Please do not quote without permissions from the authors
The Normative Orientations of Civil Servants in China,
Denmark, Taiwan, and the United States
Lotte Bøgh Andersen,
Aarhus University,
lotte@ps.au.dk,
Torben Beck Jørgensen (corresponding author),
University of Copenhagen,
tbj@ifs.ku.dk
Karsten Vrangbæk,
Danish Institute of Governmental Research,
kvr@akf.dk,
Yahong Zhang,
Rutgers, the State University of New Jersey
yahongzh@andromeda.rutgers.edu
Paper presented at the third International Public Values Workshop at Department of Public
Administration, University of Illinois at Chicago, June 2012
The Normative Orientations of Civil Servants in China,
Denmark, Taiwan, and the United States
Introduction
A basic but rarely addressed question in public values research is whether and how public sector
values vary from one country to another.1 We know that public values vary according to the task,
sector, and level of administration within a country (Andersen et al. 2012a; Vrangbæk, 2009), but
comparative studies of public values are rare and limited to countries that are quite similar to one
another, cf. Van der Waal, Pekur, and Vrangbæk (2008). The paper seeks to fill this gap by
comparing public values in China, Taiwan, the USA, and Denmark.2
One might begin by arguing that the core task of public administration – serving a master in ruling a
country – is basically the same everywhere, regardless of how the political master is chosen and
regardless of culture. Thus, one would expect inherently governmental functions to be identifiable
(Garofalo, 2009) and that it is thus possible to determine a set of universal public values. Indeed, a
few comparative empirical investigations point out rather similar sets of public values in a number
of countries (Beck Jørgensen & Sørensen, 2012–13; Palidauskaite, 2011).
In contrast, one might expect public values to vary considerably between countries due to
differences regarding a number of factors, including age of statehood, political history, size of
country, political system, and national culture. In this paper, we focus first on variations,
concentrating on what we see as the most promising factors: variations in the political system and
national culture. In this respect, the selected countries present us with remarkable variation. The
biggest constitutional difference is between China, a single-party state governed by the Communist
Party of China, and the other three countries, which are constitutional democracies. As to cultural
variation, the contrasts between liberal Western culture and Confucian Eastern culture – and
1
Public values are defined as the ideals, coined as principles, to be followed in the public sector when producing a
service and regulating citizens, business firms etc (Beck Jørgensen, 2003), such as legality, political loyalty, honesty
and impartiality.
2
Taiwan is considered on equal terms with a country, even though it is not recognized as such by all nations.
1
corresponding rule and moral-based cultures – are often pointed out (Yang & Van der Wal, 2011).
Moreover, numerous studies have highlighted major differences between groups of countries
(Hofstede, 2001; Schwartz, 1992; World Value Survey).
Of course, the contrasting expectations of value homogeneity versus value diversity can be
combined if we identify a set of universal public values which, in practice, are interpreted
differently because of contextual differences. Following this line of thought, looking for both
differences and similarities becomes important.
Comparative studies in public values can be carried out in several ways. A constitutional approach
to regime values would focus on constitutions and high-level court decisions on constitutional
matters (Overeem, this symposium; Rohr, 1978). A Public Value Mapping approach (Bozeman,
2006) would include a number of other written documents in the chosen countries, such as official
statements regarding policy goals, goal statements in strategic plans issued by government agencies,
and values derived from public budget documents. Standard sociological/anthropological
approaches would include methods such as observation, interviews, surveys, and case studies.
We have chosen surveys for this first preliminary study for two reasons. First, a survey provides an
overview of the differences and similarities, which guides more detailed studies in the future,
possibly utilizing mixed methods as recommended by Fitzpatrick et al. (2011). Second, comparing
public values in countries all over the world is costly, especially if qualitative methods and refined
documentary approaches are used, whereas surveys are rather cost-effective. Naturally, the main
drawback is that surveys do not allow the researcher to dwell on understandings and interpretations
pertaining to respondents’ perceptions concerning value importance.
Our study focuses on which public values civil servants assess as values that should be important in
their country. The implication of this focus is that we do not focus on the extent to which certain
public values are implemented but rather on the normative orientation of the civil servant. Why is
the normative orientation of civil servants important? First, the normative orientation is likely to
reflect the education and socialization of the civil servant in question and thus the greater context.
Second, the civil servants are unlikely to operate in ways that are totally detached from their
normative orientation. On the contrary, the normative orientation is likely to provide guiding
2
principles that can be referred to in situations of uncertainty and when resolving conflicts. Third,
and perhaps most importantly, the normative orientation possibly reflects the aspirations and
ambitions of a civil servant, thereby providing an idea of civil service self-perception.
As a proxy for civil servants, we use students enrolled in Master of Public Administration programs
in the four countries (public management trainees). These students represent a convenient and
informative group of respondents for this type of comparative study. They have working experience
from the country’s public administration and have completed some level of higher education,
rendering it likely that they are aware of abstract notions of public values and public service
motivation (Svara, 2010). Furthermore, it is convenient to gather survey data as a part of their
education. The weakness is a potential selection bias, such as their being more committed to public
sector values than other public sector employees, but this is probably the same in all of the
countries.
Development of Hypotheses3
For a number of reasons, we do not begin by looking for universal public values. First, the degree of
political (and economic) authority may vary along one or several dimensions. Subsequently, there
are degrees of “publicness,” and some organizations are thus more public than others. Thus,
absolute political authority offers a special case, and there is often a lack of clear boundaries
between public and private organizations (Bozeman, 1987). Second, even if organizations are fully
financed by public money and fully subject to political control, publicness may vary (Antonsen &
Beck Jørgensen, 1997). Third, how to serve a political master efficiently and obediently may vary
due to contextual factors. We will therefore proceed to discuss the contextual factors. But the notion
of inherently governmental functions and corresponding universal public values will serve as a nilhypothesis, and we will return to this idea in the empirical analysis.
Political System
Political systems can be categorized in numerous ways. Distinctions can be drawn between unitary
and federal states, decentralized and centralized states, corporatist regimes versus parliamentary
regimes etc. In relation to public values, we believe that the strongest contrast is between
3
The values in all of the hypotheses presented in this section are selected from the public value questionnaire, see
Appendix A.
3
democracies and dictatorships, in this paper operationalized as multi-party systems versus singleparty systems. In single-party systems, a single political party forms the government, and no other
parties or only approved minor parties are permitted to run candidates for election. Alternatively,
laws or practices prevent the opposition from legally obtaining power. In contrast, multi-party states
have at least two independent parties that are, both formally and in practice, allowed to run
candidates for election and ultimately for office.
