Preliminary results. Please do not quote
A comparative study of public values 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 prepared for The Public Value Consortium Biennial Workshop in Chicago 2012
Introduction
1
A basic but rarely addressed question in public values research is whether and how public sector
values vary from one country to another. We do know that public values vary according to task,
sector and administrative level within one country (Vrangbæk, 2009) but comparative studies of
public values are rare and limited to countries that are quite alike, cf. Van der Waal, Pekur and
Vrangbæk (2008) and Palidauskaite (2010). This paper seeks to fill this gap in the literature as it
compares public values in four countries – China, Taiwan, USA and Denmark – which differ
widely on a number of potentially important dimensions such as size, age, constitution and culture.
Our basic interest is to investigate how public values depend on political system and national
culture. One would clearly expect to find value differences between democracies and dictatorships
and between Western and Eastern cultures or between European continental and anglosaxon
cultures. In this respect the selected countries present us with large differences. The biggest
constitutional difference is between China which is a single-party state governed by the Communist
Party of China in contrast to the other three countries which are constitutional democracies. As to
cultural variation, several studies point to large differences (Hofstede; Schwartz, World Value
Survey). The possible interplay between political system and culture is also of interest as national
cultures may fit more or less well with specific political systems.
On the other hand, one may also argue that is important to look for a set of universal public values
because the core task of public administration – serving a master in ruling a country – simply is the
same everywhere, regardless of how the political master is chosen and regardless of culture
(references). Following this line of thought it is important looking for both similarities and
differences and explaining identified patterns.
Empirical studies of values are difficult because a value is such an airy and intangible phenomenon.
When it comes to comparative studies of public values in countries with different administrative
traditions and different political history the problems are immense as misunderstadings are beyond
absolute control. ”Oxygen” is the same in Denmark and China – an ”agency” or ”directorate” is
probably not, and certainly the understanding of political loyalty may depend heavily on both the
past and actual circumstances – in contrast to oxygen. This is precisely why comparative studies are
difficult - and why they are highly important (Beck Jørgensen, 2010).
2
Comparing public values in countries all over the world is costly if qualitative methods are used.
For this first preliminary study we have chosen surveys. Our study builds on survey data on public
values collected from students enrolled in a Master of Public Administration or Governance
program in the four countries. These students represent a convenient and informative group of
respondents for this type of comparative study. They have working experience from public
administration in their home countries and they have completed some degree of higher education,
which makes 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 from them as a part of
their education and this is likely to have a positive impact on response rates. The weakness is a
potential selection bias, e.g. being more committed to public sector values than other public sector
employees. The questionnaire used in this study has been applied in a number of earlier
investigations, reported in papers on Danish public values (also compared to Dutch and Estonian
values), on testing the dimensions of public values, and on empirical and conceptual links between
public values and public service motivation (references). Information on the data set is enclosed.
In the next section we will develop hypotheses addressing the possible links between constitution
and culture on the one hand and public values at the other hand. We consider constitution and
culture as independent variables and public values as the dependent variable but we will discuss
whether this is always the case. The following sections present our data and methods, the data
analysis and discussion, and finally our conclusion.
How public values may be framed by contextual factors – development of hypotheses
We define public values 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). The public
values literature insists that the desirable, not only the desired, is important and public values thus
provide direction to public employees rather than drive action (Andersen et al. 2012). Public values
are often considered important, because they are expected to a) form our perceptions of reality; b)
give identity to individuals as well as organizations; c) guide behavior; d) give meaning to public
service; and e) be necessary for the maintenance of communities and societies (cf. Kluckhohn,
1962; Lawton and Rose, 1994; Maguire, 1998; Verdikommisjonen, 1999).
According to Bozeman (2007: 13) public values specify
3
•
‘the rights, benefits and prerogatives to which citizens should (and should not) be entitled;
•
the obligations of citizens to society, the state and one another; and
•
the principles on which governments and policies should be based.’
The latter – systemic or institutional – element in the definition is important, as it indicates that
values can be considered as the basic building blocks of the public sector (Beck Jørgensen 1999:
581). We delimit ourselves to the systemic/institutional understanding of public values. We further
narrow it down by excluding the policy-part of Bozeman’s definition because a broader
understanding of public values (as the dependent variable) may overlap with political culture and
possibly also with national culture, one of our independent variables. Also, and not least important,
our questionnaire is designed to capture public values in the narrow sense.
Universal governmental regime values?
One may argue that there is one and only one contextual factor influencing public values and that is
ownership. Public organizations are owned and governed by politicians and the main role of all
public administration is 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. Thus, the public apparatus has one core obligation,
i.e. serving the master. If this argument is valid we should be able to identify a universal set of
public values. Or to put it differently: Inherently governmental regime values can be identified
(references). Suggestions could be political loyalty, adaptability and rule abidance as these values
may facilitate obedient service.
For a number of reasons, we do not choose identifying inherently governmental values as our point
of departure. First, in general the degree of political authority (and economic authority) may vary
along one or several dimensions. Following that argument, absolute political authority is a special
case, and often no clear boundary exists between public and private organizations. There are
degrees of publicness and some organizations are thus more public than others (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 because of contextual factors. We will therefore proceed
to discuss the contextual factors already mentioned: political system and culture.
4
Political system
Political systems can be categorized in numerous ways. One may differentiate between unitary and
federal states, decentralized and centralized states, corporatist regimes versus parliamentary regimes
etc. For our test we have chosen the strongest contrast between political systems, i.e. democracy
versus dictatorship, in this paper operationalized as multi-party systems versus single-party 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, or laws or practices prevent the
opposition from legally getting power. In contrast, in multi-party states at least two independent
parties are, both formally and in practice, allowed to run candidates for election and ultimately for
office.
The question is then which public values we theoretically associate with a single-party respective
multi-party regime. A single-party regime calls for a monolithic top-down world with a strong
emphasis on hierarchy, loyalty towards masters, flair for interpreting politics, and vertical climbing
within the system. Thus, our first hypothesis includes public values stressing the vertical nature of a
system (“verticality”). Next, we expect that values related to task and relations between
administration and the public play a less important role. Consequently we expect values such as
legality and professionalism to be less important. Finally, a single-party regime calls for limited
public competition over power and a restrained public discourse. Therefore we do not expect
transparency and listening to the public opinion to be important public values.
Referring to our questionnaire (see appendix A) we formulate the following two hypotheses:
H1: A country with a single-party system attach more importance to ability to interpret the political
climate and signals, good relations with higher authorities, recognition from management, a high
salary and career opportunities compared to master students from multi-party countries.
H2: A country with multiple parties attach more importance to judicial values/due process,
independent professional standards, public insight and transparency and consideration of the public
opinion compared to master students from a single-party system.
5
Culture
The nature of a political system is far from the only potential reason why certain public values
would be prioritized. Most importantly, countries can be argued to be dominated by a national
culture. The main argument is that a political system is embedded in a national culture that
influences which public values are the important ones. A common sense understanding would lead
us to think that West (USA, Denmark) contrasts East (China, Taiwan) but as will be shown a more
nuanced and theoretical grounded point of departure than a common sense East West comparison is
available.
