Do values transcend cultural and geographical contexts? A comparative study of Schwartz value theory in the People’s Republic of China and Hong Kong
Alistair King (s2067552) Supervisor: Dr. Joshua Robison
BSc Political Science: International Relations and Organisations Leiden University 2019
1. Introduction and Research Question
How citizens arrive at decisions, their process of opinion formation, has been a topic of much research in political science. This process is crucial for understanding the concept of constraint and determining the levels of competence in decision making. Attitude constraint refers to the consistency of beliefs and the ability to predict these beliefs based on some overarching posture that holds belief systems together. Psychological constraint, distinct from social constraint, “helps to account for the political attitudes and preferences of the public” (Feldman, 1988, p. 417). The main premise of this view is that there are overarching postures toward man and society, which act as a form of glue which bind together more specific attitudes and beliefs, such as social justice and natural law (Feldman, 1988, p. 417). This view of psychological constraint is similar to that adopted by Malka, Lelkes and Soto (2017) who contest that it is “psychological mechanisms that lead people to hold either consistently left-wing or right-wing stances across a range of issues” (p. 1046). However, as their article states, research on constraint has found that, in general, “there do not exist psychological constraints” that would hold left and right-wing cultural attitudes to left and right-wing economic attitudes. (Malka et al., 2017, p. 1045). Consequently, the consensus is that constraint is low, implying that attitudes on concrete political issues are unstable over time. This jeopardises citizen competence, as it suggests that unstable attitudes may be subject to elite manipulation.
Nevertheless, research into value systems has shown that they can serve as alternative forms of attitude constraint and, as such, are crucial for understanding the formation of public opinions. Brewer (2003) shows this by stating that there is widespread consensus that citizens use core values as a means to determine where they stand on certain issues. Whilst many researchers differ in their interpretations of the definition of ‘values’, there is agreement that values broadly refer to abstract beliefs that relate to concrete end state goals (Schwartz, 1990; Goren Schoen, Reifler & Chittick, 2016). Schwartz (1992) value theory posits that citizens rely on ten main values in the formation of attitudes. These ten values can be further
subcategorised into four superordinate dimensions; ‘Openness to change and self
enhancement’ referring to values of individualism and values of ‘self-transcendence and conservation’ reflecting socially oriented (collective) goals (Schwartz, 1992; Goren et al., 2016). Goren et al. (2016) have taken this theory and applied it to the US context,
investigating which values affect citizens attitudes across a number of policy issues. They found that, in their research context, ‘self-transcendence’ and ‘conservation’ values played a
greater role in attitude formation than the values of ‘self-enhancement’ and ‘openness to change’ (Goren et al., 2016). These findings reflected research only in a limited, western context. Thus, there remains questions over whether findings in a context different from western centric studies would differ.
The People’s Republic of China (PRC) offers scope into a context where democracy is not the form of government and, therefore, the types of dominant values may be vastly different. This motivates the following question: ‘Which values have the greatest influence on citizens’ attitude formation in the People’s Republic of China and Hong Kong?’. In testing this question, through bivariate correlations and linear regression, it was found that the superordinate dimensions of values espoused by SVT, do not reliably map to the Chinese and Hong Kong contexts – at least as measured by the World Values Survey. However, this research will show that the Schwartz (1992) values reflecting socially oriented, collective goals have a greater impact on opinion formation on economic and political issues, than individual oriented values. This finding shows that as an individual’s affinity for socially oriented goals increases, their support for private ownership of business decreases and their support for technocratic government increases. Moreover, this paper will show that there are differences in the dominant values in the PRC and Hong Kong on opinion formation across these policy areas.
2. Theory of values
To explain why we would expect findings in the PRC to differ to the existing research, a definition of values must be given. As Goren et al. (2016) discuss, Rokeach (1973) gives a cogent definition, describing values as:
“(1) abstract beliefs about (2) desirable end states or behaviours that (3) transcend specific situations, (4) guide evaluation and behaviour, and are (5) ranked in terms of personal importance” (Goren et al., 2016, p. 979).
To this end, values reflect “inter and intra-personal” goals, thus some interpretation about an ‘end-state’ is necessary (Goren et al., 2016, p. 979). Distinguishing between these goals is central to understanding how values shape opinion formation. Brewer’s (2003) work builds upon decades of public opinion research, which came to the conclusion that core values serve as a form of constraint for citizens in helping them to “decide where they stand
on political issues” (p. 173). Brewer (2003) also explores how political knowledge shapes two of these core values, egalitarianism and traditional morality. However, studies on these individual values, such as egalitarianism, have been narrow in their scope. The relative significance of egalitarianism as a core value varies context to context and may not matter as much in contexts not akin to Western-liberal models. This has led many researchers to follow a logic of ‘universal’ value systems, as identified by Schwartz value theory (SVT).
