A Cross-Cultural Study of Leadership Attitudes in Three Baltic Sea Region Countries

Audra I. Mockaitis

Victoria University of Wellington, New Zealand

This study adds to previous research on the influence of national cultural values on leadership attitudes and introduces new findings for three countries - Lithuania, Estonia, and Poland. The attitudes of managers and employees toward various aspects of leadership in these three geographically and historically similar countries are, in fact, different. Significant differences were found across countries on the following variables: control, supervision, view toward authority, commitment, decision type, initiative, preferred leader orientation, and leadership style. It is revealed that differences in attitudes are closely related to differences in country positions on Hofstede's cultural dimensions. This study provides interesting cues for managers working in the region.

Effective leadership depends on the understanding by leaders of the attitudes, expectations, and behaviors of subordinates in various situations. Subordinates must also be motivated to satisfy leaders' expectations of them. This also means that leaders must understand the needs and expectations of subordinates. Thus, the leader should lead in ways both acceptable and encouraging for subordinates to strive for organizational and personal goals and objectives. Traditional studies of leadership, however, have generally focused on leader styles and behaviors, and not on the congruence between leader and follower behaviors, expectations, and attitudes. Many studies have concluded that leaders will be effective in certain situations, provided that certain contingencies are met. However, these contingencies have often been vaguely described.

With the internationalization of business, it became increasingly evident that certain methods or styles are not as effective elsewhere as they are at home. This, in turn, sparked interest in cross-cultural management studies in which various theories were tested or previous studies were replicated in several countries. The main focus of these studies was to test whether differences exist in attitudes toward management aspects across countries. Since the first comparative management studies (e.g., Haire, Ghiselli, & Porter, 1966; Hofstede, 1984 - study conducted from 1968-1972), a number of studies have addressed attitudes toward different aspects of leadership in different countries (Bochner & Hesketh, 1994; Brodbeck et al., 2000; Offermann & Hellmann, 1997; Smith, Peterson, & Wang, 1996; Suutari, 1996). Although they have produced varying results, most have confirmed that national culture influences attitudes toward and behaviors of leaders. Among the most researched topics are attitudes toward leadership and decision-making styles. These are also the focal topics in this study. Consequently, it is no longer necessary to prove that differences in attitudes toward leadership exist across countries, but rather, it is more important to reveal how organizational attitudes are influenced by culture and to describe this relationship.

However, we still know relatively little about employee attitudes and preferences in the former socialist countries. Cross-cultural research is a new field in Central and Eastern Europe (CEE). To date, little research has been conducted on cultural values and their influence on aspects of management in the CEE countries (Ardichvili & Kuchinke, 2002). Most of the studies on leadership, to date, have been conducted in the United States and Western Europe (Yukl, 1994). More recent studies, which have included Eastern Europe, are those of Nasierowski and Mikula (1998) (a replication of Hofstede's 1984 study in Poland); Suutari's (1998) comparison of Estonian, Russian, and Finnish leadership behaviors; a study of human resources practices in 10 countries by Aycan et al. (2000); and the GLOBE study (Brodbeck et al., 2000; Koopman et al., 1999). Little is still known about the Baltic countries. This study fills this gap by examining the influence of cultural values on attitudes toward leadership in Lithuania, Estonia, and Poland. Since this is the first such comparative study in these countries, it is important to provide an accurate picture of the relationship between culture and leadership attitudes. Thus, the aim of this exploratory study is to compare employee attitudes toward supervision and control, views of authority, organizational commitment, types of decisions preferred, taking initiative and leadership styles across the three countries.

Background

Over the last three decades, there has been witnessed a gradual increase in both quantity and quality of comparative leadership studies; however, many of the studies, which address apparently similar issues, report different findings. This may be due, in part, to the differences among studies in their measurement of leadership dimensions (Bass, 1990). Another reason for variation lies in lack of uniformity in the designs of the studies. The studies of the 1960s and 1970s generally attributed variations in country mean scores to cultural differences, without operationalizing the independent variables or theoretically justifying cross-national differences. More recently, studies have related aspects of leadership to specific cultural variables. The following sections introduce the independent and dependent variables in this study and provide an overview of some of the earlier comparative studies from which the dependent variables are drawn.

The Independent Variables

In this study, culture is defined as a set of learned values shared by a group, which influences the group's way of life, including its perceptions (beliefs and attitudes) and behavior, and distinguishes one human group from another (Mockaitis, 2002). Hofstede's (1984) dimensions of culture are used to operationalize the independent variables. Although more recent studies, such as the GLOBE project (Koopman et al., 1999), the dimensions of Schwartz (1994) or Trompenaars (1993) exist, the dimensions of Hofstede (1984) have been chosen to operationalize culture in this study, as this is the only study, to date, in which data are provided for Lithuania, Estonia, and Poland.

