides financial reports of organizations in ninety-four countries all over the world. Five countries from this database were selected, in order to research remuneration systems in organizations. The countries were selected making allowances for the following criteria;
• quality of information available • annual report 2007
• geographic location
• scores on Hofstede’s dimensions • size organization
These criteria are important in that this study tries to test in which content and in what way national culture influences remuneration systems. In order to research this, there have to be clear cultural distinctions between selected countries. That makes it simpler to perform proper regression analyses.
The first criterion mentioned is essential in that there should be significant data for proper research. In this case, data is appropriate when organizations disclose the amounts of remuneration on
4
the different components mentioned in the introduction section5. Companies which did not clearly disclose information about the basis salary, annual bonus, retirement compensation and stock plans are dropped. Especially the last two components are hard to measure. Therefore only the organizations whose disclose these components clearly are selected for research. Further, this research focuses on remuneration of executive directors from the board of directors. Director’s remuneration is chosen instead of remuneration for lower management/subordinates, because information about the rewarding of the lower management/subordinates is harder to trace.
Remuneration of board of directors in the selected organizations should be in relation to fiscal year 2007. This criterion tries to except influences of the credit crunch. One could expect that because of this crisis, remuneration systems are redesigned. Due to stricter corporate governance rules and regulations, and pressure for lowering bonuses, smaller differences between organizations and countries could be expected. Consequently, this could make the findings of this study weaker. The reason for leaving out the years before 2007 is topicality. Using data from the most recent year makes this study more actual and therefore more interesting in order to draw conclusions.
The third criterion is applied to make differences between countries as large as possible. Researchers have suggested that entering countries that are psychically close reduces the level of uncertainty firm’s face in the new market (Johanson and Vahlne, 1992); and that psychically close countries are easier for companies to learn about, because one can expect smaller differences (Kogut and Singh, 1998). By selecting countries which are physically far away from each other one can expect the biggest differences.
The criterion related to the dimensions of national culture is important in the way it influences the statistical output. Using countries with almost similar scores on the dimensions of national culture will show no proper relations regarding remuneration systems. Therefore this study selects countries which score quite different on the five dimensions of national culture. This would make the relations more clear.
The last criterion is applied in order to improve the probability of proper results during the reconnoitring expedition. Large organizations generally disclose more information regarding remuneration of directors and therefore it is expected that more data will be available for this research.
Taken into account the criteria mentioned above, the following countries are selected from the database:
I. Hong Kong II. the Netherlands
5
III. South Africa IV. Sweden
V. United Kingdom
The scores on the dimensions of national culture of the selected countries and their ranks compared to all countries studied by Hofstede are provided in table 3 below. The determination of the scores is already explained in the literature review. In general there is a significant distribution of the rankings of the different countries. Only on UAI, all selected countries score relatively low.
Table 3
Scores on cultural dimensions Hofstede
Country Power distance Individualism Masculinity Uncertainty avoidance
Long-term orientation
Score Rank Score Rank Score Rank Score Rank Score Rank
Hong Kong 68 23/26 25 46 57 21/23 29 61/62 96 2 Netherlands 38 52 80 4 14 63 53 44 44 10 South Africa 49 48 65 15 63 16/17 49 51/52 - - Sweden 31 59/60 71 9/10 5 68 29 61/62 33 11 United Kingdom 35 54/56 89 3 66 10/12 35 59/60 25 16/20
Source: Clearly Cultural
As already pointed out, the organizations in the above portrayed countries are checked on completeness of necessary information regarding the remuneration elements.
Table 4
Distribution organizations over countries
Country Hong Kong Netherlands South Africa Sweden United Kingdom
Number of
Organizations 10 13 11 14 10
Data for Hong Kong organizations were most difficult to obtain. However, Hong Kong is selected for research because this thesis wants to include an Asian country in order to create as much as possible geographic distribution of countries. Cheung (1999) indicated that Hong Kong maintains significantly higher standards of corporate governance and at the same time has developed more sophisticated and adequate legal systems to protect property rights than the rest of the countries in the Asian region (Nam et al. 1999). Taken this into account one can expect Hong Kong to be the best Asian country to select in order to obtain data regarding remuneration of executive director’s. Although Hong Kong provides the most appropriate data of all Asian countries, Cheung (1999) recommend for more disclosure of the boards remuneration. This recommendation explained the difficulty of obtaining proper information regarding rewarding in Hong Kong organizations.