The theoretical question, then, is which public values we associate with single-party and multi-party
regimes, respectively. A single-party regime calls for a monolithic, top-down world with a strong
emphasis on hierarchy, flair for interpreting the motives and wishes of the masters, and vertical
climbing within the system. Thus, our first hypothesis includes public values stressing the internal
vertical nature of a system.
H1: Civil servants in countries with a single-party system attach greater importance to the ability to
interpret the political climate and signals, good relations with higher authorities, recognition from
management, a high salary, and career opportunities as compared to civil servants from multi-party
countries.
Next, we expect that a strong presence of these internal values related to verticality crowds out the
attention to external matters, such as citizens and the public-at-large. First, a value conflict between
the internal and external orientation can easily be imagined. Second, a single-party regime calls for
limited public competition over power and a disciplined public discourse.
H2: Civil servants in countries with a single-party system attach less importance to satisfying
immediate users’ needs, public insight and transparency, and consideration of the public opinion as
compared to civil servants from countries with a multi-party system.
Culture
But there are many potential reasons other than the nature of the political system for why certain
public values would be prioritized. Most importantly, a political system can be embedded in a
national culture that influences which public values are important. A common sense understanding
would lead us to think that the West (USA, Denmark) stands in contrast to the East (China,
4
Taiwan). As we will see, however, a more nuanced and theoretical grounded point of departure than
a common sense East–West comparison is available.
Hofstede and his associates (Hofstede, 2001; Hofstede & Bond, 1984) have developed cultural
dimensions such as individualism–collectivism, masculinity–feminism, high–low power distance
and high–low uncertainty avoidance. The USA and Denmark have individualistic cultures, whereas
China and Taiwan have collectivistic cultures. Here, we find an East–West divide. Conversely,
China and the USA have masculine cultures, whereas Denmark has a more feminine culture, and
Taiwan is in-between. On the uncertainty-avoidance dimension, there is very limited difference
between the four countries.
The World Values Survey positions countries on two dimensions: traditional values vs.
secular/rational values and survival values vs. self-expression values. While there is a clear
difference between East and West on the survival/self-expression dimension – which can actually
be related to wealth – both East and West have high scores on secular/rational values.
Schwartz (1992, 1994, 1999) has developed a multidimensional measure of cultural values. He
argues that countries can be placed according to the average scores of their citizens on seven
dimensions: Embeddedness/conservatism, hierarchy, mastery, affective autonomy, egalitarianism,
harmony and intellectual autonomy. Based on an analysis of 74 countries, Schwartz creates a map
with relatively distinct groups of countries. Our four case countries are placed in three different
groups based on their relative emphasis on the seven value dimensions. China and Taiwan belong to
a Confucian cultural group with strong emphasis on hierarchy. The USA belongs to a group of
English-speaking countries emphasizing “mastery.” Denmark is part of the group of European
countries with two core values of “egalitarianism” and “intellectual autonomy.”
There are two reasons for beginning with Schwartz’ cultural theory. First, it is easier to relate our
public values in the questionnaire to the Schwartz theory compared to Hofstede and the World
Value Survey. Second, empirical investigations have confirmed it to be a well-argued theory.
The Confucian cultural group (here, China/Taiwan) is characterized by a hierarchical, differential
allocation of fixed roles and resources being the legitimate, desirable way to regulate
5
interdependencies. People are socialized to comply with the obligations and rules—and sanctioned
if they do not. The cultural emphasis is on the legitimacy of an unequal distribution of power, roles
and resources (social power, authority, humility, wealth). We therefore expect civil servants from
this group to stress public values relating to loyalty, tradition, and the willingness to accept
authority.
H3: Civil servants in a country belonging to the Confucian cultural group attach greater importance
to political loyalty, loyalty to rules, continuity, and adaptability compared to civil servants from
other countries.
The mastery cultural group (here, the USA) is characterized by groups and individuals supposedly
mastering, controlling, and changing the social and natural environment through assertive action in
order to further personal or group interests. The cultural emphasis is on getting ahead via active
self-assertion (ambition, success, daring, competence).
H4: Civil servants in a country belonging to the mastery cultural group attach greater importance to
innovation, the ability to innovate, business-like operations, willingness to take risks, and learning
and development on the job compared to civil servants from other countries.
The cultural orientation towards egalitarianism (here, Denmark) is characterized by individuals
being portrayed as moral equals who share basic interests and are socialized to transcend selfish
interests, cooperate voluntarily with others, and display concern for the welfare of others.
Intellectual autonomy has a cultural emphasis on the desirability of individuals independently
pursuing their own ideas and intellectual directions. Public values relating to egalitarianism would
emphasize treating citizens as equals rather than subjects, for example, and a preference for
recognition from peers rather than managers. Public values relating to intellectual autonomy would
stress the ability to act in a professional manner independent of other actors.
H5: Civil servants in a country belonging to the egalitarianism/intellectual autonomy cultural group
attach greater importance to user democracy, recognition from peers, personal integrity, and
independent professional standards as compared to the civil servants from other countries.
6
Comparing mainland China, Denmark, the USA, and Taiwan
Summing up, the biggest difference in the political systems is between single-party mainland China
and the other three multi-party countries. In terms of culture, the four countries belong to three
different cultural groups, as shown in Table 1, which also lists the countries’ population sizes,
political systems, and the number of respondents from each country in our survey.
Table 1: Investigated countries
Cultural group
Number of
Country
Inhabitants
Political system
China
1341m
Single-party state
Confucian
Taiwan
23m
Multi-party state
Confucian
223
United States
310m
Multi-party state
Mastery
265
Denmark
6m
Multi-party state
Egalitarianism/intellectual
autonomy
respondents
307
455
Research Design and Methods
Analyzing Master’s of Public Administration students
We investigate master’s students for several reasons. First, Master’s of Public Administration
students in the four countries are relatively comparable, even in these very different contexts. They
have completed some measure of higher education and are working toward employment in the
public sector. The educational programs under investigation differ slightly in their relative weight
on different policy analysis and public administration/management approaches to public service
education, but Soto, Opheim, and Tajalli (1999) find limited differences between these types of
educational programs, indicating that this is inconsequential. Still, the comparison between the
questionnaire answers in different countries should be done carefully due to the known differences
in cross-cultural response behavior. “Response set” means the general tendencies to systematically
agree more (or less) with all questions, and it is known to vary between countries (e.g. Hofstede &
Bond, 1984).
7
Second, many of the master’s students have practical working experience in the public sector, and
their values thus reflect the different values and public service motivation. Svara (2010) found that
public administration students are oriented toward contributing to society based on his survey of
Master’s of Public Administration students at five universities, and more than 80 percent agreed
that meaningful public service and the opportunity to help others are important (ibid.). Svara (2010,
p. 361) further argues that generational change in government employment presents numerical,
attitudinal, and organizational challenges, and more knowledge about future public administrators
around the world might help us handle these challenges.