Based on survey research conducted from 1968-1972 on IBM subsidiaries in 40 countries, Hofstede (1984) was able to distinguish four factors, or dimensions, on which countries more or less differ: power distance, uncertainty avoidance, individualism, and masculinity. Power distance (PDI) is defined as "the extent to which the less powerful members of institutions and organizations within a country expect and accept that power is distributed unequally" (Hofstede, 1991, p. 28). Uncertainty avoidance (UAI) refers to "the extent to which the members of a culture feel threatened by uncertain or unknown situations" (Hofstede, 1991, p. 113) and their ways of coping with this uncertainty. Individualism (IDV) refers to the relationships between an individual and society, whether these ties are loose (high IDV) or whether individuals form tight-knit groups early on in their lives (low IDV). Masculinity (MAS) distinguishes between those societies in which gender roles are distinct and "masculine" values dominate (high MAS) and those in which gender roles overlap (low MAS). A fifth dimension, long term orientation, was later added by a team of researchers (Chinese Culture Connection, 1987), but data concerning this dimension is unavailable for Lithuania, Estonia, and Poland.

Although Hofstede's work has been criticized for its use of one multinational corporation in drawing generalized conclusions about societies (Triandis, 1982) and that the corporate culture may have had an affect on the results (Shackleton & Ali, 1990), the patterns of relationships found by Hofstede among countries have been shown to remain over time in replication studies (Hoppe, 1998). Hofstede (2001) has provided over 400 external validations of his dimensions. To date, this is the largest and most cited study on national cultural values. The inclusion of findings from several extension and replication studies in new countries currently brings the total number of countries (and regions) in the database to 70.

Table 1
Indices for Lithuania, Estonia, and Poland on Four of Hofstede's Cultural Dimensions

Ronen and Shenkar (1985) have grouped countries together into country clusters based on previous cross-cultural studies. Since no data were available at the time on Lithuania, Estonia, and Poland, these are not included in any country clusters. Ronen and Shenkar (1985) speculated that countries similar in geographic proximity and institutional environment also share cultural values. It can be seen from Table 1 that there are quite a few similarities between the three countries on certain cultural dimensions. For example, the individualism dimension is similar for Estonia and Poland, masculinity is similar for Lithuania and Poland, and the power distance and uncertainty avoidance dimensions are similar for Lithuania and Estonia. However, although the three countries shared a similar institutional environment, this may actually be used to better reveal true cultural differences.

The Dependent Variables

As with the definition of culture, there are about as many definitions of leadership as there are authors attempting to define it (Stogdill, 1974). The fact that the concept of leadership contains many dimensions adds to this difficulty. Leadership has been defined as a focus of group processes, personality, compliance, influence, behaviors, persuasion, and power relations (Bass, 1990). Here, leadership is defined as "an interaction between two or more members of a group that often involves a structuring or restructuring of the situation and the perceptions and expectations of group members" (Bass, 1990, p. 19). An important aspect of this definition is the stress on perceptions and expectations of group members, which is particularly important in cross-cultural interactions. Consequently, actual behaviors of managers and subordinates will not be addressed in this study, but rather, their general beliefs and preferences.

The dependent variables in this study were drawn from several earlier comparative studies. Table 2 provides an overview of some of these earlier studies and the items used to measure leadership variables. Because of the dearth of research in Central and Eastern Europe, I sought to address those aspects of leadership which have been previously measured in other comparative studies in order to provide a better theoretical basis from which to formulate hypotheses.

All of the studies, with the exception of Smith, Peterson, and Wang (1996), measured relatively similar aspects across five or more countries; however, rather different results were obtained. Bass, Burger, Doktor, and Barrett (1979) discussed cross-cultural differences (specifically Hofstede's power distance and uncertainty avoidance dimensions); however, they did not discuss their findings in relation to cultural differences. The later studies, however, all made predictions regarding how their results should differ based on country differences in cultural dimensions, mainly those of Hofstede (1984). Although only Suutari (1996) defined culture, these studies all operationalized culture to some extent by providing definitions of Hofstede's variables and relating their findings to these.

Another noticeable difference lies in the measurement of the variables. Some authors (e.g., Suutari, 1996) measured the degree of agreement with general statements, while others (Bass et al., 1979; Offermann & Hellmann, 1997; Pavett & Morris, 1995) measured both behavior and attitudes. Effective leadership depends on the manager's or leader's ability to recognize subordinate needs and expectations and adjust methods to subordinate perceptions of effective leadership. Of the above studies, however, only Bochner and Hesketh (1994) addressed employee attitudes, while the remaining studies involved managers. Here too is a gap in comparative studies, where studies of employee attitudes are few. In light of these differences, it was not possible to directly apply the measures used by previous researchers in this study; however, where possible, similar statements were formulated regarding ideal or preferred, but not actual, leadership behavior.

Table 2
Summary of Leadership Aspects Addressed by Earlier Studies

Research Hypotheses

Culture will influence perceptions about the role of the leader in organizations, the interaction between leader and follower, the types of decisions preferred by superiors and subordinates, and the degree of participative decision-making seen as acceptable. It is expected that Hofstede's cultural dimensions will be useful in predicting differences in Lithuanian, Estonian, and Polish attitudes on different leadership factors. Previous studies (Bochner & Hesketh, 1994, Offermann & Hellmann, 1997; Suutari, 1996) have already found these dimensions to be useful in explaining preferences for participation in decisions, conflict management, supervision and control, leadership styles, and leader orientation. Thus, these studies provide a useful base from which to draw conclusions regarding the relative views of employees in the three Baltic countries toward similar aspects.