Several theories locate high numbers of countries on cultural maps. Hofstede and associates have
developed cultural dimensions such as individualism-collectivism, masculinity-feminism, high-low
power distance, and high-low uncertainty avoidance (Hofstede, 2001; Hofstede & Bond 1984).
United States and Denmark have individualistic cultures, and China and Taiwan have collectivistic
cultures. Here we find an East-West divide. On the other hand, China and United States have
masculine cultures, while Denmark have 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 locates countries on two dimensions: traditional values vs.
secular/rational values and survival values vs. self-expression values. Both East and West have high
scores on secular/rational values while there is a clear difference between East and West on the
survival/self-expression dimension. The different scores on this dimension may however be related
to wealth.
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. The seven values are: 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 into three different groups based on relative emphasis on the seven value dimensions. China
and Taiwan belongs to a “Confucian” cultural group with a strong emphasis on hierarchy. 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”.
6
The “Confucian”/hierarchical group (China/Taiwan) has the following characteristics: A
hierarchical, differential allocation of fixed roles and of resources is the legitimate, desirable way to
regulate 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).
USA on the other hand belongs to a group of “English speaking” countries where the dominant
value is “mastery”. Mastery is characterized by: Groups and individuals should master, control, and
change the social and natural environment through assertive action in order to further personal or
group interests. The cultural emphasis is on getting ahead through active self-assertion (ambition,
success, daring, competence).
Denmark is located in the group of Western European countries with a strong emphasis on
“egalitarianism” and “intellectual autonomy”. The cultural orientation towards egalitarianism is
characterized by: Individuals are portrayed as moral equals, who share basic interests and who are
socialized to transcend selfish interests, cooperate voluntarily with others, and show concern for
everyone's welfare (equality, social justice, freedom, responsibility, honesty). Intellectual autonomy
is characterized by: The person is an autonomous, bounded entity and finds meaning in his / her
own uniqueness, seeking to express own internal attributes (preferences, traits, feelings) and is
encouraged to do so. Intellectual Autonomy has a cultural emphasis on the desirability of
individuals independently pursuing their own ideas and intellectual directions (curiosity,
broadmindedness, creativity).
Schwartz’ cultural theory is chosen as our point of departure for three reasons. First, our four
countries are located in three categories and this distribution constitutes an interesting contrast to
the distribution on political system; second, it is a well argued theory, which has been confirmed
through empirical investigations; and, third, it is easier to relate our public values in the
questionnaire to the Schwartz theory.
Referring to our questionnaire (see appendix A) we formulate the following three hypotheses:
7
H3: A country belonging to the confucianist/hierarchical group attach more importance to political
loyalty, loyalty to rules, good relations with higher authorities, recognition from management and
adaptability compared to master students from other countries.
H4: A country belonging to the mastery group attach more importance to innovation, the ability to
innovate, business-like operations, willingness to take risks, learning and development on the job
compared to master students from other countries.
H5: A country belonging to the egalitarianism/intellectual autonomy group attach more importance
to judicial values/due process, independent professional standards, equal treatment, ethical
awareness and personal integrity compared to master students from other countries.
Note that H1 and H5 to some extent unfortunately are overlapping because of the vertical nature
both of the Chinese political system and the Confucianist culture.
Research design and methods
The hypotheses are investigated using surveys of Public Administration students in China, Taiwan,
USA and Denmark, and this section explains why we compare masters of public administration
students and why we have chosen the four investigated countries. It then describes the empirical
data and our operationalizations. Finally, we discuss and explain our choice of methods to analyze
the data.
Analyzing Masters of Public Administration students
We investigate master students for the following four reasons which will be discussed further
below: Master of Public Administration students in the four countries are relatively comparable
even in these very different contexts, they have practical experience in the public sector, they are
aware of public values and public service motivation, and it is practically possible to gather survey
data from them as a part of their education.
We argue that Master of Public Administration students in the four countries are relatively
comparable even in these very different contexts. The investigated educational programs differ
8
slightly in their relative weight of the 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 because of the known difference 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 and Bond, 1984).
Many of the master students have practical experience in the public sector, and their values thus
reflect the different values in the four countries’ public sectors. Svara (2010) found that Public
Administration students are oriented toward contributing to society. He based his investigation on a
survey of Masters of Public Administration students at five universities, and more than 80 percent
agreed that meaningful public service and opportunities to help others are important them (ibid.).
Svara (2010: 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 students as a part of their education. As
discussed below, the response rate was high, and we were able to get data from a relative high
number of respondents.
The weakness is a potential selection bias as our respondents have actively selected to pursue public
management training and careers and/or receive support from their workplace. This can imply that
they are more committed to public sector values than other public sector employees. Furthermore it
may 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 the results.
Comparing mainland China, Denmark, United States, and Taiwan
The biggest difference in political system is between Mainland China which is a single-party state
(where only the Communist Party and a limited number of approved minor parties under “the
9
United Front” is permitted to run candidates for election) and the other three countries which are
multi-party systems. In terms of culture, China and Taiwan belong to the “Confucian”group, USA
belongs to the “mastery” group, and Denmark is part of the group of European countries with two
core values of “egalitarianism” and “intellectual autonomy”. Differences in culture and political
systems are not, of course, the only differences between the countries, but we argue that they
capture the most relevant distinctions. Clearly, one could also look to size and age. Two countries
are small (Denmark and Taiwan) and two countries are big (China and the US) and China as a
nation is the far oldest of the four countries.
Table 1 sums up the differences between China, Taiwan, the Unites States and Denmark in culture,
number of inhabitants and political systems.
Table 1: Investigated countries
Country Inhabitants
China
Taiwan
United
States
Denmark
Political system
Cultural group
Number of
respondents
1,341m
Single-party state
Confucian/hierarchical
307
23m
Multi-party state
Confucian/hierarchical
223
310m
Multi-party state
Mastery
265
6m
Multi-party state
Egalitarianism/intellectual
autonomy
455
Data and operationalizations
In Denmark, the questionnaires were sent out by post as part of their education. 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 present students in
the other three countries completed the questionnaires. The number of respondents in each of the
10
countries can be seen in Table 1 above. The translations were done by native speaking researchers
and language specialists (the English version of the questionnaire is attached).
In order to measure how much importance the students attached to different values, the respondents
were asked to evaluate the importance of firstly general organizational values in pursuing the daily
tasks of their organization, secondly values related to the desirability of certain staff competences
and thirdly 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 staying within the budget (values related to staff
motives). The general values are linked to performance of daily organizational tasks in the text and
the respondents are thus encouraged to consider the desirability of abstract values in light of
specific working tasks. The options for answering were for general values and staff competences:
Should be fundamental: very important (5); should normally be important (4); should occasionally
be important (3); shouldn’t be prominent (2); shouldn’t play any role (1) and don’t know (coded as
missing). The options for answering for the questions about motives were: extremely important (5),
very important (4), fairly important (3), less important (2), not important (1) and don’t know
(coded as missing). The answering options for staff motivation are thus more neutral in tone, while
the questions on general values and staff competencies more clearly aims to elicit respondents’
perspectives on desirability.