Finally, it is convenient to gather survey data from master’s students as a part of their education. As
discussed below, the response rate was high, and we were able to obtain data from a relatively large
number of respondents.
The weakness is a potential selection bias, as our respondents have actively decided to pursue
public management training and careers and/or receive support from their workplace. This possibly
implies a greater commitment to public sector values than other public sector employees.
Furthermore, it might mean that they are more focused on management values than professional or
user values. However, since we sample from similar respondent groups in all four countries, we do
not expect this to affect our conclusions concerning the similarities and differences between the
countries.
Data and Operationalizations
The questionnaire used in this study has been applied in a number of earlier investigations testing
the dimensions of public values (Andersen et al., 2012a) and the empirical and conceptual links
between public values and public service motivation (Andersen et al., 2012b).
In Denmark, the questionnaires were sent out by post as part of the master’s degree program. A
stamped return envelope was enclosed. In the other three countries, data was collected as part of
lectures held in Public Administration. The response rate in Denmark was 62, while almost all of
the students in the other three countries completed the questionnaires. The number of respondents in
8
each of the countries can be seen in Table 1 above. The questionnaires were translated by nativespeaking researchers and language specialists (the English version of the questionnaire is attached).
In order to measure how much importance the respondents attached to different values, the
respondents were asked to evaluate the importance of (1) general organizational values in pursuing
the daily tasks of their organization, (2) values related to the desirability of certain staff
competences and (3) values about the dominant motivational factors for the staff. Examples are
political loyalty and high productivity (general values), ethical awareness and adaptability (values
related to staff competences), and good relations to users and remaining within the budget (values
related to staff motives). The general values are linked to the performance of daily organizational
tasks, and the respondents are thus encouraged to consider the desirability of abstract values in light
of specific working tasks.
Note again that all questions address what should be, not what is. Our data thus reflects the
desirability of certain normative universes, not the actual presence of certain public values.
Methods of Analysis
Analytically, our strategy is to give an overview of what the respondents actually answered and test
whether their answers differed between the countries. The mean score on a given value in each of
the countries provides a parsimonious summary of the respondents’ answers, but the comparison
involves four challenges. The first challenge is that differences between countries may be due to
differences in language rather than political system or culture. Although the translation has been
validated with native-speaking researchers, culture and language are so closely intertwined that all
interpretations between countries must be carried out very carefully. Still, we can analyze the
respondents’ answers and then discuss whether they reflect an actual difference in normative
orientations. This is also the main reason for primarily presenting descriptive analyses (graphs
showing the mean scores for each of the four countries) and only secondarily using more advanced
statistics.
The second challenge is that value importance is measured on ordinal scales (see Appendix A),
suggesting that advanced statistics are necessary as a supplement to the descriptive graphs. The
available measure of the central tendency for ordinal variables (the median) does not capture
9
nuances in the distribution of answers, which is the reason for using mean scores in the graphs. But
this makes it necessary to supplement with ordinal regressions to make sure that our results are
robust when we use a proper ordinal scale technique. While the key results from these regressions
are mentioned in the result section, all of the details are in Appendix B.
The ordinal regressions also handle the third challenge, namely that the gender composition and
average age differ between the countries. The ordinal regressions control for these variables to
make sure that the differences, if any, are not due to these personal characteristics.
The fourth challenge relates to the fact that the questionnaire included 16 general values, 10 values
linked to staff competences, and 11 values related to staff motives. Comparisons of all these values
between four countries without adjusting for the number of comparisons performed would mean
that we could obtain statistically significant differences by chance, only because of the considerable
number of comparisons. We have therefore theoretically selected a more limited number of
comparisons to present in the paper, and we performed post-hoc tests to adjust the p-values of the
comparisons to the fact that we perform many comparisons. We consistently use Scheffe post-hoc
tests.
Results
The presentation of the results follows the hypotheses, starting with the expectations concerning
single-party versus multi-party systems followed by the hypotheses concerning culture.
Does the Political System Matter?
Following Hypothesis 1, we expect the Chinese civil servants to attach more importance to the
ability to interpret the political climate and signals, good relations with higher authorities,
recognition from management, high salary, and career opportunities compared to the civil servants
from the other three countries.
Concerning the value of being able to interpret political climate and signals, the variation between
the four countries is as expected according to Figure 1. In a Scheffe post-hoc test, these differences
are all highly statistically significant (p < 0.001). This is also the case if the differences are analyzed
using ordinal regression (Table B1 in Appendix B).
10
Figure 1: The Importance of Ability to Interpret Political Climate and Signals
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: Comparison of average importance reported by the participating Master’s of Public Administration students from
the four countries. The averages are calculated from responses between 5 (highest level of importance) and 1 (lowest
level of importance). See Appendix A for the exact question and response categories.
The next value concerns the value of good relations with higher authorities as a factor of
motivation. Figure 2 shows that although the average importance attached to this value is greatest
for China, the difference to the US and Taiwan is not very substantial. Although the differences are
statistically significant in the ordinal regression (Table B1), post-hoc tests indicate that the
difference between Chinese and US respondents is not robust.
Figure 2: The Importance of Good Relations with Higher Authorities
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
11
Recognition from management is expected to be seen as a more important motivational factor for
Chinese core employees. Figure 3 shows that this is actually the case, and post-hoc tests and ordinal
regression confirm that Chinese respondents find this value significantly more important than
respondents from the three other countries.
Figure 3: The Importance of Recognition from Management
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Concerning high salary as a motive for core employees, our expectations are met in Figure 4, and
the logistic regression (Table B1) also shows that the differences are statistically significant
controlled for age and gender. Still, the post-hoc test shows that the difference between China and
Taiwan is not robust when we control for the many comparisons performed.
Figure 4: The Importance of High Salary
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
12
Finally, career opportunities are expected to be seen as more important in China than in the other
three countries. Figure 5 shows that Chinese respondents (as expected) attach greater importance to
career opportunities as a staff motive than the other respondents (right side, yellow bars), and this
difference is also statistically significant for all of the countries (both in post-hoc tests and in Table
B1). However, the respondents from the four countries do not differ in the importance they attach to
whether the organization should ensure career opportunities (left side, blue bars). The ordinal
regression (Table B1) indicates that the difference between Chinese and Danish respondents is
statistically significant after controlling for age and gender, but the substantial difference is
inconsequential.