Schwartz (1992) suggests that “clusters of values expressing similar goals” can be arranged into “higher-order value types” and this is the basis for the development of his model (Goren et al., 2016, p. 979). Schwartz and Blisky (1990) distinguish that there are two types of goals represented by values. Terminal goals refer to the end states, mentioned above, whereas instrumental goals relate to “modes of behaviour” (Schwartz & Blisky, 1990, p. 878). Schwartz (1992) elaborates on this interpretation of values and subsequently proposes that there ten broad types of values that emerge; universalism, benevolence, conformity, tradition, security, power, achievement, hedonism, stimulation, and self-tradition (Goren et al., 2016, p. 979). These ten value types can then be categorised into four superordinate dimensions; “self-transcendence” and “conservation” values which relate to “how the individual relates to broader society” and “openness to change” and “self-enhancement” values which relate to the individual’s private interests (Goren et al., 2016, p. 979). SVT, therefore, dictates that the dominant values influencing attitude formation can be
distinguished dichotomously on the goals they represent – i.e. whether they equate to socially oriented or individual goals.
2.1 Understanding Schwartz values
As mentioned, the ten Schwartz (1992) values can be combined to fit into four superordinate dimensions. Table 1 lists each of the ten Schwartz (1992) values and gives a brief description of the principles they espouse.
Table 1
Schwartz (1992) values and their definitions
1. Universalism: understanding, appreciation, tolerance – having a sense of stewardship for the well-being of everyone and nature.
2. Benevolence: maintaining and enhancing the welfare of those whom one is close, and frequent, contact with.
3. Conformity: the restraint of actions, inclinations, and impulses that are likely to upset or harm others and break established social norms, conventions, or
expectations.
4. Tradition: respect for, commitment to, and acceptance of the customs and ideas embodied in one’s own culture or religion.
5. Security: safety, harmony, and stability of the self, personal relationships and society – maintenance of order and the status quo.
6. Power: social status and prestige – having control or dominance over others and resources.
7. Achievement: personal success gained through the demonstration of competence according to established social standards.
8. Hedonism: pleasure or sensuous gratification for oneself – pursuing self-indulgence 9. Stimulation: excitement, novelty, and challenges in life.
10. Self-direction: independent thought and action – having autonomy in choosing, creating and exploring.
Source: Adapted from (Goren et al., 2016, p. 980; Schwartz, 1994, p. 22)
Figure 1 illustrates how these values can be separated along the ‘individual’
‘collective’ distinction. In general, as Cieucich (2011) describes, the distribution of values in the model, where they are placed, is based on how similar or dissimilar the “motivational goals that are the foundation of given value types” are (p. 206). Schwartz value theory holds that values that fit together in their superordinate dimension should be strongly correlated with each other. For example, an individual who scores highly on the value of
‘self-direction’, is expected to score commensurately highly on ‘stimulation’ because they are both under the superordinate dimension of ‘Openness to change’. Political scientists have explored the effects of these dimensions on attitude formation and concluded that the collective
dimensions, ‘Self-Transcendence’ and ‘Conservation’ have the greatest effect, whilst the individual dimensions matter less in this regard (Goren et al., 2016). As Schwartz’s (1992) circular model of values shows, Hedonism is treated tentatively. Many researchers have debated where it fits, whether in ‘Openness to change’ or ‘Self Enhancement’, and as such, many omit it from any analysis. However, for the purpose of this research, Hedonism will be tested as a singular independent variable to better gauge the effect of individualistic values on opinion formation.
Figure 1
Adapted model of Schwartz’s circular model of values
Source: Adapted from (Goren et al., 2016, p. 980; Schwartz, 1994, p. 24; Schwartz & Blisky, 1990, p. 880)
Existing research on Schwartz value theory (SVT) has highlighted the lack of
significance of these ‘individual values’ in opinion formation in many Western contexts. The applicability of SVT with other value systems has also been tested. Cieciuch (2011)
investigates whether SVT can be integrated with Schelerian values, concluding that these values mapped appropriately to their hypothesised places on Schwartz’s (1992) circular model. This finding shows the SVT is compatible with other theories of value distribution. However, many of the conclusions on SVT, have focussed on Western contexts and little attention has been given to other contexts. This study will consequently test whether SVT is as universally applicable as claimed, shedding light on the emergence of value systems in the PRC and comparing this with Hong Kong.