Note that all questions address what should be, not what is. Our data thus reflect 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 to
test whether their answers differed between the countries. The mean score on a given value in each
of the countries gives a parsimonious summary of the students’ 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 need to be done very carefully. Still, we can analyze what answers
the respondents gave and then discuss whether should be seen as an actual difference in their
11
values. This is also the main reason for primarily presenting descriptive analyses (graphs showing
the mean scores for each of the four countries) and only secondary using more advanced statistics.
The second challenge is, however, that value importance is measured on ordinal scales (see the
options for answering above), and this suggests that advanced statistics are necessary as a
supplement to the descriptive graphs. The available measure of central tendency for ordinal
variables (the median) does not capture 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 (shown in the appendix) to make sure that our results are robust when we use a proper
ordinal scale technique. When we compare one country to the other three countries, this country is
reference category, and the country coefficients express estimated difference from the students from
this country. If we compare groups of countries (e.g. the two countries belonging to the
Confucianist/hierarchical group with the two non-Confucianist countries) we construct a dummy
variable for whether a given student comes from the group in question (here: a Confucianist
country). If there is only one country in a group (e.g. United States in the mastery group) we use the
country dummy variable instead of coding countries together.
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 differences, if any, are not due to these personal characteristics. 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 fourth challenge relates to the fact that we measured 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 adjustment for the number of comparisons performed would mean that we could
get significant differences by chance, only because of the big number of comparisons. We have
therefore theoretically selected a more limited number of comparisons to present in the paper and
we also 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.
12
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.
Political system matters
According to hypothesis 1, the expectation is that master students from a country with a singleparty system 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 master students from multi-party countries. Given that China is the only
investigated single-party system, we test whether master students from China attach more
importance to the mentioned values compared to master students from Denmark, United States and
Taiwan.
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. The students from China attach higher
importance to this value than students from the other three countries. In a post hoc test (Scheffe,
including all items related to staff competences), these differences are all highly significant
(p<0.001). This is also the case if the differences are analyzed using ordinal regression (see Table
A1 in the appendix).
Figure 1: Comparison of importance of being able to interpret political climate and signals
Importance
5
Ability to interpret the political climate and signals
5=Very important
4
4= Should normally be important
3
3= Should occasionally be impo.
2
2= Shouldn’t be important
1
1= Shouldn’t play any role
0
Denmark
China
US
Taiwan
Note: Comparison of average importance reported by the participating Master of Public Administration students from
the four countries. The averages are calculated from the following classification: Very important (5); should normally
be important (4); should occasionally be important (3); shouldn’t be important (2); shouldn’t play any role (1).
13
The next value mentioned in hypothesis 1 concerns good relations with higher authorities, and the
students have been asked whether this is an important motive for core employees. Figure 2 shows
that although the average importance attached to this value is highest for China, the difference to
US and Taiwan is not very substantial. Still, according to the logistic regression in table A2 Chinese
students attach significantly higher importance to good relations with higher authorities as a motive
among core employees compared to students from all the other countries. Post hoc tests do,
however, indicate that only the difference from Danish and Taiwanese students is robust.
Figure 2: Comparison of good relations with higher authorities as a motive
5= Extremely important
4= Very important
3= Fairly impo.
2= Less important
1= not important
Note: See the note to Figure 1.
Recognition from management is also expected to be seen as a more important motive for core
employees for the Chinese students, and Figure 3 shows that this is actually the case. Post hoc tests
also show that Chinese students on average find this value significantly more important than
students from the three other countries, and the ordinal regression (table A3) confirms this.
14
Figure 3: Comparison of importance of recognition from management as a motive
5=Extremely important
4= Very important
3= fairly important
2= Less important
1= Not important
Note: See the note to Figure 1.
The next value expected to be more important for Chinese students is high salary as a motive for
core employees. Again, our expectations are met in Figure 4, and the logistic regression (table A4)
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: Comparison of importance of high salary as a motive
5=Extremely important
4= Very important
3= fairly important
2= Less important
1= Not important
Note: See the note to Figure 1.
Finally, career opportunities are expected to be seen as more important in China compared to the
other three countries. Figure 5 shows that Chinese students (as expected) attach more importance to
career opportunities as a staff motive than the other students (right hand, yellow bars), and this
difference is also statistically significant for all the countries (both in post hoc tests and in Table
A6). However, the students from the four countries do not differ in their importance attached to
15
whether the organization should ensure career opportunities (left hand, blue bars). The ordinal
regression (Table A5) indicates that the difference between Chinese and Danish students is
statistically significant after control for age and gender, but the substantial difference is
inconsequential.
Figure 5: Comparison of importance of career opportunities as general value and motive
5=Very important/extremely important (motive)
4= Should normally be important/very important
(motive)
3= Should occasionally be impo./fairly importan
(motive)
2= Shouldn’t be important/less important (motive
1= Shouldn’t play any role/not important
Note: See the note to Figure 1.
In sum, hypothesis 1 on the relation between a single party system and public values reflecting
verticality is confirmed.
Hypothesis 2 expects that a country with a multi-party system attach more importance to values
related to task and relations between system and environment: judicial values/due process,
independent professional standards, public insight and transparency and consideration of the public
opinion compared to master students from a single-party system. In this study, the implication is
that master students from Denmark, United States and Taiwan are expected to attach more
importance to the above mentioned values compared to master students from China.
Concerning the first value, judicial values/due process, our expectations are definitely not met. On
the contrary, Figure 6 shows that Chinese students attach substantially more importance this value
than the other students, and both logistic regression (Table A7) and post hoc tests confirm this. On
the other hand, it is worth mentioning that all countries attach very high importance to judicial
values.
16
Figure 6: Comparison of judicial values/due process as general value
5
Judicial values/due process
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Concerning the importance of independent professional standards, our results are not in line with
the expectations either. Figure 7 shows that Danish students attach more importance to this value
than all other master students, but although Chinese students attach slightly less importance to this
value than US students and Taiwanese students, these differences are not statistically significant in
either ordinal regression (Table A8) or post hoc tests.
Figure 7: Comparison of independent professional standards as general value
Independent professional standards
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Chinese master students were expected to attach less importance to public insight and transparency
than master students from Denmark, United States and Taiwan. This expectation is not met. There
is no difference between Denmark, China and Taiwan, while students from United States score
17
significantly higher than all the other students (see Figure 8 below, and Table A9 for the relevant
ordinal regression).
Figure 8: Comparison of public insight and transparency as general value
Public insight and transparency
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
The last part of hypothesis 2 was that master students from multi-party states were expected to think
that consideration of the public opinion was more important than students from a single-party state
(China). Again, this expectation is not met (see Figure 9). Similar to public insight and
transparence, master students from United States attach higher importance to considerations of the
public opinion than students from the other three countries, while Danish students attached
significantly lower importance to this value than the other students (Table A10).
Figure 9: Comparison of consideration of the public opinion as a general value
Consider the public opinion
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
18
These results leave us with a highly interesting discussion. While hypothesis 1 concerning public
values in single-party countries is confirmed, hypothesis 2 concerning public values in multi-party
countries is not met. One may say that the hierarchical/vertical orientation in a single-party country
(hypothesis 1) is not at the expense of other values we believed to be present in a democracy
(hypothesis 2). Other factors may be present. Also note that all countries attach very high
importance to all values in hypothesis 2 with one exception: considering the public opinion where
Denmark has an interestingly low score. Possibly, this leaves us with three candidates to a set of
universal public values.