Figure 5: Comparison of Importance of Career Opportunities as General Value and Motive
Organisation should ensure career
opportunities
Career opportunities as a motive
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
In sum, the bulk of the analyses support Hypothesis 1: Political systems matter. The single-party
system is associated with public values stressing the internal vertical nature of a system. Note also
that Denmark rather systematically scored lowest in most comparisons, cf. Figures 1, 2, 4, and 5
(with regard to career opportunities as a motivational factor), indicating that the Danish political
system is the least vertically oriented system. Finally, we found a clear East–West distribution on
high salary. We will return to these special findings later.
13
Hypothesis 2 expects that civil servants in countries with a single-party system attach less
importance to external actors, such as users and the public-at-large, than the civil servants in
countries with a multi-party system.
Concerning satisfying the immediate needs of users, our expectations are definitely not met. On the
contrary, Figure 6 shows how Chinese respondents attach substantially greater importance to this
value than the other respondents, and both logistic regression (Table B2 in Appendix B) and posthoc tests confirm this. Moreover, a rather clear East–West distribution is visible.
Figure 6: The Importance of Satisfying the Immediate Needs of Users
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Chinese respondents were expected to attach less importance to public insight and transparency
than respondents from the other countries. This expectation is not met. There is no difference
between Denmark, China, and Taiwan, whereas the US respondents score higher than all the other
respondents, as seen in Figure 7 below (Table B2 shows that this difference is statistically
significant).
14
Figure 7: The Importance of Public Insight and Transparency
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
The last part of Hypothesis 2 implies that respondents from states with multi-party systems find
consideration of the public opinion more important than the respondents from a single-party system
(China). Again, this expectation is not met (see Figure 8). Similar to public insight and
transparency, US respondents attach greater importance to granting consideration to the public
opinion than respondents from the other three countries, while Danish respondents attached
significantly lower importance to this value than the other respondents (Table B2).
Figure 8: The Importance of Considering the Public Opinion
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
15
These results leave us with a highly interesting discussion. While Hypothesis 1 concerning public
values in single-party countries is supported, Hypothesis 2 concerning public values in multi-party
countries is rejected. One might say that the vertical orientation in a single-party country
(Hypothesis 1) is not at the expense of the other values we believed to be present in a democracy
(Hypothesis 2). Other factors may be present.
Does Culture Matter?
Hypothesis 3 expected civil servants in a country belonging to the Confucian cultural group (China,
Taiwan) to attach more importance to political loyalty, loyalty to rules, continuity, and adaptability
than civil servants from other countries.
Figure 9: The Importance of Political Loyalty
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Figure 9 shows that China has the highest level of importance attached to political loyalty followed
by Taiwan and Denmark, with approximately the same level and the US with the lowest level. This
implies that although the respondents from the Confucian cultural group on average have higher
scores (as also illustrated in the significant dummy variable in Table B3 in Appendix B), it does not
explain why Taiwan does not differ from Denmark, as otherwise expected.
Loyalty to rules was also expected to be seen as more important for respondents from the Confucian
cultural group, and if we combine the respondents from China and Taiwan in a category in the
16
ordinal regression, there is a statistically significant (positive) difference from the other
respondents. As Figure 10 also illustrates, however, the differences are inconsequential, and China
is the only country where the difference is statistically significant in the post-hoc tests.
Figure 10: The Importance of Rule Loyalty
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
As Figure 11 illustrates, Chinese respondents attach greater importance to continuity than the other
respondents followed by respondents from Taiwan. Table B3 confirms that respondents from the
Confucian cultural group attach higher importance to continuity than other respondents. The
difference between Taiwanese and US respondents is, however, negligible.
Figure 11: The Importance of Continuity
Note: See the note to Figure 1.
17
The last value expected to be more important for respondents from countries belonging to the
Confucian cultural group is adaptability. Here, the results again defy expectations (Figure 12 and
Table B3). The US respondents attach the greatest importance to this value followed by the Danish
and Chinese respondents, with approximately the same average importance, while Taiwanese
respondents tend to see this value as less important.
Figure 12: The Importance of Adaptability
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Thus far, the findings indicate that culture does not matter, at least not in the expected manner. An
alternative conclusion could be that culture matters contingent upon regime type. The high Chinese
scores on both political loyalty and rule loyalty―resembling the verticality values from Hypothesis
1―suggest that the Confucian culture has been diluted in the Taiwanese case, while it remains
stronger in China (cf. also the high Chinese score on continuity).
Hypothesis 4 expects that civil servants in a country belonging to the mastery cultural group (USA)
attach more importance to innovation, the ability to innovate, business-like operations, willingness
to take risks, learning, and development on the job than respondents from other countries.
We have asked the respondents two questions about innovation: Both as a general value and about
the importance of the ability to innovate. Concerning innovation as a general value, Figure 13
18
shows that respondents from the US and Denmark attach the greatest importance to this value and
that US respondents, as expected, attach more importance to the value than Chinese and Taiwanese
respondents (and these differences are statistically significant in both the post-hoc tests and ordinal
regression, see Table B4 in Appendix B). There is no significant difference between Denmark and
the US. For the ability to innovate, we find the exact same pattern (the yellow bars in Figure 13).
This indicates an East–West divide.
Figure 13: The Importance of Innovation
Innovation as a general value
5
Ability to innovate
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Concerning business-like operations (Figure 14), Danish respondents score higher than the other
respondents followed by respondents from the US, then Taiwan, and finally China. Our expectation
is not met, given that US respondents score lower than Danish respondents and do not differ from
Taiwanese respondents (although they score higher than Chinese respondents, Table B4). Given
China’s low score score, the explanation may be differences in political and economic systems.
19
Figure 14: The Importance of a Business-Like Approach
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
Concerning the willingness to take risks (Figure 15), Chinese and Taiwanese respondents see this as
more important than both US and Danish respondents. Our expectation is therefore not met here
either (see also Table B4). Note that we find a clear Confucian versus non-Confucian divide but the
opposite of what one might have expected.
Figure 15: The Importance of Willingness to Take Risks
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
20
In contrast, our expectation is met for learning and development on the job, as the US respondents
score higher than the respondents from all of the other countries in Figure 16 (Table B4 shows that
these differences are statistically significant).
Figure 16: The Importance of On-the-Job Learning and Development
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
We did not find much support for Hypothesis 4. Only one value―the value of learning and
developing on the job―was considered more important in the US as compared to the three nonmastery countries. On the other hand, it should be noted that some of the results point to the
importance of the Confucian–non-Confucian cultural split. Compared to Denmark and the US,
China and Taiwan score lower on innovation and business-like operations and higher on willingness
to take risks. The low scores can be interpreted as being in accordance with Confucianism, but we
cannot explain the high score on the willingness to take risks.