2.2 The emergence of value systems in the PRC
The emergence of dominant value systems in the PRC is complex as, for one, its unique history makes it difficult to map inherently Western values onto the Chinese model
and, secondly, the reinvention of what constitutes modern-day China is an area of much debate. Consequently, it is prudent to highlight this debate in discerning how the PRC’s fractured history has led to its distinct value systems.
Dirlik (2015) discusses how the labels “China” or “Zhongguo” are merely Western creations. The geographical region of China has “been the site of ongoing conflicts over power and control between peoples of different origins” and these conflicts have varied over time in both “geographical scope and demographic composition” as the various dynasties shifted in their respective spheres of influence (Dirlik, 2015). Dirlik (2015) shows that it is difficult to define what is meant by a homogenous notion of “China” or “Chinese people” because these dynasties varied in the demographic composition of their populations. The American political scientist, Lucian Pye, echoed this, claiming that modern-day China remains “a civilization pretending to be a nation-state” (Kissinger, 2012, p. 11). However, over time, there has been a common diffusion of uniquely ‘Chinese’ values that have spread throughout the region we today call the PRC. Kissinger (2012) suggests that, regardless of debates about the concept of China as a state, a common Chinese civilisation can be defined by the “outward radiance of Chinese culture throughout East Asia” (pp. 10-11). To this end, identifying the diffusion of Chinese culture and values over time is akin to identifying a common ‘Chinese identity’.
This section will discuss a few of the forces that have shaped the emergence of the dominant value systems that have come to define the modern-day PRC.
2.2.1 Confucianism.
Though only formally existing as a state, in its current form, since 1949, the PRC has a history of rebellions, dynasties, unification and reunification. Driving development and change throughout these years of fracture and conflict was a distinct “community of values fostered among its population and its government of scholar officials” that were “secular in nature” (Kissinger, 2012, p. 13). The origins of the predominant values of Chinese society were shaped by the teachings and principles of the ancient philosopher Kong Fu-zi –
Confucius (Kissinger, 2012, p. 13). Confucius teachings espoused a common set of principles that have become part of the value system of the modern-day PRC. Confucius was concerned with the “cultivation of social harmony” more than “machinations of power” (Kissinger, 2012, p. 14). This goal is not too dissimilar to that of early classical liberal thinkers. John Stuart Mill, for example, believed that democracy promoted the “highest and most
harmonious development of human capacities” and in the absence of democracy, “ignorance and brutality will prevail” (Heywood, 2012, p. 42). To this end, the end-goals of early liberal and Confucius thought are similar. Some scholars have noted how early Confucian thinkers espoused certain “individualistic and liberal themes” and have even contested that
Confucianism is “compatible with popular democracy and human rights” (Nathan, 2015). However, a clear distinction must be made. In liberal democracy, there is a ‘social contract’, in which consent is given for government to govern to protect the freedoms of the individual. By way of contrast, Confucianism emphasises the “subordination of individual interests”, the deference to elders and authority, and the oppression of women in the pursuit of the “Way” to social harmony (Nathan, 2015). Consequently, liberal democratic values and Confucian values can be distinguished in the methods and goals that they espouse.
Confucius advocated that a “hierarchical social creed” was necessary so as to achieve social harmony (Kissinger, 2012, p. 15). Therefore, subservience to authority and “knowing thy place” are values imbued within Confucianism (Kissinger, 2012, p. 15). As the teachings of Confucius were passed on through the dynasties, it evolved into an ideology which was “secular, hierarchical, authoritarian and family-centred” (Mackerras, 2008, p. 10). These ideas were constitutionally fused to the ruling dynasties who used Confucian values a means of regime legitimation. The Han Dynasty (206B.C-A.D. 220), for example, “adopted Confucian thought as an official state philosophy” thereby merging Confucian values into the political system (Kissinger, 2012, p. 14).