Once the assortment is defined, remuneration systems of the selected organizations have to be analysed. In order to evaluate and classify the reward elements, this study makes a distinction between:
• short-term vs. long-term compensation • variable vs. fixed compensation
Figure 2 below, expresses the distinction of remuneration elements used in this research.
Figure 2
Remuneration domains
In this research, short-term compensation consists of the rewards paid over one year or less. This automatically defines long-term compensation as rewards paid over a period of one or more than one year.
Variable compensation contains all elements of the remuneration system that are determined by meeting performance goals. Fixed payments on the other hand are those rewards paid without relation to certain performances and therefore will be paid out anytime.
Table 5 gives an overview of the above described classification of remuneration elements.
One important comment on this table is regarding to stocks. This remuneration element is used in different ways in the organizations that are analysed. In some organizations the amount of stocks is based on company performance in the previous year, while in other organizations this payment is based on the longer term.
Table 5
Remuneration elements
Remuneration elements Variable Fixed
Short-term § Annual bonus
§ Stock
§ Base salary § Benefits
According to Taft and Singh (2003), the used remuneration elements are determined as follows:
Annual bonus: Annual bonus target, determined by performance measures, performance standards and pay performance structures.
Base salary: Base salaries are typically determined through competitive benchmarking, based primarily on industry salary surveys and supplemented with data from peers within the industry.
Stock: All remuneration that is referred to financial instruments. Shares, options and bonds are part of this compensation. One remark that has to be made according to this type of reward is that stocks belong to two categories of remuneration. In some companies, directors are paid in stocks over a one year period (short-term), while in other organizations directors are paid in shares over a 3-5 year period (long-term).
Benefits: Elements of benefits are representation allowances, contribution to health insurance plans, school fees, car allowances, transport allowances and tax compensation.
Pension: Retirement benefits are commonly set using a certain fixed formula related to the base salary.
Table 6, presented in the appendix demonstrates the selected organizations for each of the chosen countries.
After investigating director’s rewarding in all selected organizations, the data remained is tested with statistics program ‘Statistic Package for Social Sciences 14’ (SPSS 14). By performing a regression analysis, the relation between the national cultures for the chosen countries and the remuneration elements used by the organizations in these countries could be described.
4. Results
In this section, the results of the data study in this research will be discussed. First the general findings will be presented. Hereafter the statistical results for testing hypothesises will be discussed.
Before presenting these results, it is important to explain how the different remuneration components are computed. These components’ presenting the dependent variable in this research and therefore models this study.
For all remuneration elements applied in this thesis, ratios are used in order to express the proportion between those elements. Two formulas are used to obtain those ratios. The percentage of short-term compensation is calculated by applying formula three which is presented below. One comment has to be made in relation to this formula. As explained in the previous section, the remuneration element stocks could also be a short-term compensation. If this was the case for companies, the formula is extended with the value of stocks paid out. The percentage long-term compensation is therefore automatically 100% minus the percentage short term compensation.
Formula four is used for calculating the percentage of the variable compensation. All variable payments are summed up and at the end divided by the total remuneration for fiscal year 2007. In this case the percentage of fixed compensation is 100% minus the percentage of variable compensation.
Percentage short term compensation =
((base salary + annual bonus + benefits) / total remuneration)) * 100% (3)
Percentage variable compensation =
((annual bonus + stock) / total remuneration)) * 100% (4)
4.1. General findings
4.1.1. Hong Kong
Hong Kong is the country from the sample that scores the highest on the PDI. Therefore, following the developed hypothesis, the expected result is that organizations from Hong Kong use a high ratio of fixed compensation for directors. This expectation is based on the arguments presented to develop hypothesis two. As already mentioned in the previous part, information about remuneration paid to executive directors in organizations from this country was most difficult to obtain compared to the other selected countries. Remuneration disclosure is a common problem in Asia (Cheung, 1999), which is already referred to in the research method section. The author states that this lack of transparency and disclosure is a danger that acts as a feature of many Asian companies. Despite of this lack of transparency and disclosure, it was possible to select ten Hong Kong companies that met the criteria for selection.
Table 7 shows the distribution of remuneration elements in the selected Hong Kong organizations.