Hypothesis 5 expects the civil servants in a country belonging to the egalitarianism/intellectual
autonomy cultural group (Denmark) to attach more importance to user democracy, recognition from
peers, personal integrity, and independent professional standards than the civil servants from other
countries.
21
The expectation is not confirmed for the importance of strengthening user democracy. As illustrated
in Figure 17 (and tested in Table B5 in Appendix B), the difference between the countries is
inconsequential.
Figure 17: The Importance of Strengthening User Democracy
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
There is no substantial difference between Denmark, China, and Taiwan when examining
recognition from peers (Figure 18), while US respondents attach less importance to this value (see
also Table B5).
22
Figure 18: The Importance of Peer Recognition
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
For personal integrity, the country mean scores can be seen in Figure 19. Taiwanese respondents
score lower than Danish respondents, but US respondents score significantly higher (according to
ordinal regression, although the difference is substantially inconsequential and not statistically
significant in post-hoc testing).
Figure 19: The Importance of Personal Integrity
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
23
For independent professional standards, our expectations are met. According to Figure 20, Danish
respondents have higher scores than respondents from the other three countries (post-hoc testing
and Table B5 show that this difference is statistically significant).
Figure 20: The Importance of Independent Professional Standards
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Note: See the note to Figure 1.
In sum, the evidence concerning Hypothesis 5 is mixed. We found the expected distribution in the
case of independent professional values, but the other expectations were not met. On the other hand,
if we consider the values discussion in relation to Hypothesis 1 (see also Table B1), Denmark
consistently has the lowest score on all public values related to the vertical nature of a single-party
system. Together with the high score on independent professional standards, this fits nicely with the
Schwartz location of Denmark in the egalitarian/intellectual autonomy cultural group.
Discussion: Diversity or Homogeneity?
Differences Between Countries
Our first two hypotheses were derived from theoretical expectations concerning differences in the
value importance between respondents from different political systems. We expected respondents
from the single-party Chinese system to differ systematically from respondents in the three
countries with multi-party systems. The first hypothesis concerned expectations about the
importance of the internal values related to verticality: “ability to interpret the political climate and
signals,” “good relations with higher authorities,” “recognition from management,” “high salary,”
24
and “career opportunities.” The analysis confirmed that Chinese respondents attach higher
importance to these values.
Our second hypothesis expected the emphasis on internal values related to verticality to crowd out
the external values: the importance of satisfying users’ immediate needs, public insight and
transparency, and consideration of the public opinion. However, these expectations were not met.
Chinese respondents actually emphasized satisfying users’ needs more than other respondents.
In sum, regime type is clearly reflected in the importance of values related to internal verticality, but
regime type does not explain all of the normative orientations. Chinese respondents do not differ
significantly from their counterparts in multiparty systems with regard to external related values.
The results from our analysis of the hypotheses derived from culture theory were also quite mixed.
The first group of expectations (in Hypothesis 3) concerned political loyalty, loyalty to rules,
continuity, and adaptability. We argued that respondents from countries belonging to the Confucian
cultural group (Taiwan and China) were likely to attach higher importance to these values than their
counterparts in Denmark and the US. The overall picture is that these expectations were confirmed
for China but not for Taiwan. This mixed picture can be interpreted differently. First, it may be
argued that culture matters but that the impact depends on the political regime, as the values which
can be theoretically connected with Confucianism may have better conditions for thriving in some
settings (China) than in others (Taiwan). Second, Taiwan may have moved away from a classical
Confucian culture and towards a more hybrid cultural form due to stronger interaction with Western
countries than in China, as Schwarz and Bardi (2001) also find.
Hypothesis 4 stated that the respondents from a country belonging to the mastery cultural group
attach more importance to innovation, the ability to innovate, business-like operations, willingness
to take risks, and learning and development on the job. These expectations were not met, except for
learning and development on the job.
Hypothesis 5 expected that the respondents from a country belonging to the
egalitarianism/intellectual autonomy cultural group would attach more importance to user
democracy, recognition from peers, personal integrity, and independent professional standards.
25
These expectations were not confirmed, with the exception of the high importance attached to
independent professional standards by Danish respondents. However, if we combine this result with
the consistently low Danish scores on all public values related to the verticality, the emerging
picture fits nicely with the Schwartz location of Denmark in the egalitarian/intellectual autonomy
cultural group.
The overall conclusion is that we find mixed support for regime-based and culture-based
hypotheses. The most unambiguous result is the identified relationship between regime type and
values related to verticality, with China and Denmark as opposites. Next, the lack of clear support
for most hypotheses suggests that we should consider other explanations for the reported value
profiles in the four countries. First, our results indicate interaction between regime type and culture.
The Chinese regime may create a more benign environment for Confucianism to have an impact,
just as the low Danish scores on vertical-related values (regime) fit well the high score on
independent professional standards (culture). Second, we often found an unexpected and rather
clear East–West distribution. This finding suggests that the Schwartz cultural theory has
weaknesses, e.g. Denmark and USA have unexpectedly similar scores in many cases, and that a
more traditional East–West categorization seems more robust. Third, factors other than regime type
and culture are present. The very low Chinese score on business approach suggests that the type of
economic system should be included as a contextual factor.
Finally, one might imagine alternative routes of analyzing the identified normative orientations of
the civil servants in the four countries. One possibility is to highlight the specifics of the single
country. For example, USA has the highest scores on public insight and transparency, considering
the public opinion and accountability to society in general, and a remarkably low score on political
loyalty, suggesting an open public administration. China has a similar high score on accountability
to society in general, which, combined with the highest score on political loyalty, suggests a
completely different interpretation of a general accountability to society.
Another path of investigation―given the fact that all countries score very high on certain
values―is analyzing the nature of the similarities of the normative orientations of the civil servants
in the four countries. We thus turn to our nil-hypothesis.
26
Universal Public Values?
Public administrations in all countries share so many characteristics in common that they are likely
to share a set of universal public values or―putting it differently―a set of inherently governmental
values. One line of thought stresses that the public administration is owned and governed by
politicians. The main role of public administration anywhere is thus to be loyal towards the
legitimate masters with no regard to whether the master is elected in open competition between
several parties, selected by a single party, designated by tradition, or has inherited a kingdom by
birth. The public apparatus has a single core obligation: serving the master. If this argument is
valid, we should be able to identify a universal set of public values consisting of values facilitating
obedient service, such as political loyalty and good relations with higher authorities.
However, one might also argue that all public administrations share tasks such as rule
implementation, law enforcement, and service production in common. This implies that public
organizations follow rules and professional standards with little or no attention to the political
master. If this argument is valid, the core obligation is independent professional decision making,
and a set of universal public values takes another shape, e.g. stressing legal values integrity, and
independent professionalism, i.e. a corps d’esprit of an independent civil service.