Power distance is relevant to the study of leadership because it deals with preferences for, expectations of, and relationships to authority (Offermann & Hellmann, 1997). In high power distance societies, inequality, expressed as differences in status, is the societal norm. On the other hand, where power distance is low, status differences are not accentuated and authority is based on the skills and knowledge of the individual (Hofstede, 1984, 2001). Smith et al. (1994) asserted that managers in high power distance societies use more rules and procedures than those from lower power distance countries. In addition, high uncertainty avoidance induces subordinates to be reluctant to exercise autonomy and accept responsibility, which leads the manager to exercise greater control and provide more detailed instructions than are actually required (Kanungo & Mendonca, 1996). In fact, in their study of managers from 19 countries, Offermann and Hellmann (1997) found statistically significant positive correlations between uncertainty avoidance and leader control and negative correlations between both uncertainty avoidance and power distance and leader delegation and approachability. Den Hartog et al. (1997) found that Poles have a more positive view of autocratic behavior and status consciousness and risk avoidance than in countries where power distance and uncertainty avoidance are lower. In line with these views, the following two hypotheses are presented.

Hypothesis 1: Supervision and control will be acceptable in Poland the most and acceptable in Estonia the least.

Hypothesis 2: Leadership based on status will be acceptable in Poland the most and acceptable in Estonia the least.

In individualist societies, leadership is about leading individuals who seek their own self-interests based on their needs. In collectivist societies, leadership is a group phenomenon. The group must be an effective unit in order to be led. Subordinates will bring forth loyalty in exchange for security and protection from their superior (Hofstede, 2001). Loyalty will also be higher in societies with higher uncertainty avoidance where security is highly valued. The dependence relationship between superior and subordinate in high power distance societies may also be manifested in increased employee commitment when employees have to depend on superiors for instructions, rewards, and sanctions. Considering that Poland is far higher than Estonia and Lithuania on uncertainty avoidance and power distance and only slightly higher than Lithuania on individualism, the following hypothesis is presented.

Hypothesis 3: Poland will score the highest on preference for leadership based on commitment and loyalty followed by Lithuania and Estonia.

Differences are also hypothesized regarding the way decisions are made, that is, whether decisions are based on logic or intuition and experience. Hofstede (1984, 2001) proposed that this is exemplified in the masculinity dimension. In more feminine societies, decisions will be based on intuition and consensus, while in highly masculine societies, they will be based on facts and logic. However, higher uncertainty avoidance also implies that decisions are based on facts and information, as this provides greater control over the decision outcomes. Since Estonia is lowest on masculinity and uncertainty avoidance and Poland scores highest on uncertainty avoidance and high on masculinity, it is expected that Estonians will most and Poles will least prefer intuitive decisions.

Hofstede (1984) and other researchers (e.g., Pavett & Morris, 1995; Suutari, 1996) have found correlations between power distance and preferred leadership styles and participation in decision-making. A higher power distance has been found to be associated with more authoritarian attitudes, while lower power distance societies prefer more consultative styles. On the other hand, individuals in high power distance societies also often indicate a preference for a participative/democratic leader and participation in decision making, indicating a higher enthusiasm for what they least have. In countries where power distance is lowest, initiative in participative decision-making is expected to be taken by subordinates, and we would, thus, expect a consultative or democratic manager to be most preferred. In feminine societies, as already mentioned above, there is a preference for cooperation, but also for decisions by consensus. Thus, participative decisions may be more preferred in these societies, as well as a consultative or participative manager most valued.

The masculinity dimension is also useful in explaining preferences for a task- versus relations-oriented manager. However, the higher the power distance, the more formal the relationship is between superior and subordinate. Hofstede (1984) found a relationship between the preference for task- and relations-oriented managers and uncertainty avoidance. In high uncertainty avoiding societies, managers are seen as more task-oriented, as a lower tolerance for ambiguity is expressed in controlling tasks, deadlines, and meeting performance objectives. Here, Estonia scored the lowest on power distance, uncertainty avoidance, and masculinity, while Poland scored the highest and Lithuania in between. Thus, participation in decisions and a democratic or participative manager will be preferred in Estonia the most and preferred in Poland the least.

Hypothesis 4: Estonia will score highest and Poland the lowest on preference for intuitive decisions.

Hypothesis 5: Estonians will prefer a relations-oriented manager the most, and Poles will prefer this type of manager the least.

Hypothesis 6: Estonians will prefer a democratic/participative leadership style the most, and Poles will prefer this leadership style the least.

Initiative is characteristic of individualist societies. It should not be expected that members in lower individualist societies display initiative, as this would jeopardize their membership to the in-group. When an individual displays initiative, he/she demonstrates a willingness to accept greater responsibility for one's actions and a desire to perform. This is also characteristic of masculine societies, which stress performance and challenge - the more uncertain the outcome of a certain decision, the greater the risk. Thus, an individual's ability to take the risk involved depends on his/her ability to cope with uncertainty. Although Estonia and Poland scored higher on individualism than Lithuania, and both Lithuania and Poland were relatively masculine, Poland scored considerably higher on uncertainty avoidance compared to Lithuania and Estonia, indicating that Poles will be least inclined to demonstrate initiative. Based on these country positions, the following hypothesis is proposed.

Hypothesis 7: Estonians will favor opportunities to show initiative the most, and Poles will favor this the least.