Culture matters
Hypothesis 3 expected that a country belonging to the Confucianist group attach more importance
to political loyalty, loyalty to rules, good relations with higher authorities, recognition from
management and adaptability compared to master students from other countries. This implies that
master students from China and Taiwan attach more importance to these values compared to master
students from United States and Denmark.
Figure 10 shows that China has the highest level of importance attached to this value followed by
Taiwan and Denmark with approximately the same level and United States with the lowest level.
This implies that although the students from Confucianist countries on average have higher scores
(as also illustrated in the significant dummy variable in Table A11), it does not explain why Taiwan
does not, as expected, differ from Denmark.
19
Figure 10: Comparison of political loyalty as a general value
Political loyalty
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Loyalty to rules was also expected to be seen as more important in Confucianist countries, and it is
correct that if we combine the students from China and Taiwan in a category in the ordinal
regression, then they score significantly higher than the other students. But as Figure 11 also
illustrates, the differences are inconsequential, and China is the only country which differs
significantly from the other countries (all three of them) in the post hoc tests.
Figure 11: Comparison of importance of being loyal to rules
Loyalty to rules
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
We already discussed good relations with higher authorities as a motive in relation to hypothesis 1
(which expected higher scores for students from single-party systems), and although the ordinal
regression finds that students from Confucianist countries score higher on average, this is primarily
20
due to the high score in China. Students from United States and Taiwan do not differ (see also
figure 2 above), and our interpretation is therefore that the high score in China has more to do with
the political system than with the Confucianist culture.
The conclusion is basically similar for recognition from management as a motive. The significant
difference between Confucianist countries and other countries can be seen in Table A14, but Figure
3 above illustrates that the high score of Chinese students on this value may have more to do with
political system than with culture (given that the difference between US and Taiwan is neither
statistically nor substantially significant, while China and Taiwan differ significantly (both in post
hoc test and in Table A3).
The last value expected to be more important for student from Confucianist countries compared to
other countries is adaptability. Here the results show the opposite pattern (Table A15). Students
from the United States attach highest importance to this value followed by Danish and Chinese
students with approximately the same average importance, while Taiwanese students see this value
as significantly less important compared to students from the other three countries.
Figure 11: Comparison of importance of being adaptable
Adaptability
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
So far the conclusion is that culture does not matter, at least not in the expected way. Note that with
regard to political loyalty and loyalty towards rules we see the highest scores from Chinese
students, fitting well with the results from testing hypothesis 1. An alternative conclusion could thus
21
be, that culture matters, but that the confucianist culture has been diluted in the Taiwanese case,
while it remains stronger in China.
Hypothesis 4 expects that a country belonging to the mastery group attach more importance to
innovation, the ability to innovate, business-like operations, willingness to take risks, learning and
development on the job compared to master students from other countries. This implies that master
students from United States attach more importance to these values compared to master students
from Denmark, China and Taiwan.
We have asked the students 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 12 shows
that students from United States and Denmark attach most importance to this values and that
students from United States as expected attach significantly more importance to the value than
students from China and Taiwan (both in post hoc tests and ordinal regression, see Table A16).
There is no significant difference between Denmark and United States. For the ability to innovate,
we find the exact same pattern (see Figure 12 and Table A17). This points to an East-West divide.
Figure 12: Comparison of importance of innovation
Innovation
5
Innovation as a general value
Ability to innovate
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Concerning business-like operations (Figure 13), Danish students score significantly higher than the
other students followed by students from United States and then Taiwan and finally China. Our
expectation is not met, given that US students score significantly lower than Danish students and do
22
not differ from Taiwanese students (although they score higher than Chinese students, Table A18).
Given China’s solitary score, the explanation may differences in economic systems.
Figure 13: Comparison of importance of business-like operations as a general value
Business-like operations
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Concerning willingness to take risks (Figure 14), Chinese and Taiwanese students see this as more
important than both US and Danish students. Our expectation is therefore not met here either (see
also table A19).
Figure 14: Comparison of importance of willingness to take risks as a staff competence
Willingness to take risks
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
23
In contrast, our expectation is met for learning and development on the job where students from the
United States scores significantly higher than students from all other countries (Figure 15 and table
A20).
Figure 15: Comparison of importance of learning and development on the job as a staff motive
Learning and development on the job
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
We didn’t find much support to hypothesis 4. Only one value, the value of learning and developing
on the job, was considered more important in the US compared to the three non-mastery countries.
On the other hand, it should be noted that we identified results that point to the importance of the
Confucianist - non-Confucianist cultural split. Both China and Taiwan score lower than Denmark
and the US on innovation and business-like operations and they similarly score higher than
Denmark and the US on willingness to take risks. The low scores can be interpreted as in
accordance Confucianism but the high score on willingness to take risks demands a special
explanation which we don’t have.
Hypothesis 5 expects that a country belonging to the egalitarianism/intellectual autonomy group
attach more importance to judicial values/due process, independent professional standards, equal
treatment, ethical awareness and personal integrity compared to master students from other
countries. This implies that master students from Denmark are expected to attach more importance
to these values compared to master students from China, Taiwan and United States.
Concerning judicial values/due process, our expectations are not met, given that master students
from China have the highest score followed by students from Denmark and United State and finally
24
Taiwan (see Figure 6 above). Both post hoc tests and the ordinal regression (see table A7) show that
Danish students have lower scores than Chinese students and higher scores than Taiwan (there is no
significant difference between Denmark and United States).
For independent professional standards, our expectations are met. According to Figure 7 (above),
Danish students have the highest score, and they differ significantly from students from the other
three countries in post hoc test and ordinal regression (see also Table A8).
As can be seen from Figure 16, students from the United States attach higher importance to equal
treatment than the other countries, while there is no difference between students from Denmark,
Taiwan and China (Table A22).
Figure 16: Comparison of importance of equal treatment as a general value
Equal treatment
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
Taiwanese students do, as expected attach less importance to ethical awareness than Danish
students, but this is not the case for students from United States and China (according to Figure 17,
the ordinal regression (table A24) and post hoc tests). Denmark (as a country from the
egalitarianism/intellectual autonomy group) can therefore not be said to be characterized by a
higher level of importance of this value compared to other countries.
25
Figure 17: Comparison of importance of having ethical awareness as a staff motive
Having ethical awareness
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
The same conclusion can be drawn for personal integrity where the country mean scores can be
seen in Figure 18. Taiwanese students score lower than Danish students, but students from the
United States score significantly higher (according to ordinal regression, although the difference is
substantially inconsequential and not significant in post hoc test).
Figure 18: Comparison of importance of personal integrity as a staff competence
Personal integrity
5
Importance
4
3
2
1
0
Denmark
China
US
Taiwan
In sum, hypothesis 5 can not be confirmed. We found the expected distribution only in the case of
independent professional values. If we on the other hand take a look at tables 1, 2, 3, 4, and 5,
Denmark is rather consistently the country that shows the lowest score on all public values related
to the vertical nature of a single-party country. This fits nicely with the Schwartz location of
Denmark among egalitarian countries just as well as the high score on professional standards fits
well with Denmark’s location in the intellectual autonomy group. (Nice test – bad hypothesis!)