Table 7
Major part of the amounts presented is in $HK. However, a number of organizations present their results in a foreign currency. This is not an issue for this study since ratios instead of absolute amounts are used in order to perform statistical analyses. As a consequence the research is not biased by using different currencies.
Table 8 shows the ratio’s which ultimately function as the input for the various statistical analyses applied.
Table 8
Distribution remuneration Hong Kong
Organization Short-term ratio Long-term ration Variable ratio Fixed ratio
1 62,36% 37,64% 72,20% 27,80% 2 60,24% 39,76% 56,36% 43,64% 3 76,21% 23,79% 60,18% 39,82% 4 87,14% 12,86% 41,50% 58,50% 5 98,81% 1,19% 51,97% 48,03% 6 84,96% 15,04% 39,00% 61,00% 7 67,06% 33,94% 71,56% 28,44% 8 100% 0% 0% 100% 9 86,98% 13,02% 23,74% 76,26% 10 99,82% 0,18% 0% 100% Average 82,36% 17,64% 41,65% 58,35%
Based on this table, the first notion tells us that there is an obvious emphasis on short-term remuneration. The average of these ten companies gives a percentage of 82,36% short-term remuneration. Looking at hypothesis 6 - which expects a focus on long-term remuneration in countries with a high score on LTO - this is the opposite of what could be expected. In the statistical part and in the conclusion these results will be discussed more in depth.
4.1.2. The Netherlands
The Netherlands scores high on IND and low on MAS compared to the other selected countries. Following the developed hypothesises, this high score on IND should lead to a focus on variable remuneration. On the other side, the low score on MAS is matched to fixed remuneration. This is an interesting contradiction, which will try to explain in the next section. Table 9 shows the distribution of remuneration over the elements in Dutch organizations.
Table 9
Formula three and four are used for computing the percentages of the different compensation elements, which are demonstrated in table 10 below.
Table 10
Distribution remuneration Netherlands
Organization Short term Long-term Variable Fixed
1 45,56% 54,44% 81,04% 18,96% 2 90,3% 9,7% 53,66% 46,34% 3 78,02% 21,98% 66,37% 33,63% 4 60,2% 39,8% 56,3% 43,7% 5 64,46% 35,54% 61,49% 38,51% 6 53,39% 46,61% 63,55% 36,45% 7 67,34% 32,66% 47,24% 52,76% 8 74,24% 25,76% 51,07% 48,93% 9 78,98% 21,02% 42,18% 57,82% 10 62,99% 37,01% 60,96% 39,04% 11 87,35% 12,65% 21,04% 78,96% 12 56,58% 43,42% 49,82% 50,18% 13 84,07% 56,58% 11,25% 88,75% Average 69,5% 30,5% 51,23% 48,77%
4.1.3. South Africa
South Africa finds itself in de middle of all scores on the cultural dimensions. Therefore, it is supposed that the design of remuneration systems in South African countries is not exceptional. One could expect that the scores on different elements of compensation lie in between scores of organizations in the other countries. Table 11 in the appendix shows the distribution of absolute remuneration amounts in South Africa.
Table 11
Looking at table 12 below, on the first sight the difference between variable and fixed compensation is not significant. This is expected looking at the ranks on the dimensions for South Africa. Since no score was computed for the long-term orientation dimension of national culture, no hypothesises were developed regarding the term of remuneration for this country.
Table 12
Distribution remuneration South Africa
Organization Short-term ratio Long-term ratio Variable ratio Fixed ratio
1 81,69% 18,31% 55,18% 44,82% 2 80,6% 19,4% 53,06% 46,94% 3 53,29% 46,71% 73,54% 26,46% 4 71,35% 28,65% 58,49% 41,51% 5 85,94% 14,06% 33,54% 66,46% 6 76,23% 23,77% 25,59% 74,41% 7 74,05% 25,95% 71,13% 28,87% 8 80,64% 19,36% 40,71% 59,29% 9 96,72% 3,28% 49,02% 50,98% 10 34,94% 65,06% 62,67% 37,33% 11 76,76% 23,24% 45,54% 54,46% Average 73,88% 26,12% 51,68% 48,32%
4.1.4. Sweden
The companies and the distribution to elements of the remuneration system for Sweden are given in table 13 in the appendix.