Methods

Many empirical studies may be termed national studies because of the lack of attention to methodological requirements, namely, the isolation of the influence of culture (Kelley & Worthley, 1981). Several steps have been taken to isolate and measure the effect of culture on the dependent variables in this study and are outlined in detail below.

Sample

One of the main problems in cross-cultural research is controlling that variance in the data that is truly attributed to cultural differences. Age, gender, and occupation are among some of the variables that may influence the results. Thus, controlling for background variables is a necessary step in cross-cultural research. One way of ensuring that demographic differences are minimized is to match samples as closely on as many variables as possible. Another method is to apply control of background variables in the statistical analysis, in this way "teasing" out any differences in samples. Both of these methods were applied in this study. The sample consisted of 90 respondents in eight organizations in three countries: four companies in Lithuania, three in Estonia, and one in Poland. Not only were respondents in the organizations matched, but the organizations themselves were similar as well. The influence of contextual factors, such as historical or institutional environment influences were also considered to be minimal, as the three chosen countries share a similar institutional environment, level of economic development, and the like. Because of the similarities between countries, any observed differences are more likely to be a result of cultural differences (Van de Vijver & Leung, 1997).

Choice of organizations. Study participants worked in comparatively young organizations in a single industry, advertising. Because the advertising sector is relatively new in the Baltic countries, it was expected that the average age of employees in the organizations would be relatively low and similar across all organizations. Since the studies of cultural values in the three countries were based on student samples, I sought companies in which the average age of employees was similar to the average age of respondents in the cultural values studies, thus, minimizing the possibility of generation effects, which are especially apparent in post-socialist societies. For example, the replication study in Lithuania (Mockaitis, 2002) reported a mean age of 23, and in Poland the mean age of the student sample was 29 (Nasierowski & Mikula, 1998).

Second, the newness of the advertising sector in the three countries and its specific nature ensured that the agencies engaged in similar activities, used similar technologies and had similar organizational structures. This was confirmed through interviews with company representatives.

Third, because the organizations are relatively small, it was decided to include several organizations in each country, while still ensuring that they were matched on as many aspects as possible. The organizations were all locally-owned, and shared similar ownership or affiliated status. Three of the organizations in Lithuania and two in Estonia belonged to the same parent group of companies, which was owned by Estonian and Lithuanian nationals and was also part stakeholder in the remaining organizations. One Lithuanian and one Estonian organization were also linked to the Polish organizations by partnership association.

All of the organizations employed mostly nationals. Although the Polish organization was head by an expatriate, it had local ownership, as did its affiliated companies in Lithuania and Estonia. This strategy of choosing organizations with shared ownership or affiliation was the only available option, as multinational advertising agencies have entered these markets only recently. All of the companies also shared a similar client base across the three countries and were all engaged in both creative advertising and media strategy. At the time of the study, there were relatively few leading agencies in the three countries. Most advertising companies served mainly local clients and did not have clear strategies. The companies in this study were all leading companies in the three countries, which had adopted uniform methods of working with clients. Although the companies technically belonged to different parent organizations, their main stakeholders were shared across all organizations in the three countries. This strategy of establishing technically different companies to serve a wider client base was not unusual in these countries, at a time when taxation laws were still changing quite often and profits were kept at a minimum.

Finally, all of the organizations had a similar market share at the time of the study. Thus, with the exception of size, the organizations were matched on several contextual factors: technology, market share, ownership, and organizational structure. This information was also confirmed during preliminary interviews with company management.

Respondent characteristics. The respondent sample consisted of 18 respondents in Poland, 44 in Lithuania, and 28 in Estonia. It was expected that samples would be matched on gender, age, level of education, and job position, considering the specifics of the sector. However, some discrepancies were found. The levels of education in each of the countries were fairly similar, with 75% of respondents in Lithuania, 70% in Estonia, and 78% in Poland having a university degree. The mean age of respondents was 27 in Lithuania and Poland and 29 in Estonia. The gender ratio was 59%, 78%, and 67% female in Lithuania, Estonia, and Poland, respectively. The percentages of respondents in managerial positions were 41% in Lithuania, 29% in Estonia, and 44% in Poland. Because of the uneven distribution of respondents on demographic variables, it was decided to adjust the country means to account for differences in demographic variables by treating them as covariates in the analysis, in this way, ensuring that any remaining differences between the samples be accounted for by culture.

Research Instrument

The data for the study were collected in 2001. The main section of the questionnaire consisted of 34 items asking respondents to indicate the extent of agreement to statements regarding general beliefs about effective leadership on a 6-point scale ranging from 1

(strongly disagree) to 6 (strongly agree). The choice for the use of a 6-point scale was made because the items measured opinions and not actual behaviors, and an opinion was sought for each item. It was deliberately sought to avoid labels such as "neither agree nor disagree," as there is still a higher overall prevailing uncertainty in these countries as compared to western European countries and hesitancy to participate in survey research (Michailova & Liuhto, 2000). This was taken as a precautionary measure even though employees were assured that their individual responses would not be revealed to their colleagues or management.

The next section included two questions pertaining to leadership styles and task versus relations orientation. Respondents were asked to choose, from various descriptions, the manager for whom they would most like to work. The last section contained six demographic questions measuring job position, level of education, gender, age, nationality, and department.