26
Discussion and conclusion
Our first two hypotheses were derived from a theoretical expectation of differences in values
between students of different political regimes. We expected students from the single-party Chinese
system to differ systematically from students in the other three countries with multi-party systems.
The first hypothesis concerned expectations about the importance of the values “ability to interpret
the political climate and signals”, “good relations with higher authorities” and the motivational
factors “recognition from management”, “high salary” and “career opportunities”. The analysis
confirmed our expectations that Chinese students attach higher importance to these values and
motivational factors above students from the other three countries.
Our second hypothesis related to political regimes included a set of expectations concerning judicial
values/due process, independent professional standards, public insight and transparency and
consideration of the public opinion. Chinese students were expected to attach lower importance to
these values than their counterparts in multiparty systems. However, these expectations were not
met. Chinese students actually emphasized judicial values/due process higher than other students,
and they do not differ significantly from the other three countries in regards to the other values in
this hypothesis.
What is our overall conclusion concerning political regime then? Firstly that it the regime type
seems to have some importance, as a fair number of expectations are met. But secondly that it does
not provide a perfect theory for understanding value differences. The theory may be wrong, or our
translation into survey questions could be imprecise. Considering the high importance attached to
judicial values/due process by Chinese students it seems fairly plausible that Chinese students
perceive the question differently from the others. We speculate that they do not consider individual
citizen rights and rule based administration, but rather refer to general anti-corruption and system
preservation values. Judicial values might for example refer more to “what you as a citizen or civil
servant should do for the system” than what the system should deliver in terms of upholding and
supporting the rights of the individual. Chinese students do not differ significantly from their
counterparts in multiparty systems in regards to independent professional standards, public insight
and transparency and consideration of the public opinion. This could indicate that even in
multiparty systems public administrators must consider external interests, although not necessarily
27
act according to them. Finding a way of coping with this is probably part of the professional
standards emphasized in all four countries. It thus points toward our null hypothesis of a common
set of value orientations among civil servants throughout the globe, given similarities in their core
job functions.
The results from our analysis of hypotheses derived from culture theory were also quite mixed. The
first group of expectations concerned political loyalty, loyalty to rules, good relations with higher
authorities, recognition from management and adaptability. We argued that Confucianist countries
(Taiwan and China) were likely to attach higher importance to these values than their counterparts
in Denmark and the United states. The overall picture is that these expectations were confirmed for
China, but not for Taiwan. This could lead to several different conclusions. Firstly we can speculate
that culture matters, but that the impact depends on the political regime, as partly confirmed in our
first two hypotheses. Confucianist values may have better conditions for thriving in some settings
(China) than in others (Taiwan). A second possible interpretation is that Taiwan has moved away
from a classical confucianist culture and towards a more hybrid cultural form due to stronger
interaction with western countries than in China. Schwarz indicates this in his overview studies of
cultural characteristics across countries, where China and Taiwan belong to the same family, but
with China as a cultural “outlier” and Taiwan closer to western countries (Schwarz and Bardi 2001).
The relatively lower importance placed on adaptability by “confucianist” students seems to falsify
our initial hypothesis, which was based on the premise that a high degree of willingness to accept
hierarchical decisions and accommodate system needs was an inherent part of this culture. Yet, our
formulation of the question is ambiguous and it is possible that the high score given by other
students reflects an entirely different understanding of adaptability, which is more in line with
innovation and entrepreneurship than adapting to system needs.
Hypothesis four stated that master students from a country belonging to the mastery group attach
more importance to innovation, the ability to innovate, business-like operations, willingness to take
risks, learning and development on the job. These expectations were not met, except for learning
and development on the job. Both Danish and US students place high importance on innovation. It
is possible that they attach slightly different meanings to this concept. Referring to hypothesis five it
could be that Danish students link innovation to the personal integrity and development, while US
students perhaps are more clearly linking innovation to “mastery” concepts. But we cannot confirm
28
these speculations. That Danish students attach high importance to “business like” operation may
reflect the general public administration rhetoric in the country over the past two decades, whereas
US students either take this for granted, or have grown weary of this after the early introduction of
“entrepreneurship” ideas in the 1980s.
The final hypothesis derived from culture theory was that master students from a country belonging
to the egalitarianism/intellectual autonomy group would attach more importance to judicial
values/due process, independent professional standards, equal treatment, ethical awareness and
personal integrity. These expectations were not confirmed, except for the high importance attached
to independent professional standards by Danish students. This result can be seen in light of the
consistently low scores on all public values related to the vertical nature of a single-party country in
table 1,2,3,4 and 5. The emerging picture when looking across these results fits nicely with the
Schwartz location of Denmark among egalitarian countries just as well as the high score on
professional standards fits well with Denmark’s location in the intellectual autonomy group.
The overall conclusion from our deductively based analysis must be that we find some support for
regime based hypothesis and some for culture based hypothesis. Furthermore, we speculate that
there might be interaction between the two types of explanations. In particular we suggested that
regimes may create more or less benign environments for cultural factors to have an impact.
A more inductive approach might reveal other patterns or possible interpretations, but we do not
have space to move into this in greater detail here, except for a brief consideration of our initial
discussion about the possibility of finding values that are shared across the public administrators in
all four countries. One way to operationalize this is to look for values that receive an average score
of importance above four in all countries. The following values meet this criterion: judicial
values/due process; independent professional standards; public insight and transparency; loyalty
towards rules; adaptability; learning and development on the job; equal treatment; ethical awareness
and personal integrity. Interestingly, these values are also found in Codes of Good Government in
14 countries around the world (Beck Jørgensen & Sørensen (forthcoming). Quite a few of these
values are related to a Weberian understanding of bureaucracy, and taken together they point to an
understanding of the public machinery as self-propelled and self-disciplined, not to say selfreferential. This indicates that future research needs to look at similarities as well as differences,
29
even when very different countries are compared.
30
References
Andersen, L.B., Beck Jørgensen, T., Kjeldsen, A.M., Pedersen, L.H., Vrangbæk, K. (in press):
"Public Service Motivation and Public Values: Conceptual and Empirical Relationships".
Accepted to be published in American Review of Public Administration January 9th. 2012.
Antonsen, M. & Beck Jørgensen, T. (1997): The ’publicness’ of public organizations. Public
Administration, vol. 75, no. 2, 337-358
Beck Jørgensen, T. 1999. The public sector in an in-between time: searching for new public values.
Public Administration, 77(3), 565–84.
Beck Jørgensen, T. (2003): Værdier og den offentlige sector. I: Beck Jørgensen, T. (ed): På sporet
af en offentlig identitet. Århus: Aarhus Universitetsforlag, 11-38.
Beck Jørgensen, T. (2010). ”Danish PA and PA in Danish”, In G. Bouckaert & W. van de Donk:
The Europaen Group of Public Administration (1975-2010). Perspectives for the Future.
Bruxelles: Bruylant, 223-227.
Beck Jørgensen, T., & Sørensen, D.-L. (forthcoming). Codes of good governance: National or
global public values? Accepted for Public Integrity.