How do we then identify universal values? One practical strategy is simply to look for values that
received average scores of importance above four in all countries. Several values included in our
hypotheses met this criterion. However, we must obviously include all of the values in our
questionnaire, not limiting ourselves to the values related to our five hypotheses on diversity.
15 values meet our criterion for universalism. Although the differences in score are modest (the
average scores for all respondents range from 4.65 for personal integrity to 4.14 for good social
working environment), it is notable how the seven values with the highest total scores fit nicely into
a single group, which most aptly can be labeled as corps d’esprit values: personal integrity, ethical
awareness, judicial values, accountability to society in general, professional commitment, equal
treatment, and independent professional standards. Interestingly, this value type is omnipresent in
codes of good government in 14 countries around the world (Beck Jørgensen & Sørensen, 2012-13)
and confirms this finding.
27
Six values can be categorized in a group of organizational values: commitment to organizational
task/mission, adaptability, productivity, learning and development on the job, loyalty to rules, and
good social working environment. Finally, two values concern external relations: public
insight/transparency and good relations to users.
Note that none of the values related to verticality survived, meaning that public values such as
interpreting the political climate and signals, good relations with higher authorities, recognition
from management, salary, career opportunities, and political loyalty cannot be included in a set of
universal public values. Their importance depends on the national setting. Apparently, the vision of
a public administration devoted to the obedient service of the political masters is not a dominant
part of civil servants’ normative orientation.
Note also that the “modern values” reflecting New Public Management, networked governance, and
participation do not meet our criterion of universality. Although productivity is included, it cannot
be taken for granted that this value reflects NPM. A businesslike approach as a value does not pass
our test and a content analysis of economy related values (productivity, effectiveness, parsimony) in
codes of good governance reveals that the economy is related to the proper use of public money and
avoiding corruption rather than running public organizations as efficient business firms (Beck
Jørgensen & Sørensen, 2012-13) .
All in all, we get a picture of an understanding of the public machinery as self-propelled and selfdisciplined. This is not to say that the normative orientation of our civil servants is exclusively
introvert. The universal values include accountability to society in general, public insight and
transparency, and good relations to users but characteristically exclude listening to the public
opinion, satisfying immediate users’ needs, and strengthening user democracy. Civil servants expect
themselves to serve others but not to please others: They are superior to others in order to serve
others.
Conclusion
Contrary to our initial expectations, the main conclusion seems to be that despite important
differences between China, Taiwan, USA, and Denmark, civil servants appear to subscribe to a
28
common set of universal public values―a public service ethos―centered around now-classic
public values such as personal integrity, accountability to society in general, and professional
commitment. These values are situated in different contexts in terms of the type of political system
and culture, and the enactment and interpretation of values may therefore assume different routes.
Naturally, this study is marked by several weaknesses. Four of our five hypotheses―especially the
culture-based hypotheses―received mixed support, suggesting that the Schwartz classification had
weaknesses or that culture is too broad and imprecise a concept. However, one should recall that we
found several rather clear East–West differences.
Next, survey data have weaknesses. While there are often only small differences in score, the
abstract values in the questionnaire can be interpreted quite differently in the four countries.
Further, we only know civil servants’ assessment of which public values they thought should be
important. We have no other data. Cutting to the bone, we might only know something about civil
servants’ sweet but unrealistic dreams.
Future research should thus incorporate other data and methods such as written material, case
studies, observation of behavior, and historical studies as also suggested by Toon Kerkhof in this
symposium. Finally, it would be interesting to conduct studies in countries from other continents,
such as Africa and Latin America.
29
References
Andersen, L. B., Beck Jørgensen, T., Kjeldsen, A. M., Pedersen, L. H., & Vrangbæk, K. (2012a).
Public value dimensions: Developing and testing a multidimensional classification.
International Journal of Public Administration, 35, 715-728.
Andersen, L. B., Beck Jørgensen, T., Kjeldsen, A. M., Pedersen, L. H., & Vrangbæk, K. (2012b).
Public service motivation and public values: Conceptual and empirical relationships.
Published online in American Review of Public Administration, 28 March 2012.
Antonsen, M., & Beck Jørgensen, T. (1997). The ‘publicness’ of public organizations. Public
Administration, 75, 337-358.
Beck Jørgensen, T. (2003). Værdier og den offentlige sektor. In T. Beck Jørgensen (Ed.), På sporet
af en offentlig identitet (pp. 11 38). Aarhus: Aarhus University Press.
Beck Jørgensen, T., & Sørensen, D.-L. (2012-13). Codes of good governance: National or global
public values? Public Integrity. Scheduled for publication in 15 (1).
Bozeman, B. (1987). All organizations are public: Bridging public and private organizational
theories. San Francisco: Jossey-Bass.
Bozeman, B. (2006). Public value mapping: An outline of a normative method. In H. Foss Hansen
(Ed.), Den organiserede forvaltning. Politik, viden og værdier i samspil (pp. xxx-xxx).
Copenhagen: Forlaget Politiske Studier.
Fitzpatrick, J., Goggin, M., Heikkila, T., Klingner, D., Machado, J., & Martell, C. (2011). A new
look at comparative public administration: Trends in research and an agenda for the future.
Public Administration Review, 71, 821-830.
Garofalo, C. (20XX). American governance and outsourcing: Public values, prisons, and national
security. Paper presented at the Conference on Governing Good and Governing Well, Vrije
Universiteit, Amsterdam, May 2009.
Hofstede, G., & Bond, M. H. (1984). Hofstede’s culture dimensions: An independent validation
using rokeach’s value survey. Journal of Cross-Cultural Psychology, 15, 417-433.
Hofstede, G. (2001). Culture’s consequences. Comparing values, behaviors, institutions and
organizations across nations, second edition. Thousand Oaks, London & New Delhi: Sage
Publications.
Liu, Y., & Cohen, A. (2010). Values, commitment, and OCB among Chinese employees.
International Journal of Intercultural Relations, 34, 493-506.
30
Meyer, P. J., & Allen, J. N. (1984). Testing the ‘side-bet theory’ of organizational commitment:
Some methodological considerations. Journal of Applied Psychology, 69, 372-378.
Overeem, P. (2008). Regime change for regime values? Applying the constitutionalist approach to
public administration outside America. Copenhagen Research Workshop on Public Values and
Public Interest.
Palidauskaite, J. (2011). Value profile of civil servants in new European democracies through the
lens of embedded ethics. In M. S. De Vries & P. S. Kim (Eds), Value and virtue in public
administration: A comparative perspective (pp. xxx-xxx). Houndmills, Basingstoke, Palgrave
Macmillan.