The original version of the questionnaire was developed in English. Back-translation was used to check for inconsistencies in the Lithuanian language version, and small changes were made to the questionnaire before distributing the final version. A drawback of the research instrument was that it was not translated into Polish and Estonian, but only Lithuanian, due to the inability to find competent translators in the languages familiar with the topics, and the inability to then back-translate these versions to check for agreement between versions. It could not be guaranteed that meaning equivalence would be maintained if the surveys were translated without back-translation. Although the use of English-language questionnaires in cross-national surveys has been criticized because of the risk of cultural accommodation (Harzing et al., 2002), it was decided that the original English language questionnaire would be used for Estonian and Polish respondents after interviews with company management in these countries which revealed that the language of communication between companies was English and that command of English by Estonian and Polish respondents was far better than Russian, another possible choice for the language of the questionnaires. As a precaution, the questionnaires were checked for meaning equivalence and instrument and item bias by examining reliabilities, histograms, variance, and standard deviations, as well as theoretical reasoning (comparing the results to hypothesized outcomes). No significant effects of bias on responses were found.

Measures

The constructs in this study were developed based on the analysis of previous cross-national leadership studies.  Questionnaire items were formulated to reflect concepts previously addressed in the literature. Most of the measures were inspired by the studies presented in Table 2, with wording changed to reflect general attitudes to avoid reference to one's direct superior and ensure that items assess employee attitudes toward leadership and not actual behaviors, which might be more a reflection of the working environment of respondents. Suutari's (1996) role clarification measure and Haire et al.'s (1966) sharing information and objectives construct were useful in this study in measuring supervision and control. The studies of Laurent (1983) and Bochner and Hesketh (1994) were useful in developing items reflecting views toward authority and conflict avoidance. Other studies provided help, as well, in developing items measuring preferences for a task versus relations orientation (e.g., Bass et al., 1979; Bochner & Hesketh, 1994) and leadership styles (Sadler & Hofstede, 1972). However, most of the previous studies either did not test their scales through factor analysis or reliability analysis, or they did not report these results. For these reasons and because of the lack of uniformity in samples, constructs and measurement of items in previous studies, it was necessary to develop additional measures.

Table 3 depicts the dependent variables of this study. All questionnaire items were first analyzed and grouped into constructs on the basis of intercorrelations and theoretical reasoning. The internal consistency of constructs was tested using Cronbach's alpha. To ensure that constructs were reliable measures in all three countries, the overall alpha reliabilities were first calculated across countries and then analyzed separately within each country. This strategy allowed for the detection of constructs, which at first glance appeared reliable, but showed low reliabilities in one or more countries.  Constructs with alpha reliabilities lower than 0.60 overall or in more than one country were subsequently removed from the analysis. This resulted in the six composite variables in the table with alpha reliabilities all over 0.62 and two additional variables: task versus relations orientation and leadership style, which were each measured with a single question asking respondents to indicate preferences for leader type from among two choices for a task versus relations orientation and four choices regarding leadership style.

Although the reliabilities within countries for each of the constructs were satisfactory, there were some discrepancies. The 3-item control scale had the highest overall alpha reliability of 0.86 and high reliabilities in Estonia and Poland, yet lower in Lithuania (a = 0.63). The same is true of decision type (a = 0.76) and initiative (a = 0.73), where reliabilities in Estonia and Lithuania were a = 0.58, respectively. The highest consistency, although with a lower overall alpha reliability (a = .63), was for the supervision scale. Although the overall reliabilities are rather satisfactory, where the reliabilities of constructs vary significantly within countries, the results should be approached with some caution.

Table 3
Alpha Reliabilities of Leadership Constructs  

Findings

Table 4 depicts the results for all of the dependent variables. Mean scores for composite variables were calculated as the average of items comprising the scale. A comparison of hypothesized to observed outcomes, highlighting significant differences in mean scores between pairs of countries, is presented in Table 5.

Poland scored the highest on control and supervision and Estonia scored the lowest, as was predicted in Hypothesis 1. Post-hoc tests indicated that differences in mean scores between all countries were significant for control, and that Poland scored significantly higher on supervision than Estonia, as suggested in the hypothesis.

The next variable in Table 4 is view toward authority. Hypothesis 2 predicted that status (as expressed here in acceptance of authority and privileges) will be acceptable in Poland the most and in Estonia the least. Although post-hoc tests revealed that Poland scored significantly higher than both Estonia and Lithuania, the mean difference between Lithuania and Estonia was not significant, and Estonia scored even slightly higher than Lithuania, contrary to what was predicted.

It was predicted in Hypothesis 3 that Polish respondents would demonstrate a clear preference for loyalty to the organization and group commitment. These aspects were grouped in commitment, where Poland scored the highest and Lithuania the lowest, and post-hoc tests indicated that Poland scored significantly higher than both of the remaining countries. Hypothesis 3 was not fully supported by these results.

Table 4
Adjusted Country Mean Scores and Culture Effects for Leadership

A higher score on decision type indicated that intuitive decisions were preferred to facts-based decisions. This was predicted in Hypothesis 4 to be associated with the masculinity and uncertainty avoidance dimensions, where Estonia was expected to score the highest and Poland the lowest on a preference for intuitive decisions. The results were in line with this proposition, and the differences in mean scores between Poland and Estonia were statistically significant. It was also expected that Estonians would not only prefer intuitive decisions the most, but also a relations-oriented leader (Hypothesis 5), indicated by a higher score on the task versus relations orientation. This hypothesis was supported by the findings.