Bozeman, B. (1987): All Organizations are Public. Bridging Public and Private Organizational
Theories. San Francisco: Jossey-Bass.
Bozeman, B. 2007. Public Value and Public Interest: Counterbalancing Economic Individualism.
Washington DC: Georgetown University Press.
Hofstede, G. & M. H. Bond (1984) “Hofstede's Culture Dimensions. An Independent Validation
Using Rokeach's Value Survey” Journal of Cross-Cultural Psychology 15 (4): 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.
Kluckhohn, C. (1962). Values and value-orientations in the theory of action: An exploration in
definition and classification. In T. Parsons & E. A. Shils (Eds.), Toward a general theory of
action (pp. 388-433). Cambridge, MA: Harvard University Press.
Lawton, A., & Rose, A. G. (1994). Organisation and management in the public sector. London:
Pitman.
Liu, Ying & Aaron Cohen (2010) Values, commitment, and OCB among Chinese employees
International Journal of Intercultural Relations 34: 493–506.
31
Maguire, M. (1998). Ethics in the public service: Current issues and practice. In A. Hondeghem
(Ed.), Ethics and accountability in a context of governance and new public management. EGPA
Yearbook (pp. 23-34). Amsterdam: IOS Press.
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.
Mintzberg, H. (1983). Structure in Fives. Designing Effective Organizations. Englewood Cliffs:
Prentice-Hall.
Palidauskaite, J. 2010. “Value profile of civil servants in new European democracies through the
lens of embedded ethics”. Paper presented at the European Group for Public Administrationconference, Toulouse.
Soto, W. D.; C. Opheim, H.T. Tajalli (1999) Apples and Oranges? Comparing the Attitudes of
Public Policy Versus Public Administration Students The American Review of Public
Administration 29 (1): 77-91.
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. 1–65).
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., & A, Bardi, A. (1997). Influences of adaptation to communist rule on value
priorities in Eastern Europe. Political Psychology, 18, 385-410.
Schwartz, S. H. & A. Bardi, A. (2001). Value hierarchies across cultures: Taking a similarities
perspective. Journal of Cross-Cultural Psychology, 32, 268-290.
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(3), 361-377.
Van der Waal, Z., Pekur, A. and 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(4), 317–33.
Verdikommisjonen [The National Value Commission]. (1999). Verdier – fellesskap og mangfold.
Verdikommisjonens midtveisrapport [Values – community and diversity. The national value
commission mid-term report]. Verdikommisjonens styringsgruppe, Oslo.
32
Vrangbæk, K. (2009). Public sector values in Denmark: A survey analysis. International Journal of
Public Administration, 32(6), 508-535.
33
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
Fundamenta Should
occasionl, very
normally be nally be
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 us over the 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 (look at 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:
34
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’ Shouldn’
t be
t play
Don’t
important any role 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:
35
Appendix B
Table A1: Ordinal regression of ability to interpret the political climate and signals (variable 2a)
Estimate
Std. Error
Wald
df
Sig.
[variable 2a = 1]
.257
.359
.513
1
.474
[variable 2a = 2]
1.309
.336
15.211
1
.000
[variable 2a = 3]
2.949
.328
80.594
1
.000
[variable 2a = 4]
4.804
.347
191.553
1
.000
Age
-.014
.009
2.253
1
.133
.011
.112
.010
1
.920
Taiwan=1
-1.708
.184
86.426
1
.000
US=1
-1.468
.175
70.362
1
.000
Female =1
Denmark=1
-2.773
.303
83.716
1
.000
Note: Logit link function. Nagelkerke=0.21. Reference category is China and the country coefficients express estimated difference
from China.
Table A2: Ordinal regression of importance of good relations with higher authorities as a motive (variable 3G)
Estimate
Std. Error
Wald
df
Sig.
[variable 3G = 1]
-2.687
.411
42.827
1
.000
[variable 3G = 2]
-.371
.317
1.365
1
.243
[variable 3G = 3]
1.662
.313
28.201
1
.000
[variable 3G = 4]
3.898
.333
137.155
1
.000
Age
-.004
.009
.185
1
.667
Female =1
.429
.112
14.660
1
.000
Taiwan=1
-.747
.176
18.030
1
.000
US=1
-.451
.166
7.376
1
.007
Denmark=1
-2.648
.304
76.060
1
.000
Note: Logit link function. Nagelkerke=0.27. Reference category is China and the country coefficients express estimated difference
from China.
Table A3: Ordinal regression of importance of recognition from management as a motive (var.3i)
Estimate
Std. Error
Wald
df
Sig.
[variable 3i = 1]
-5.512
.514
114.796
1
.000
[variable 3i = 2]
-3.577
.351
104.044
1
.000
[variable 3i = 3]
-1.361
.319
18.169
1
.000
[variable 3i = 4]
.534
.317
2.830
1
.093
Age
.010
.009
1.175
1
.278
Female =1
.392
.111
12.559
1
.000
Taiwan=1
-.632
.174
13.111
1
.000
US=1
-.981
.166
34.855
1
.000
Denmark=1
-.698
.291
5.746
1
.017
Note: Logit link function. Nagelkerke=0.057. Reference category is China and the country coefficients express estimated difference
from China.
36
Table A4: Ordinal regression of importance of high salary as a motive (variable 3d)
Estimate
Std. Error
Wald
df
Sig.
[variable 3d = 1]
-5.561
.393
200.565
1
.000
[variable 3d = 2]
-2.853
.331
74.244
1
.000
[variable 3d = 3]
-.759
.322
5.562
1
.018
[variable 3d = 4]
1.588
.324
23.967
1
.000
Age
.016
.009
2.965
1
.085
Female =1
.060
.111
.289
1
.591
Taiwan=1
-.412
.176
5.475
1
.019
-1.791
.174
106.022
1
.000
US=1
Denmark=1
-2.380
.301
62.668
1
.000
Note: Logit link function. Nagelkerke=0.314. Reference category is China and the country coefficients express estimated difference
from China.
Table A5: Ordinal regression of importance of the organization ensuring career opportunities (variable 1n)
Estimate
Std. Error
Wald
df
Sig.
[variable 1n = 1]
-5.563
.496
125.922
1
.000
[variable 1n = 2]
-3.815
.358
113.272
1
.000
[variable 1n = 3]
-1.651
.325
25.745
1
.000
[variable 1n = 4]
.653
.322
4.116
1
.042
-.019
.009
3.961
1
.047
Female =1
.398
.113
12.491
1
.000
Taiwan=1
-.211
.175
1.457
1
.227
.177
.167
1.125
1
.289
Age
US=1
Denmark=1
-.626
.296
4.481
1
.034
Note: Logit link function. Nagelkerke=0.021. Reference category is China and the country coefficients express estimated difference
from China.
Table A6: Ordinal regression of importance of career opportunities as a motive (variable 3e)
Estimate
Std. Error
Wald
df
Sig.
[variable 3d = 1]
-6.918
.552
157.065
1
.000
[variable 3d = 2]
-4.235
.345
150.510
1
.000
[variable 3d = 3]
-1.692
.322
27.531
1
.000
[variable 3d = 4]
.336
.317
1.120
1
.290
Age
.001
.009
.012
1
.912
Female =1
.256
.111
5.272
1
.022
Taiwan=1
-.733
.175
17.602
1
.000
US=1
-.942
.166
32.109
1
.000
Denmark=1
-2.331
.298
61.100
1
.000
Note: Logit link function. Nagelkerke=0.223. Reference category is China and the country coefficients express estimated difference
from China.