Table 13
A remarkable finding when looking at relative distribution in table 14 below is the high score on fixed compensation. The average percentage of fixed remuneration of the analysed Swedish organizations was 72,65%. This is in line with the hypothesis that countries which score low on the dimensions IND and MAS, using more fixed remuneration, since Sweden scores low on these dimensions. The average score on short term remuneration was 72,80%, which is quite high for a country that has a medium score on LTO.
Table 14
Distribution remuneration Sweden
Organization Short term Long-term Variable Fixed
4.1.5. United Kingdom
The distribution to the different elements of the remuneration system for British organizations is illustrated in table 15 of the appendix.
Table 15
This country scores low on PDI and LTO, while it ranked high on the dimensions IND and MAS. Based on the literature review discussed before, one could expect British remuneration systems to be highly focused on short-term –and variable compensation. Table 16 gives an overview of the relative distribution of the remuneration in the United Kingdom.
Taking a view on table 16 below, the average percentage short-term compensation paid to directors is 54,66%. This is the bigger part, but not an outstanding amount. Looking at the average of variable payments, one could find an average of 63,29%, which is the highest percentage variable compensation found in this research, and therefore in line with hypothesis 3.
Table 16
Distribution remuneration United Kingdom
Organization Short-term ratio Long-term ratio Variable ratio Fixed ratio
In the section above, a complete description of remuneration paid to directors in the selected organizations is given. In addition to these amounts, charts of the results of the analysis of the companies can be found in figure 3 of the appendix.
Figure 3
Although Hong Kong and South Africa are part of the former British Empire, the scores on the different dimensions of national culture differ from each other. Hofstede (2005) suggests that for the individualism dimension an explanation could be found in wealth. The more wealth in a country, the higher the score on IND. Beside this argument, colonization could also fortify the differences in culture. It is imaginable that inhabitants of Hong Kong and South Africa deduct influences from their former masters.
The above described explanations may account for differences in scores on cultural dimensions between countries.
In the next part of this chapter, statistical analysis is applied in order to test the relation of the scores on national culture and the scores on the remuneration elements.
For a general view of the popularity for certain remuneration elements, table 17 below presents the average scores on remuneration dimensions for the selected countries.
Table 17 Average scores remuneration dimensions
Country Variable remuneration Short-term remuneration
Hong Kong 41,65% 82,36%
The Netherlands 51,23% 69,5%
Sweden 51,68% 73,88%
United Kingdom 27,35% 72,84%
4.2. Empirical results
To test the relationship between the five dimensions of national culture and the established elements of the remuneration system, the statistical software program SPSS 16.0 is used.
The research made use of a regression analysis. This is a proper analysis when a (linear) relation between two (or more) interval- or ratio scale variables is expected. In this research it is supposed that the higher/lower the score on a cultural dimension, the higher/lower the score on a remuneration elements. Therefore a regression analysis is a proper tool to apply in this study. The regression analysis tries to find a numeric explanation for the dependent variable by one or more independent variables (Huizingh, 2006). In this case the dependent variable is the remuneration element, while the independent variable is the score on the dimensions of national culture. The assumption made for this analysis is the causal relation between national culture and a specified remuneration element.
Regression analysis is possible because national culture as well remuneration elements are measured on an interval/ratio scale base.
First of all the dimensions of national culture are tested for their mutual correlations. Although the regression model doesn’t make use of combinations of these variables, the correlations can give insights for explaining certain results on hypothesises.
A strong negative relation between the variables power distance individualism and individualism -long-term orientation is found. Since these combinations of variables are opposites of each other, this relation was expected. Other combinations of the independent variables don’t show significant relations.
4.2.1. Regression analysis
This research is especially interested in the influence of the different dimensions of national culture on the type of remuneration. This relation is for all 5 dimensions tested by a linear regression.
The first linear regression tested (H1), which argues that organizations from countries with a high score on UAI have a preference for fixed remuneration. Table 18 in the appendix shows the result of this regression analysis.
Table 18
Looking at the box plot given in figure 4 in the appendix, one could see that the countries Hong Kong and Sweden with both a score of 29 on UAI are responsible for breaking-trough a straight line.
Figure 4
Filtering both countries out of the data and by performing a new regression analysis, the expected positive relation between uncertainty avoidance and fixed remuneration is found. This is showed in table 19 of the appendix.