The preference for leadership styles differed significantly across countries. It was hypothesized (Hypothesis 6) that Estonians would prefer a democratic/participative leadership style the most (Type 4 of the four leader descriptions in the questionnaire) and Poles the least. Here, Estonia indicated a preference for a Type 3 (consultative) or Type 4 (democratic/ participative) manager, while Poland most preferred a persuasive (Type 2) manager. Lithuanian respondents preferred a persuasive or consultative style (closer to Type 3). These results, thus, supported Hypothesis 6.

The final hypothesis (Hypothesis 7) pertained to initiative, where Estonia was predicted to score the highest, followed by Lithuania and Poland. The ordering of countries is in line with predicted positions, and post-hoc tests indicated that Estonians do agree far more than Poles that showing initiative leads to individual and organizational success. This could be expected, given the large relative country differences on the uncertainty avoidance dimension. The final hypothesis was, thus, supported by the findings.

Table 5
Pairwise Comparisons of Country Mean Scores and Hypothesized Differences

Main Effects

Most of the findings presented above supported the hypotheses in so far as the ordering of countries was in line with predictions on the dependent variables. However, this did not fully reveal why these differences exist. The proposals regarding country positions on the dependent variables were based on reasoning regarding country positions on four of Hofstede's cultural dimensions. Thus, this reasoning may be justified if culture effects are found on the dependent variables.

Analysis of Variance was used to test the main effect of culture on the dependent variables. However, since background variables may also explain some amount of cross-cultural differences, a procedure suggested by Van de Vijver and Leung (1997) was used to test the main effects of culture in the presence of other variables. The procedure consisted of two stages. An ANOVA was first conducted to test the null hypothesis of no cultural differences on the dependent variable and the effect of culture is measured (F1). Demographic variables were next introduced as covariates and the effect of culture after control of these variables was measured again (F2). These values were then compared. Where the effect of culture before and after the introduction of covariates does not differ significantly, cross-cultural differences cannot be accounted for by the context variables. Where F1 was significant, F2 is smalle, yet significant as well, the context variables provided a partial explanation for mean differences. And, if F2 is not significant, mean differences may be entirely attributed to the context variables. In those instances where background variables are seen to have an effect on mean differences, separate ANOVAs were conducted entering the background variables each separately as covariates, and the F -ratios before and after their introduction were compared to the original F1 values once again, to isolate those variables, which have the strongest effects. To verify that these background variables indeed influence mean scores, separate ANOVAs were then conducted, treating the identified demographic variables as independent factors. This method provides an efficient way to identify which external variables affect mean differences.

It may be seen in Table 4 that the most significant initial effects of culture (F1) were found for the variables control, view toward authority, decision type, initiative, task vs. relations orientation, and leadership style. A smaller effect was found for commitment, and the effect on supervision was insignificant. It became clear from the table that were we not to isolate culture and conduct the analysis again with the inclusion of covariates, the results could be misleading. It is clear that for some of the variables, differences on the demographic variables interfere and do not allow us to see the clear picture.

After the introduction of covariates (F2), it is seen that some of the variables were influenced wholly by culture (e.g., view toward authority, initiative, task vs. relations orientation, and leadership style). Where the influence on supervision was not significant before statistical control, mean differences on this variable were now attributed to cultural differences. Here, only education (F = 8.17, p < .001) had a large initial effect on supervision, which is not surprising, as Hofstede (1984, 2001) had also found a negative relationship between power distance and level of education. Demographic variables partially explained differences on commitment and decision type. It may be said that the evidence of culture effects for the above variables assisted in confirming hypotheses 1, 4, 5, 6, and 7; and the lack of significant effect of culture on some of the factors helped to explain why the remaining hypotheses were not fully supported.

It makes sense to identify other explanatory variables and identify their effects. Table 6 depicts the two variables identified as partially explained by demographic variables and their F2 values. It became evident that commitment was slightly affected by age and more so by job position, and decision type was influenced by age and more so by gender. Separate ANOVAs were conducted, treating each of the identified demographic variables as independent factors, to determine which of these were most accountable for differences in mean scores. The ANOVAs revealed that commitment was most affected by job position, while the mean differences on decision type were partially explained by gender. None of the F -ratios for the remaining demographic variables exceeded 1.00. Also, the only highly significant effect was that of gender on decision type (F = 10.71, p < .01). Thus, the effects of demographic variables on the dependent variables were only slight. Aside from the influence of gender on decision type, it may be concluded that the cultural influences overrode influences of demographic variables.

Table 6
The Effects of Demographic Variables on Select Leadership Factors

Discussion

This study suggests that even apparently culturally "close" countries exhibit variations in attitudes about leadership, which may be explained by cultural differences. The former command economy countries, despite a relatively shared institutional environment, were quite different with respect to national cultural values and attitudes with Lithuania midway between Estonia and Poland on almost all dimensions (with the exception of the masculinity dimension). Culture was found to wholly explain mean differences on five dependent variables and partially explain differences on three. The relative country positions on many of the leadership factors were in line with expectations, which were based on previous research and lend support to the usefulness of Hofstede's cultural dimensions. Some implications of the results are discussed next.