37
Table A7: Ordinal regression of importance of judicial values/due process as general value (variable 1b)
Estimate
Std. Error
Wald
df
Sig.
[variable 1b = 1]
-6.185
.538
132.374
1
.000
[variable 1b = 2]
-5.076
.422
144.913
1
.000
[variable 1b = 3]
-3.471
.366
90.062
1
.000
[variable 1b = 4]
-1.221
.350
12.190
1
.000
Age
.004
.010
.154
1
.694
Female =1
.016
.123
.016
1
.898
Taiwan=1
-1.841
.201
83.734
1
.000
US=1
-1.064
.195
29.792
1
.000
Denmark=1
-.731
.323
5.110
1
.024
Note: Logit link function. Nagelkerke=0.094. Reference category is China and the country coefficients express estimated difference
from China.
Table A8: Ordinal regression of importance of independent professional standards (variable 1c)
Estimate
Std. Error
Wald
df
Sig.
[variable 3d = 1]
-6.942
1.050
43.715
1
.000
[variable 3d = 2]
-4.363
.424
105.927
1
.000
[variable 3d = 3]
-1.922
.332
33.452
1
.000
[variable 3d = 4]
.375
.325
1.334
1
.248
-.010
.009
1.114
1
.291
Female =1
.154
.116
1.766
1
.184
Taiwan=1
.198
.176
1.261
1
.262
US=1
.024
.168
.020
1
.888
Age
Denmark=1
.945
.300
9.923
1
.002
Note: Logit link function. Nagelkerke=0.096. Reference category is China and the country coefficients express estimated difference
from China.
Table A9: Ordinal regression of importance of public insight and transparency (variable 1i)
Estimate
Std. Error
Wald
df
Sig.
[variable 1i = 1]
-7.170
1.050
46.591
1
.000
[variable 1i = 2]
-3.852
.375
105.595
1
.000
[variable 1i = 3]
-1.759
.330
28.434
1
.000
[variable 1i = 4]
.190
.325
.342
1
.559
-.006
.009
.451
1
.502
Female =1
.215
.112
3.684
1
.055
Taiwan=1
-.206
.173
1.420
1
.233
.651
.170
14.606
1
.000
Age
US=1
Denmark=1
-.338
.298
1.289
1
.256
Note: Logit link function. Nagelkerke=0.034. Reference category is China and the country coefficients express estimated difference
from China.
38
Table A10: Ordinal regression of importance of listening to the public opinion (variable 1j)
Estimate
Std. Error
Wald
df
Sig.
[variable 1j = 1]
-4.326
.355
148.185
1
.000
[variable 1i = 2]
-3.072
.332
85.861
1
.000
[variable 1j = 3]
-1.277
.320
15.961
1
.000
[variable 1j = 4]
.740
.317
5.433
1
.020
-.007
.009
.638
1
.424
Female =1
.081
.110
.538
1
.463
Taiwan=1
.028
.171
.026
1
.872
US=1
.498
.164
9.242
1
.002
Age
Denmark=1
-1.848
.295
39.226
1
.000
Note: Logit link function. Nagelkerke=0.204. Ref. category is China. Country coefficients express estimated difference from China.
Table A11: Ordinal regression of importance of political loyalty (variable 1a)
Estimate
Std. Error
Wald
df
Sig.
[variable 1a = 1]
-2.998
.161
346.367
1
.000
[variable 1a = 2]
-1.849
.133
194.194
1
.000
[variable 1a = 3]
-1.014
.122
69.468
1
.000
[variable 1a = 4]
.481
.120
16.164
1
.000
Age
-.052
.005
125.509
1
.000
Female =1
-.159
.110
2.105
1
.147
Confucianism =1
2.427
.152
255.399
1
Note: Logit link function. Nagelkerke=0.223. Non- Confucianist culture is reference (students from Denmark and US)
Table A12: Ordinal regression of importance of loyalty to rules (variable 2d)
Estimate
Std. Error
.000
Wald
df
Sig.
[variable 2d = 1]
-6.878
1.006
46.791
1
.000
[variable 2d = 2]
-3.680
.230
256.419
1
.000
[variable 2d = 3]
-1.587
.132
145.107
1
.000
[variable 2d = 4]
.764
.123
38.812
1
.000
Age
.001
.005
.019
1
.890
Female =1
.133
.112
1.414
1
.234
Confucianism =1
.290
.142
4.208
1
Note: Logit link function. Nagelkerke=0.008 Non- Confucianist culture is reference (students from Denmark and US)
Table A13: Ordinal regression of importance of good relations to higher authorities as a motive (variable 3g)
Estimate
Std. Error
Wald
df
.040
Sig.
[variable 3g = 1]
-3.406
.295
133.358
1
.000
[variable 3g = 2]
-1.112
.135
68.133
1
.000
[variable 3g = 3]
.827
.123
45.489
1
.000
[variable 3g = 4]
2.955
.150
386.032
1
.000
Age
.051
.005
112.591
1
.000
Female =1
.386
.111
12.133
1
.000
Confucianism =1
.434
.140
9.579
1
Note: Logit link function. Nagelkerke=0.208 Non- Confucianist culture is reference (students from Denmark and US)
.002
39
Table A14: Ordinal regression of recognition from management as a motive (variable 3i)
Estimate
Std. Error
Wald
df
Sig.
[variable 3i = 1]
-4.841
.420
132.727
1
.000
[variable 3i = 2]
-2.906
.186
242.846
1
.000
[variable 3i = 3]
-.692
.120
33.345
1
.000
[variable 3i = 4]
1.187
.124
92.419
1
.000
Age
.001
.005
.085
1
.771
Female =1
.358
.110
10.616
1
.001
Confucianism =1
.672
.139
23.217
1
Note: Logit link function. Nagelkerke=0.045 Non- Confucianist culture is reference (students from Denmark and US)
.000
Table A15: Ordinal regression of importance of being adaptable (variable 2e)
Estimate
Std. Error
Wald
df
Sig.
[variable 2e = 2]
-4.352
.281
239.693
1
.000
[variable 2e = 3]
-2.211
.145
231.576
1
.000
[variable 2e = 4]
.322
.122
6.942
1
.008
Age
.009
.005
3.610
1
.057
Female =1
.137
.114
1.455
1
.228
Confucianism =1
-.595
.145
16.824
1
Note: Logit link function. Nagelkerke=0.019 Non-Confucianist culture is reference (students from Denmark and US)
.000
Table A16: Ordinal regression of innovation as a general value (variable 1m)
Estimate
Std. Error
Wald
df
Sig.
[variable 1m = 1]
-6.118
.666
84.474
1
.000
[variable 1m = 2]
-4.307
.406
112.387
1
.000
[variable 1m = 3]
-1.314
.336
15.257
1
.000
[variable 1m = 4]
.746
.334
4.989
1
.026
Age
.008
.009
.755
1
.385
Female =1
.234
.112
4.353
1
.037
-1.025
.180
32.308
1
.000
.571
.306
3.478
1
.062
Taiwan=1
Denmark=1
China=1
-1.139
.169
45.467
1
.000
Note: Logit link function. Nagelkerke=0.127. Reference category is US and the country coefficients express estimated difference
from US students.