Table 19
This relation (t = 0,2020, p < 0,052, n = 33) in the new regression model is even significant at a significance level of 10%. Given this regression model, (H1) would be accepted. This new finding will be reviewed and declared in the conclusion/discussion part. In that part there also will be given an explanation for the fact that filtering organizations from Sweden and Hong Kong resulted in a significant relation which accept (H1).
The second regression analysis is executed in order to test (H2). This hypothesis states that the higher the score on PDI in a country, the higher the preference for fixed remuneration in organizations from these countries. Table 20 indicates that there is no significant relation (t = -0,335,
p < 0,739, n = 58) between these variables.
Table 20
On the basis of table 20, it can be concluded that (H2) is not supported. Looking at figure 5 of the appendix, there is a clear increasing line visible, which indicates that the higher the score on PDI, the more preference for fixed remuneration.
Figure 5
However, the remuneration in Swedish organizations, which is represented by the first cube, is exceptional. Performing a new regression analysis without the companies from Sweden, gives different outcomes as one can see in table 21 (t = 2,301, p = 0,031, n = 45).
Filtering out Swedish organizations result in a significant positive relation with a R square of 8,5% on a significance level of 5%. This new model would therefore have accepted (H2). An explanation for this finding will be given in the conclusion/discussion part of this study.
A third regression was performed in order to test (H3). In general this hypothesis argues that organizations from countries with a high score on IND will have a preference for variable remuneration. From table 22 it can be concluded that this relation is a positive one (t = 1,633, p = 0,108, n = 58), but not significant at a level of 10%.
Table 22
Although the relation is not significant, the sign of the coefficient is positive, which could be expected referring to the evoked hypothesis based on arguments from literature review. However, because of the insignificance of the relation, (H3) is rejected.
Even as with (H1) and (H2), Sweden is exceptional in this hypothesis. One could observe this by looking at figure 6, which draws the box plot of this premise.
Figure 6
Sweden is the country that scores 71 on IND. Leaving the observations of Swedish organizations out of the test, a significant positive relation (t = 2,312, p = 0,026, n = 45) is the consequence. Therefore in the case that Swedish companies are filtered out of the test; hypothesis 3 would have been accepted. In the conclusion/discussion part of this thesis an explanation for this finding will be given.
(H4) is tested with a regression that contains the variables masculinity and variable remuneration. The results in table 23 reveal that there is a significant positive relation (t = 3,147, p < 0,002, n = 58) between those variables.
Table 23
Because of the significant and positive relation between the variables it can be concluded that organizations in countries with a high score on the MAS dimension have a preference for variable remuneration. On the basis of table 23, it can be concluded that (H4) is accepted. The R squared for this model is 0,137 = 13,7%.
score on PDI will be averse to share payments in comparison with organizations from countries with a low score on this dimension. Table 24 indicates that this relation is absolutely not significant (p < 0,963). Therefore hypothesis five is rejected.
Looking at figure 7 in the appendix, Sweden with a score of 31 on the PDI is responsible for the rejection of hypothesis five.
Figure 7
Filtering out Sweden from this regression model results in the consequences described below. In this new model, the negative relation (t = 1,639, p < 0,109, n = 43) that one could expect is almost significant. Even as in hypothesis one, two, three and four, Sweden is responsible for an interference of the original regression model. So by excluding Sweden from the regression model, the expected negative relation between a high score on PDI and share-based payments is found. An explanation for the fact that filtering Sweden out of the model results in this negative relation will be given in the conclusion/discussion part.
The last hypothesis that is tested states that organizations in countries with a high score on long-term orientation will be more likely to pay their directors on the long-term. Table 25 shows that the findings are in opposite direction as the expectation. The negative significant relation (t = -3,337, p < 0,002, n = 47) consequently rejects (H6). In the conclusion/discussion part explanation for this unexpected relation will be discussed.