Indeed, the political future of modern China is likely to be determined by these Confucian traditions (Qing & Bell, 2012). In their article, Qing and Bell (2012) discuss how, in the PRC, legitimacy is seen through a different lens relative to how traditional Western thinkers may see it. According to the ‘Gongyang Zhuan’ – a work on New Text
Confucianism – “political power can be justified through three sources: the legitimacy of heaven” the “legitimacy of earth” and the “legitimacy of the human” (Qing & Bell, 2012). These three facets of legitimacy can be achieved through a tricameral legislature, in which a ‘House of Exemplary Persons’ represents ‘sacred’ legitimacy, a ‘House of the Nation’ represents ‘cultural’ legitimacy and a ‘House of the People’ that represents popular legitimacy (Qing & Bell, 2012).
Whilst from the description above, it may seem logical to conflate Confucianism with Western collectivist ideals and values, it does not map so easily. Confucianism is by no means a unified, single doctrine. It has undergone reinventions and adaptions to encompass a “whole array of ideas and practices” which have been “a source of continuous political innovation throughout most of Chinese history” (Schneider, 2019, p. 6). Schneider (2019) seems to suggest that as the classical Confucian works were “rewritten” over time, the central teachings have been diluted (p. 6). However, throughout the remodelling of Confucianism, what has remained central, and undiluted, is the focus on the promotion social harmony. The 2008 Beijing Olympics serve as a testament to how the PRC is attempting to integrate Confucian teachings into modern Chinese society. The “Hu-Wen” administration embedded this idea of social harmony in their use of quotations from classic “Confucian Analects” (Schneider, 2019, pp. 8-9). The Beijing Olympics were, in many ways, designed as a spectacle for the rest of the world to see what direction modern-day China was taking.
Consequently, Confucianism remains a central force underpinning value systems in the PRC. Where Western empires had sought to “spread trade and values throughout the
world”, the Ming Dynasty espoused insular policies, abandoning its once great navy in the late 1400s and severing itself from “the general historical development” of the West (Kissinger, 2012, p. 32). Thus, culture and society within modern-day China developed insularly until the 1800s and the years of foreign encroachment. Yet, in spite of this largely ‘autarchic’ story of development, the dominant values embedded in Confucianism share similarities with values that developed in the West. As discussed, Confucian values advocate the suppression of self-interest in the pursuit of greater social harmony. This is comparable to Schwartz (1992) values that are oriented towards the “collective”, such as ‘conformity’, ‘tradition’ ‘security’ and ‘universalism’. These values espouse common themes, such as the “restraint of actions” and the promotion of social “harmony” (Goren et al., 2016, p. 980). Therefore, whilst Confucian values do not directly map to equating collectivist ideals, they do share similar end goals. Consequently, there is reason to suggest that collectivist values will play a dominant role in opinion formation across all tested issue areas; political, cultural and economic.
2.2.2 The legacy of ‘Maoism’.
Whilst Confucianism was a dominant force shaping the emergence of a common Chinese identity, and remains central to much of modern-day PRC, the tenure of Mao
Zedong influenced the nature of value systems in the PRC, uprooting much of the traditional Confucian teachings. When Mao came to power, he embraced, an albeit vague, vision of “Great Harmony”. (Kissinger, 2012, p. 92). However, in accomplishing this, Mao sought to dismantle ancient Chinese traditions and build a new social order that was beholden to him. Kissinger (2012) outlines the split between Confucian thought and Maoism:
“… where the Confucian tradition prized universal harmony, Mao idealized upheaval and the clash of opposing forces, in both domestic and foreign affairs […]. The Confucian tradition prized the doctrine of the mean and the cultivation of balance and moderation; when reform occurred, it was incremental and put forward as the
“restoration” of previously held values. Mao, by contrast, sought radical and instant transformation and a total break with the past. […] Where traditional China revered the past and cherished a rich literary culture, Mao declared war on China’s traditional art, culture, and modes of thought” (p. 94).
It is, therefore, clear that Mao was ardently ‘anti-Confucian’, attempting to mould a new China in his own image. Yet, in spite of this, Mao’s goal was still akin to traditional Confucian teachings; the concept of “da tong, or the Great Harmony” (Kissinger, 2012 p, 95). This dismantlement of the past is most evident in the Cultural revolution (1966), in which Mao reasserted himself as the central Chinese authority figure. Other political opponents of Mao, were exiled, and in some instances killed, and “Historic buildings, writings and works of art were […] destroyed” (O’Neil, Fields, & Share, 2018, p. 399). The Cultural revolution is seen as the pinnacle moment at which the ‘Cult of Mao’ took hold and created a reverent and omnipotent image around Mao.