Supervision

Strict supervision will cause the most negative consequences in Estonia, where relationships are less formal and superior-subordinate relationships are based on interdependence and trust. Employees in Estonia are willing to show initiative, express their opinion, and prefer a democratic leader. On the other hand, Lithuanians prefer a manager who is persuasive or consultative, who depends less on subordinates than in Estonia. Consequently, relationships may be more formal and the manager will be given more control than in Estonia. In Poland, even stricter supervision will be acceptable in comparison, as this leaves little room for subordinates to make mistakes and lessens the need for initiative.

Leader Style and Authority

Employees in Poland prefer a persuasive leader who makes the decisions and informs employees about them. Participative decision-making will be most encouraged by superiors and accepted by subordinates in Estonia. Here, a low power distance implies that superiors and subordinates are interdependent, and subordinates are willing to take the individual initiative associated with decision making because of high individualism. The low masculinity of Estonia is also characterized by cooperation and consensus. In Lithuania, with a higher power distance, high masculinity, and lower individualism, employees may prefer participative decision-making; however this will take a different form. The superior may be expected to take the initiative or decide when participation is necessary and provide guidance in the process. In Poland, on the contrary, supervision and decision-making are inherent in the hierarchy and prerogatives of authority.

Teamwork

It is also likely that reaching consensus in teams will take the longest in Estonia, where the opinions of each group member will be sought. Individuals are likely to express their opinions and have a lower hesitancy to disagree with other group members. Commitment to the team may not be as high as in the other two countries, as each group member also has individual interests to defend, however optimal solutions will be sought. In Lithuania, membership to the group also determines group members' opinions to some extent, and group members may be more likely to go along with the group despite their personal opinions. Employees may also depend on the leader to have the final say in decisions. The highest commitment to the group was expressed in Poland, however, just as in decision-making, group members may not be able to arrive at solutions without the guidance of someone in authority.

Limitations and Future Research

This study is not without limitations. First of all, the small sample size was a result of regional specifics, as nonexistence of multinational corporations with offices in all three countries limited the ability to find perfectly matched organizations in all three countries. Thus, I had to rely on the next best strategy of finding relatively matched organizations, albeit small in size. However, the control of background variables helped to minimize any remaining differences in the samples. Support was also found for many of the hypotheses, thus lending support to the relevance of the country samples. Another limitation was the inability to find qualified translators (and back-translators), at the time of the study, in all three languages. Although the overall instrument reliabilities were high in all three countries, some minor discrepancies in construct reliabilities were found within countries. Third, the results represent relative similarities and differences in employee attitudes toward various leadership concepts. Thus, the conclusions should be viewed only relative to the other countries in the study. The actual practices in organizations in the three countries and a comparison, not only to cultural values, but to employee preferences, is an area for future research in the region. Replication would also improve the generalizability of the findings.

It is expected that this study will enhance understanding of emerging management practices in the Baltic States and neighboring countries - at least in as much as an understanding of relative similarities and differences in values and attitudes is provided, and that it will not only spark interest in this, as yet, neglected region in comparative management, but also establish a foundation for further comparative research in the region. This may provide new insights for policy-formation by multinationals in the region and especially for those companies which have, thus far, treated the region as a single market.

About the Author

Dr. Audra I. Mockaitis holds a Ph.D. in comparative/cross-cultural management from Vilnius University in Lithuania and a B.A. from the University of Illinois at Urbana-Champaign, USA. Until recently, she has been an associate professor of international business and marketing at Vilnius University. In 2006, she will transition to Victoria University of Wellington, New Zealand as their senior lecturer in international business.

Dr. Mockaitis' current research focuses on the internationalization of firms (SMEs), consumer ethnocentrism, cultural values and their influence on consumer behavior, communication, and leadership in the Central and Eastern European region.

Email: audra@mockaitis.com
URL: www.mockaitis.com

References

Aycan, Z., Kanungo, R. N., Mendonca, M., Yu, K., Deller, J., Stahl, G., & Kurshid, A. (2000). Impact of culture on human resource management practices: A 10-country comparison. Applied Psychology: An International Review, 49(1), 192-221.

Ardichvili, A., & Kuchinke, K. P. (2002). Leadership styles and cultural values among managers and subordinates: A comparative study of four countries of the former Soviet Union, Germany and the US. Human Resource Development International, 5(1), 99-117.

Bass, B. M. (1990). Bass and Stogdill's handbook of leadership: Theory, research and managerial application (3rd ed.). New York: The Free Press.

Bass, B. M., Burger, P. C., Doktor, R., & Barrett, G. K. (1979). Assessment of managers: An international comparison. New York: The Free Press.

Bochner, S., & Hesketh, B. (1994). Power distance, individualism/collectivism and job-related attitudes in a culturally diverse work group. Journal of Cross-Cultural Psychology, 25(2), 233-57.

Brodbeck, F. C., Freese, M., Akerblom, S., Audia, G., Bakacsi, G., Bendova, H., et al. (2000). Cultural variation of leadership prototypes across 22 European countries. Journal of Occupational and Organizational Psychology, 73, 1-29.