Rohr, J. (1978). Ethics for bureaucrats: An essay on law and values. New York: Marcel Dekker.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theory and empirical tests
in 20 countries. In M. Zanna (Ed.), Advances in experimental social psychology (pp. xxx-xxx).
New York: Academic Press.
Schwartz, S. H. (1994). Are there universal aspects in the content and structure of values? Journal
of Social Issues, 50, 19-45.
Schwartz, S. H. (1999). Cultural value differences: Some implications for work. Applied
Psychology: An International Review, 48, 23-47.
Schwartz, S. H., & Bardi, A. (1997). Influences of adaptation to communist rule on value priorities
in Eastern Europe. Political Psychology, 18, 385-410.
Schwartz, S. H. & Bardi, A. (2001). Value hierarchies across cultures: Taking a similarities
perspective. Journal of Cross-Cultural Psychology, 32, 268-290.
Soto, W. D., Opheim, C., & Tajalli, H. T. (1999). Apples and oranges? Comparing the attitudes of
public policy versus public administration students. The American Review of Public
Administration, 29, 77-91.
Svara, J. H. (2010). The next generation challenge: Finding and incorporating the local (and other)
government managers of the future. Journal of Public Affairs Education, 16, 361-377.
Van der Waal, Z., Pekur, A., & Vrangbæk, K. (2008). Public sector value congruence among old
and new EU member-states? Empirical evidence from the Netherlands, Denmark and Estonia.
Public Integrity, 10, 317-333.
Vrangbæk, K. (2009). Public sector values in Denmark: A survey analysis. International Journal of
Public Administration, 32, 508-535.
31
Yang, L., & van der Wal, Z. (2011). Rule of morality vs. rule of law: Establishing a set of values to
compare the civil service ethos in China and the Netherlands. Paper presented at the Symposium
on Ethical Leadership: Issues, Challenges and Opportunities 22-24 October, Chengdu, P. R.
China.
32
Appendix A: Questionnaire Concerning Values, Motives and Competences in
the Public Sector
1. How important do you think these values should be in your organization?
Please mark one option in each row
Should
Should be
Should
occasionfundamental; normally be nally be
very important important
important
Shouldn’t
be
important
Shouldn’t
play any
role
Don’t
know
A. Political loyalty
B. Judicial values/due process
C. Independent professional standards
(freedom to use professional discretion
in a qualified manner)
D. Balance societal interests (e.g.,
considering special interests without
letting them dominate)
E. Satisfying immediate needs of users
F. High productivity
G. Businesslike operations
H. Accountability to society in general
I. Public insight and transparency
(citizens and others can always ask
questions and look over our shoulder)
J. Listen to the public opinion (ensure
alignment between public opinion and
our activity)
K. Equal treatment (avoid discrimination
based on gender, ethnicity, and religion
in encounters with users or recruitment)
L. Continuity (consider the long term
and maintain traditions)
M. Innovation
N. Ensure good career opportunities for
employees
O. Strengthen user democracy (e.g., help
users gain influence using existing
channels or creating new ones)
P. Networking (moving beyond sector
limits and traditional jurisdictions)
Q. Other values, please specify:
33
2. In your opinion, how should core employees in a public organization rate the importance of
the following competences?
Please mark one option in each row
Should be
fundamental;
very important
Should
Should
normally be occasionally
important
be important
Shouldn’t Shouldn’t
be
play any
important role
Don’t
know
A. Ability to interpret the political
climate and signals
B. Ethical awareness
C. Willingness to take risks
D. Loyalty to rules
E. Adaptability
F. Personal integrity
(e.g., honesty and credibility)
G. Professional drive
H. Economic awareness
J. Strong interpersonal skills
I. Ability to innovate
K. Other competences, please specify:
3. In your opinion, how should core employees in a public organization rate the importance of
the following motives?
Please mark one option in each row
Should be
extremely
important
Should be Should be
very
fairly
important important
Should be Shouldn’t
less
be
important important
Don’t
know
A. Professional commitment
B. Commitment to the organizational
task/mission
C. Good social work environment
D. A high salary
E. Career opportunities
F. Staying within budget
G. Good relations with higher authorities
H. Learning and development on the job
I. Recognition from management
J. Good relations to users
K. Recognition from peers
L. Other types of motives, please specify:
34
Appendix B: Ordinal Regressions
To make sure that the results are robust when using a technique appropriate for ordinal scale, this
appendix explains and shows ordinal regressions corresponding to the hypotheses in the text. The
ordinal regressions also control for the fact that the gender composition and average age differ
between the countries.
When comparing one country to the other three countries, this country becomes the reference
category, and the country coefficients express estimated differences from the respondents from this
country. When comparing groups of countries (e.g., the two countries belonging to the Confucian
cultural group with the other two countries), we construct a dummy variable for whether a given
respondent comes from the group in question. If there is only one country in a group (e.g., the US in
the mastery cultural group), we use the country dummy variable instead of coding countries
together. We use a logit transformation in all models, and all models are acceptable in term of
significant tests of model fitting information (-2 log likelihood).
The ordinal regressions (Table B1-B5-6) can be interpreted as follows: The estimates labeled
“Location” are the coefficients for the predictor variables, and the country dummy variables are of
primary interest for the hypotheses. Compared to Chinese respondents (the reference category),
Danish respondents have 2.773 lower-ordered log odds of finding “Importance of ability to interpret
the political climate and signals” one category more important. The ratio of the odds for lower to
higher value importance for Danish and Chinese respondents are calculated by e–β, in this case exp
(2.733) = 15.37. The estimates in the Table B1-B5 are easily interpreted, given that negative values
mean that respondents from the country in question attach lower importance to the value in question
than respondents from the reference country. Reference countries are chosen to be the country
which the hypothesis expects to differ from the other. In Table 3, China and Taiwan are combined
to allow for a comparison with the countries which do not belong to the Confucian cultural group.
The interpretation of the control variables follows the same procedure. In Table 2, age is not
significantly related to “Importance of ability to interpret the political climate and signals” but to
“Importance of the organization ensuring career opportunities,” the parameter estimate for age (–
0.019) meaning that for a one-year increase in age, we expect a 0.014 decrease in the ordered log
odds of the finding, given all of the other variables in the model are held constant. Gender has only
two values (0 and 1), and the parameter estimate thus expresses the difference in expected ordered
35
log odds between women and men (reference category). The estimates labeled threshold are the
intercept equivalent terms.