The question remains, what was the impact of Maoism on Chinese values? As Kissinger (2012) noted, in spite of being vehemently anti-Confucian, Mao embodied many elements that were espoused by Confucian teachings. Mao established himself as an
omnipotent figure. Therefore, citizens were not only deferent to authority, as per traditional Confucian thought, they revered Mao – as evidenced by the influence of the ‘little red book of quotations’, carried around by most Chinese citizens during Mao’s rule. This ensured a legacy in which the suppression of individual interests in pursuit of the collective good was advocated. Mao’s goal of the “Great Harmony” was tantamount to the traditional Confucian goal of Social harmony. The legacy of Maoism, therefore, suggests that values relating to individual interest will play a lesser role in opinion formation than those of collective values.
For the focus of this study, the impacts of Mao’s reign would suggest that the Schwartz (1992) value of tradition will perhaps play a lesser role in opinion formation.
2.2.3 Civil society.
The nature of the space of Chinese civil society also provides insight into which values can be expected to be seen as dominant in opinion formation. Distinct about the PRC is the fusion between party and state; the Chinese Communist Party (CCP) is, by all extents and purposes, the government and only party in the PRC. Under Mao, the CCP claimed to “represent all legitimate social interests” and, consequently, civil society “did not officially exist” (O’Neil et al., 2018, p. 418). The limited space for social engagement has had significant consequences for the diffusion of values. The restriction of any organisations, with interests out of line with that of the CCP, limits the ability of citizens to think and act freely. Thus, under Mao’s China, values associated with Western liberal democracies were not only lacking but were actively quashed in an effort to maintain social control.
Since the 1978 and the eras of reform – notably the role of Deng Xiaoping in
‘opening’ up the PRC to the rest of the world’ – Chinese social space has undergone multiple changes. The government remains pursuant in restricting social space through the use of mass organisations, or oxymoronically termed “government-operated nongovernmental
organisations” (GONGOs) which are designed to “mobilize social groups to fulfil [the CCPs] own national goals” (O’Neil et al., 2018, p. 418). On top of this, there is strict limits to the operation of foreign Non-governmental organisations (NGOs), with all of them requiring a local, state endorsed, sponsor to ensure operation (O’Neil et al., 2018).
The strict controls on social space has driven Chinese citizens to engage in
unconventional forms of political participation, such as protests. From protests over illegal land seizures to government corruption, it is estimated that around five hundred protests take place every day (O’Neil et al., 2018). The rise in prominence and frequency of protests on these issue types demonstrates citizens belief in justice, and suggests that some traditional Western, liberal notions of justice are also evident in the minds of Chinese people.
The most promising area of Chinese civil society is seen in its internet activities. It is estimated that China has over seven hundred million ‘netizens’ – active users of online resources – and these seven hundred million netizens are becoming increasingly connected with each other and the rest of the world (O’Neil et al., 2018). Whilst the PRC has banned the use of the most used, western founded, social media platforms such as Facebook, Twitter
and Instagram, Chinese netizens are using alternative forums to interact. For many years, the microblogging platform ‘Weibo’ served as a relatively open platform in which citizens could share “uncensored news and rumors” (O’Neil et al., 423). The PRC has since cracked down, no longer allowing anonymous posting on the site. Consequently, Chinese netizens have switched to other social media platforms like WeChat. This shows that whilst the PRC may try and continue to restrict and limit social space, citizens will find alternative means to interact and communicate.
This discussion of Chinese civil society reveals that Chinese citizens are pursuant in finding new ways to engage and interact in open spaces that are less stringently controlled by the CCP. As such, it would point towards the importance of the value of ‘self-direction’ and questions the importance of the value of ‘conformity’ in opinion formation. At the same time, however, the strict confinement of Chinese social space and civil society would suggest that individual oriented values will still play a lesser role than socially oriented values on opinion formation.
2.3 Half-way between China and the West: Hong Kong
Whilst China developed largely in isolation from the West, for most of its history, the extent of foreign encroachment upon Hong Kong in the would suggest that different,
Western, values may have been diffused in the region. As such, Hong Kong may serve as an example of a scenario where distinct ‘Chinese’ values have mixed with Western notions of democracy and liberalism. Therefore, a comparative analysis of both cases may reveal how different values emerge upon exposure to Western influence. As per SVT, we should expect the same values to be dominant in opinion formation across all contexts. However, the vast differences between the U.S western contexts and the Chinese and Hong Kong contexts strongly encourages testing of this claim.