Chinese Culture Connection. (1987). Chinese values and the search for culture-free dimensions of culture. Journal of Cross-Cultural Psychology, 18(2), 143-164.

Den Hartog, D., Koopman, P., Thierry, H., Wilderom, C., Maczynski, J., & Jarmuz, S. (1997). Dutch and Polish perceptions of leadership and culture: The GLOBE project. European Journal of Work and Organizational Psychology, 6(4), 387-413.

Haire, M., Ghiselli, E., & Porter, L. W. (1966). Managerial thinking: An international study. New York: John Wiley.

Harzing, A. W., Maznevski, M., Barzantny, C., Castro, F. B., Feely, A., Fischlmayr, I. , et al. (2002). The Interaction between language and culture: A test of the cultural accommodation hypothesis in seven countries. Language and Intercultural Communication, 2(2), 120-139.

Hofstede, G. (1984). Culture's consequences: International differences in work-related values. London: Sage.

Hofstede, G. (1991). Cultures and organizations: Software of the mind. London: McGraw-Hill.

Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and organizations across nations. London: Sage.

Hoppe, M. (1998). Validating the masculinity/femininity dimension on elites from 19 countries. In G. Hofstede, W. A. Arrindell, D. L. Best, M. De Mooij, M. H. Hoppe, E. Van de Vliert, et al. (Eds.), Masculinity and femininity: The taboo dimension of national cultures (pp. 29-43). Thousand Oaks: Sage.

Kanungo, R. N., & Mendonca, M. (1996). Ethical dimensions of leadership. London: Sage.

Kelley, L., & Worthley, R. (1981). The role of culture in comparative management: A cross-cultural perspective. Academy of Management Journal, 24(1), 164-173.

Koopman, P. L., Den Hartog, D. N., Konrad, E., Akerblom, S., Audia, G., Bakacsi, et al. (1999). National culture and leadership profiles in Europe: Some results from the GLOBE study. European Journal of Work and Organizational Psychology, 8(4), 503-520.

Laurent, A. (1983). The cultural diversity of western conceptions of management. International Studies of Management and Organization, 13, 75-96. 

Michailova, S., & Liuhto, K. (2000). Organisation and management research in tansition eonomies: Towards improved research methodologies. Journal of East-West Business, 6(3), 7-46.

Mockaitis, A. I. (2002). The national cultural dimensions of Lithuania. Ekonomika, 59, 67-77.

Nasierowski, W., & Mikula, B. (1998). Culture dimensions of Polish managers: Hofstede's indices. Organization Studies, 19(3), 495-509.

Offermann, L. R., & Hellmann, P. S. (1997). Culture's consequences for leadership behaviour. Journal of Cross-Cultural Psychology, 28(3), 342-51.

Pavett, C., & Morris, T. (1995). Management styles within a multinational corporation: A five country comparative study. Human Relations, 48(10), 1171-1191.

Ronen, S., & Shenkar, O. (1985). Clustering countries on attitudinal dimensions: A review and synthesis. Academy of Management Review, 10, 435-54.

Sadler, P.J., & Hofstede, G. (1972). Leadership styles: Preferences and perceptions of employees in an international firm in different countries. Mens en Ondernerning, 26, 43-63.

Schwartz, S. (1994). Beyond individualism/collectivism: New cultural dimensions of values. In U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method and applications (pp. 85-119). London: Sage.

Shackleton, V. J., & Ali, A. H. (1990). Work-related values of managers: A test of the Hofstede model. Journal of Cross-Cultural Psychology, 21(1), 109-118.

Smith, P. B., Peterson, M. F., Akande, D., Callan, V., Cho, N. G., Jesuino, J., et al. (1994). Organizational event management in 14 countries: A comparison with Hofstede's dimensions. In A. M. Bouvy, F. van de Vijver, P. Schmitz, & P. Boski (Eds.), Journeys into Cross-Cultural Psychology (pp. 364-373). Amsterdam: Swets and Zeitlinger.

Smith, P. B., Peterson, M. F., & Wang, Z. M. (1996). The manager as mediator of alternative meanings: A pilot study from China, the USA and UK. Journal of International Business Studies, 27(1), 115-38.

Smith, P. B., Dugan, S., & Trompenaars, F. (1996). National culture and the values of organizational employees. Journal of Cross-Cultural Psychology, 27, 231-64.

Stogdill, R. M. (1974). Handbook of leadership. New York: The Free Press.

Suutari, V. (1998). Leadership behaviour in Eastern Europe: Finnish expatriates' experiences in Russia and Estonia. The International Journal of Human Resource Management, 9(2), 235-58.

Suutari, V. (1996). Leadership ideologies among European managers: A comparative survey in a multinational company. Scandinavian Journal of Management, 12(4), 389-409.

Triandis, H. C. (1982). Review of culture's consequences: International differences in work-related values. Human Organization, 41, 86-90.

Trompenaars, F. (1993). Riding the waves of culture: Understanding cultural diversity in business. Avon: Bath Press.

Van de Vijver, F., & Leung, K. (1997). Methods and data analysis for cross-cultural research. Thousand Oaks, CA: Sage.

Yukl, G. (1994). Leadership in organizations. Englewood Cliffs, NJ: Prentice Hall.