40
Table A17: Ordinal regression of ability to innovate (variable 2i)
Estimate
Std. Error
Wald
df
Sig.
[variable 2i = 1]
-6.239
.778
64.236
1
.000
[variable 2i = 2]
-3.863
.391
97.449
1
.000
[variable 2i = 3]
-1.103
.332
11.055
1
.001
[variable 2i = 4]
.858
.331
6.724
1
.010
Age
.009
.009
.932
1
.334
Female =1
.268
.111
5.850
1
.016
Taiwan=1
-.631
.178
12.560
1
.000
.454
.303
2.254
1
.133
Denmark=1
China=1
-.732
.165
19.732
1
.000
Note: Logit link function. Nagelkerke=0.058. Reference category is US and the country coefficients express estimated difference
from US students.
Table A18: Ordinal regression of businesslike operations as a general value
Estimate
Std. Error
Wald
df
Sig.
[variable 1g = 1]
-3.583
.372
92.801
1
.000
[variable 1g = 2]
-1.748
.331
27.844
1
.000
[variable 1g = 3]
.154
.324
.227
1
.634
[variable 1g = 4]
2.155
.331
42.373
1
.000
Age
.015
.009
2.554
1
.110
Female =1
.286
.110
6.765
1
.009
Taiwan=1
-.214
.174
1.512
1
.219
Denmark=1
1.065
.299
12.665
1
.000
China=1
-1.512
.169
80.437
1
.000
Note: Logit link function. Nagelkerke=0.187. Reference category is US and the country coefficients express estimated difference
from US students.
Table A19: Ordinal regression of being willing to take risks (variable 2c)
Estimate
Std. Error
Wald
df
Sig.
[variable 2c = 1]
-3.225
.369
76.560
1
.000
[variable 2c = 2]
-2.189
.339
41.589
1
.000
[variable 2c = 3]
.168
.325
.267
1
.606
[variable 2c = 4]
2.121
.333
40.628
1
.000
Age
.007
.009
.562
1
.453
Female =1
.085
.111
.584
1
.445
Taiwan=1
1.456
.180
65.082
1
.000
Denmark=1
-.228
.299
.581
1
.446
China=1
1.601
.170
88.831
1
.000
Note: Logit link function. Nagelkerke=0.207. Reference category is US and the country coefficients express estimated difference
from US students.
41
Table A20: Ordinal regression of learning and development on the job as a motive
Estimate
Std. Error
Wald
df
Sig.
[variable 3h = 1]
-7.213
.784
84.558
1
.000
[variable 3h = 2]
-5.122
.422
147.369
1
.000
[variable 3h = 3]
-2.563
.348
54.391
1
.000
[variable 3h = 4]
-.163
.339
.231
1
.631
Age
-.004
.010
.198
1
.656
.404
.114
12.570
1
.000
Taiwan=1
-1.104
.184
36.014
1
.000
Denmark=1
-1.397
.314
19.848
1
.000
Female =1
China=1
-.482
.170
7.996
1
.005
Note: Logit link function. Nagelkerke=0.081. Reference category is US and the country coefficients express estimated difference
from US students.
Table A21: Ordinal regression of satisfying immediate needs of users as a general value (variable 1e)
Estimate
Std. Error
Wald
df
Sig.
[variable 2e = 1]
-5.121
.498
105.647
1
.000
[variable 2e = 2]
-3.124
.355
77.593
1
.000
[variable 2e = 3]
-.962
.330
8.497
1
.004
[variable 2e = 4]
1.203
.331
13.217
1
.000
Age
-.012
.009
1.699
1
.192
Female =1
.412
.111
13.663
1
.000
Taiwan=1
.777
.178
19.026
1
.000
-.537
.302
3.165
1
.075
Denmark=1
China=1
.920
.167
30.363
1
.000
Note: Logit link function. Nagelkerke=0.052. Reference category is US and the country coefficients express estimated difference
from US students.
Table A22: Ordinal regression of equal treatment (variable 1k)
Estimate
Std. Error
Wald
df
Sig.
[variable 1k = 1]
-5.191
.395
172.785
1
.000
[variable 1k = 2]
-3.344
.192
303.211
1
.000
[variable 1k = 3]
-1.905
.140
184.103
1
.000
[variable 1k = 4]
-.075
.124
.368
1
.544
Age
-.021
.010
4.223
1
.040
Female =1
.333
.118
7.942
1
.005
Taiwan=1
.043
.353
.015
1
.902
1.959
.362
29.314
1
.000
US=1
China=1
.011
.321
.001
1
.972
Note: Logit link function. Nagelkerke=0.131. Reference category is Denmark and the country coefficients express estimated
difference from Danish students.
42
Table A23: Ordinal regression of user democracy as a general value (variable 1o)
Estimate
Std. Error
Wald
df
Sig.
[variable 1o = 1]
-4.165
.279
222.824
1
.000
[variable 1o = 2]
-2.792
.170
270.916
1
.000
[variable 1o = 3]
-.673
.121
30.764
1
.000
[variable 1o = 4]
1.269
.126
101.227
1
.000
Age
-.011
.009
1.546
1
.214
Female =1
.367
.111
10.932
1
.001
Taiwan=1
.402
.320
1.583
1
.208
US=1
.554
.303
3.349
1
.067
China=1
.179
.292
.377
1
.539
Note: Logit link function. Nagelkerke=0.017. Reference category is Denmark and the country coefficients express estimated
difference from Danish students.
Table A24: Ordinal regression of having ethical awareness (variable 2b)
Estimate
Std. Error
Wald
df
Sig.
[variable 2b = 1]
-5.733
.588
94.915
1
.000
[variable 2b = 2]
-4.257
.303
197.563
1
.000
[variable 2b = 3]
-2.585
.174
220.906
1
.000
[variable 2b = 4]
-.260
.132
3.901
1
.048
Age
-.005
.011
.232
1
.630
Female =1
.566
.131
18.691
1
.000
Taiwan=1
-.840
.377
4.974
1
.026
US=1
1.229
.376
10.664
1
.001
China=1
1.006
.357
7.958
1
.005
Note: Logit link function. Nagelkerke=0.147. Reference category is Denmark and the country coefficients express estimated
difference from Danish students.
Table A25: Ordinal regression of having personal integrity (variable 2f)
Estimate
Std. Error
Wald
df
Sig.
[variable 2f = 2]
-6.104
.521
137.272
1
.000
[variable 2f = 3]
-3.273
.194
285.065
1
.000
[variable 2f = 4]
-1.227
.151
66.025
1
.000
-.001
.012
.004
1
.948
.320
.140
5.222
1
.022
-1.666
.404
16.994
1
.000
.826
.417
3.925
1
.048
Age
Female =1
Taiwan=1
US=1
China=1
-.423
.374
1.278
1
.258
Note: Logit link function. Nagelkerke=0.153. Reference category is Denmark and the country coefficients express estimated
difference from Danish students.
43