Table 25
Table 26
Regression model of variable remuneration, performed for each cultural dimension
Variable Predicted sign Coefficient T-Statistic P-Value R squared in %
Uncertainty Avoidance (-) 0,006 2,477 0,016 8,3% Power distance (-) 0,000 0,335 0,739 -1,6% Individualism (+) 0,002 1,633 0,108 2,8% Masculinity (+) 0,003 3,174 0,002 13,7% Table 27
Regression model of short-term remuneration, performed by cultural dimension
Variable Predicted sign Coefficient T-statistic P-value R squared in %
Long-term
orientation (-) 0,003 3,337 0,002 18,1%
Table 28
Regression model of stock remuneration, performed by cultural dimension
Variable Predicted sign Coefficient T-statistic P-value R squared in %
Power distance (-) 0,00008 0,047 0,963 -1,8%
5. Critics on Hofstede dimensions of national culture
In the past, many researchers used the dimensions of national culture evoked by Hofstede for their studies related to national culture practices. Tosi and Greckhamer made use of the dimensions in their research on the influence of national culture and CEO compensation (2004). In his study to describe differences in employee attitudes in collective versus individualist national culture, Sims used the cultural measures developed by Hofstede. And also in research on cross-cultural management, the dimensions are applied (Pennings, 1993). As the previous examples describe, in a wide variety of study areas, the Hofstede dimensions are used. Despite of Hofstede’s popularity in relation to anthropological studies, there is permanent critique on the research and development of the dimensions of national culture. In this section a closer look at these critiques will be given and these remarks will be related to this current study.
There are a number of different matters on which the frequently used dimensions could be evaluated. One of these is the research methodology used by Hofstede to come up with his four (later five) dimensions. In his article ‘Hofstede’s model of national cultural differences and their consequenses’, Brendan McSweeney (2002) evaluated this research methodology. First of all, McSweeney stated that the number of questionnaires used in the IBM research reveals that the average number per country was small, and that for some countries it was minuscule. Therefore, no proper conclusions according to these questionnaires could be made.
Another point of critique stated by McSweeney involves the old data from the IBM survey. This argument was discarded by Hofstede as recent replications show no loss of validity (2002). Hofstede states that national cultures are extremely stable over time. Rosalie R Tung (2008) counters this argument by arguing that when countries develop economically and are acting in an intensified globalization atmosphere, differences between national cultures will decline (Ralston et al. 1993, Egri and Ralston 2004, Von Glinow et al. 2004).
Another objection following McSweeney is the narrowness of the Hofstede study. All respondents were exclusively from one single company, so this can hardly reflect entire national cultures. Further, the survey is based on five assumptions. A failure of one of these means a shortcoming in the conclusions.
lot of shortcomings on the dimensions evoked by Hofstede. McSweeney is not the only author who throws light on the weaknesses of the Hofstede dimensions.
Chiang (2005) states that the dimensions of national culture ignore the potential influence of a variety of contextual factors. The study of Chiang demonstrated that cultural differences between countries could offer some insight intro the preference for certain rewards and HR practices, but that there should be many other factors that play a role in this case. This is an important note regarding to this thesis. Despite the fact that some hypothesises are accepted, some unexpected relations are found. One way to account for these unpredicted relations is by indicating the presence of other factors than cultural ones. In the discussion/conclusion part of this study, possible explanations for these unforeseen relations will be given. Now the critiques on Hofstede’s dimensions are discussed, it could be interesting to review other dimensions of national culture that could be used for study on cultural practices.
Fons Trompenaars, another Dutch researcher developed seven new dimensions of national culture with Charles Hampden-Turner (1993). Five of these dimensions are specified for human beings, while the remaining two deal with societies and the environment. Compared to the dimensions of Hofstede, only the communitarianism/individualism dimension of Trompenaars and Hampden-Turner shows significant consistent features with the collectivism/individualism dimension. Table 29 in the appendix gives an overview of the different dimensions. For completeness the dimensions developed by Schwartz (1994) are also added. These dimensions are based on a survey in 64 countries involving more than 60.000 individuals. Displaying the individual value profile or the comparison of cultural value dimensions between different countries and regions or groups (i.e. a working teams) helps to understand the context of intercultural meanings at the workplace or in private life and establish fruitful and effective relations6.
Hodgetts and Luthans (2000) stated that the differences between theories about national culture could be explained by the difference of moment of research. In other words, they stated that the Hofstede dimensions could be history and that the Hampden-Turner and Trompenaars dimensions are better to use for current research. Future research therefore could apply the one of the other theories in order to research whether this show significant differences.
There is a lot of doubt about whether the theory of Hofstede is applicable to HR practices. In their study, Chiang and Birtch (2007) found that pressure from competitive isomorphism is driving companies toward greater consistency and standardization in management practises and that therefore national culture is converging. This is confirmed by the authors mentioned above.
6