Table B1: Ordinal Regression of Values Investigated in Hypothesis 1
Importance of
Importance of
Importance of Importance Importance of
Importance
ability to
good relations
recognition
of high
the
of career
interpret the
with higher
from
salary as a
organization
opportunities
political climate authorities as a
management
motive
ensuring career as a motive
and signals
as a motive
motive
opportunities
Threshold
Should not play any role
0.257
–2.687 ***
–5.512 *** –5.561 ***
–5.563 ***
–6.918 ***
Should not be important
1.309 ***
–0.371
–3.577 *** –2.853 ***
–3.815 ***
–4.235 ***
Occasionally be important
2.949 ***
1.662 ***
–1.361 *** –0.759 *
–1.651 ***
–1.692 ***
Should normally be important
4.804 ***
3.898 ***
0.534
1.588 ***
0.653 *
0.336
Fundamental, very important
Reference
Reference
Reference
Reference
Reference
Reference
Age
–0.014
–0.004
0.010
0.016
–0.019 *
0.001
Female = 1
0.011
0.429 ***
0.392 ***
0.060
0.398 ***
0.256 *
Taiwan = 1
–1.708 ***
–0.747 ***
–0.632 *** –0.412 *
–0.211
–0.733 ***
US = 1
–1.468 ***
–0.451 ***
–0.981 *** –1.791***
0.177
–0.942 ***
Denmark = 1
–2.773 ***
–2.648 ***
–0.698 *
–2.380 ***
–0.626 *
–2.331***
Nagelkerke
0.223
0.208
0.045
0.314
0.021
0.223
N
1174
1173
1184
1182
1178
1182
Location
Note: Logit link function. Reference category is China and the country coefficients express estimated difference from China. * p <
0.05; ** p < 0.01, *** p < 0.001.
36
Table B2: Ordinal Regression of Values Investigated in Hypothesis 2
Importance of satisfying
Importance of public
immediate user needs as a insight and transparency
Importance of listening to
the public opinion
general value
Threshold
Should not play any role
–6.041 ***
–7.170 ***
–4.326 ***
Should not be important
–4.044 ***
–3.852 ***
–3.072 ***
Occasionally be important
–1.882 **
–1.759 ***
–1.277 ***
Should normally be important
0.283 ***
0.190
0.740 *
Fundamental, very important
Reference
Reference
Reference
Age
–0.012
–0.006
–0.007
Female =1
0.412 ***
0.215
0.081
Taiwan = 1
–0.143 ***
–0.206
0.028
US = 1
–0,920
0.651 *
0.498 ***
Denmark = 1
–1.457***
–0.338
–1.848 ***
Nagelkerke
0.081
0.034
0.204
N
1180
1185
1168
Location
Note: Logit link function. Reference category is China and the country coefficients express estimated difference from China. * p <
0.05; ** p < 0.01, *** p < 0.001.
37
Table B3: Ordinal Regression of Values Investigated in Hypothesis 3
Importance of
political loyalty
Importance of
loyalty to rules
Importance of
continuity
Importance of being
adaptable
–2.998 ***
–6.878 ***
–3.326 ***
No observations
Should not be important
–1.849 ***
–3.680 ***
–1.957 ***
–4.352 ***
Occasionally be important
–1.014 ***
–1.587 ***
–0.217
–2.211 ***
Should normally be important
0.481***
0.764 ***
1.620***
0.322 **
Fundamental. very important
Reference
Reference
Reference
Reference
Age
–0.052***
0.001
0.014**
0.009
Female = 1
–0.159
0.133
–0.119
0.137
Confucian group = 1
2.427***
0.290 *
0.627***
–0.595 ***
Nagelkerke
0.223
0.008
0.068
0.019
N
1174
1182
1176
1176
Threshold
Should not play any role
Location
Note: Logit link function. Non-Confucianist culture is reference (respondents from Denmark and the US). * p < 0.05; ** p < 0.01,
*** p < 0.001.
38
Table B4: Ordinal Regression of Values Investigated in Hypothesis 4
Importance of Importance
Importance of
Importance of
Importance of
innovation as
of ability to
businesslike
willingness to
learning and
a general
innovate
operations as a
take risks
development on the
value
general value
job as a motive
Threshold
Should not play any role
–6.118 ***
–6.239 ***
–3.583 ***
–3.225 ***
–7.213 ***
Should not be important
–4.307 ***
–3.863 ***
–1.748 ***
–2.189 ***
–5.122 ***
Occasionally be important
–1.314 ***
–1.103 **
0.154
0.168
–2.563 ***
Should normally be important
0.746 *
0.858 *
2.155 ***
2.121 ***
–0.163
Fundamental. very important
Reference
Reference
Reference
Reference
Reference
Age
0.008
0.009
0.015
0.007
–0.004
Female = 1
0.234 *
0.268 *
0.286 **
0.085
0.404 ***
Taiwan = 1
–1.025 ***
–0.631 ***
–0.214
1.456 ***
–1.104 ***
Denmark = 1
0.571
0.454
1.065 ***
–0.228
–1.397 ***
China = 1
–1.139 ***
–0.732 ***
–1.512 ***
1.601 ***
–0.482 **
Nagelkerke
0.127
0.058
0.187
0.207
0.081
N
1172
1177
1167
1177
1187
Location
Note: Logit link function. Reference category is the US, and the country coefficients express estimated difference from US
respondents. * p < 0.05; ** p < 0.01, *** p < 0.001.
39
Table B5: Ordinal Regression of Values Investigated in Hypothesis 5
Importance of strengthening
Importance of
Importance of
Importance
user democracy as a general
recognition from
having personal independent
value
peers
integrity
Should not play any role
–4.165 ***
–4.736 ***
No observations –7.887 ***
Should not be important
–2.792 ***
–2.599 ***
–6.104 ***
–5.309 ***
Occasionally be important
–0.673 ***
–0.510 ***
–3.273 ***
–2.867 ***
Should normally be important
1.269 ***
1.591 ***
–1.227 ***
–0.571 ***
Fundamental. Very important
Reference
Reference
Reference
Reference
Age
–0.011
–0.013
–0.001
–0.010
Female = 1
0.367 **
0.344 **
0.320 *
0.154
Taiwan = 1
0.402
–0.172
–1.666 ***
–0.748 *
US = 1
0.554
–0.776 **
0.826 *
–0.922 **
China = 1
0.179
0.160
–0.423
–0.945 **
Nagerkerke
0.017
0.040
0.153
0.096
N
1150
1179
1186
1175
prof. standards
Threshold
Location
Note: Logit link function. Reference category is Denmark, and the country coefficients express estimated difference from Danish
respondents. * p < 0.05; ** p < 0.01, *** p < 0.001.
40