Hong Kong was ceded to Britain after the first opium war in 1842 – during which foreign powers began to severely encroach upon contemporary China’s sovereignty (O’Neil et al., 2018). Consequently, Hong Kong was allowed to develop and embrace capitalist economic models earlier than mainland PRC. The British colonial government gradually liberalised “Hong Kong’s political institutions”, opening up “new venues for political participation and political contestation” (Ng & Fulda, 2017, p. 386). The establishment of a competitive political environment is something inconsonant with the Chinese model. Political participation and contestation are uniquely liberal concepts that are completely alien to the
Chinese system. Therefore, it would stand to reason then that the story of value emergence in Hong Kong would be different to that of China. The exposure to Western, liberal ideas and political systems would suggest that the values contained with the Superordinate dimension of ‘self-enhancement’ may play a greater role than in China.
2.4 Hypotheses:
From the discussion of value systems in the PRC and Hong Kong, there is reason to believe that the dominant values underpinning opinion formation may differ from other Western contexts. Consequently, it is hypothesised that collective, socially oriented values (universalism, benevolence, conformity, tradition and security) will have greater effects on influencing policy opinions than individually oriented values (H1). This logic follows from the discussion above which shows the dominance of collective values in the PRC through Confucian principles. Moreover, the previous findings of Goren et al. (2016) show that the superordinate dimensions of Self-transcendence and Conservation play greater roles in
opinion formation in the US. In line with this, it is expected that as respondents’ affinity with collective socially oriented values increase, support for private ownership of business will decrease (H2). Confucian principles discourage the pursuit of self-interest, thus it stands to reason that collective goals will be negatively correlated with private ownership of business. Moreover, as affinity for collective values increases, support for technocratic government should increase (H3) and respondents will be less inclined to be trusting of individuals from a nation that is different to their own (H4). Collective values, such ‘benevolence’, stress
principles of stewardship in ensuring the welfare of others close to the individual thus lending credence to H3 and H4. As discussed, Hong Kong is our case of comparison and, as such, it serves as a middle ground between the context of the PRC and the West. It is expected that individual values will play a greater role in opinion formation in Hong Kong, than in the PRC (H5). The diffusion of western norms throughout British occupation in Hong Kong would suggest that traditional Confucian values are intermingled with western notions of democracy and liberalism, thus suggesting that values emphasising the pursuit of the ‘self’ will be more dominant than in the Chinese context.
3. Research Design and Methodology
This research takes the form of a two-case comparative study, focussing primarily on the case of the PRC whilst using Hong Kong as a case for comparison. As outlined in the
theory section, the unique history and emergence of the PRC’s value systems would suggest that Schwartz (1992) values oriented towards collectivist goals – those values associated with ‘self-transcendence’ and ‘conservation’- will play a greater role in opinion formation across all policy areas than values of self-enhancement or openness to change. Hong Kong, by way of contrast, tells a different story of the way values have emerged. The exposure to capitalist and liberal ideas, much earlier than the PRC, would suggest Schwartz (1992) values such as ‘achievement’, ‘power’, ‘hedonism’ and ‘self-direction’ will play a greater role in opinion formation than in the context of the PRC.
In testing these assumptions, measurement must be taken of each of the ten Schwartz (1972) values. The World Values Survey (WVS) is one such source that specifically
measures these values – among other things – in most countries across the world. The WVS conducts multinational surveys across many political and demographic dimensions. The survey in the PRC was carried out between 2012 and 2013 by researchers at Peking university. The sampling method used was “GPS/GIS Assistant Area Sampling” to obtain results from all regions in the PRC, excluding Hong Kong, Macau and Taiwan, and
participants were randomly sampled via stratified, multi stage probability proportional to size (PPS) sampling. Interviews were conducted at the residences of each respondent. The sample size is 2300. However, responses to questions such as ‘don’t-know’ are coded as missing, and not included in any subsequent analyses, thus N will always be smaller than 2300.
The majority of methodological information is, unfortunately, not listed for the survey in Hong Kong. However, the survey was carried out by the University of Hong Kong and a method of random sampling is listed. The sample size for Hong Kong is 1000 and, again, N for analyses will always be slightly smaller than this number.
3.1 Survey issues
For this research, it is important to note the potential effects of social desirability bias. Studies have shown that respondents have a tendency to “present themselves in a good light” to the interviewer and will, therefore, “over-report socially ‘desirable’ behaviour” (Halperin & Heath, 2017, p. 268). This is especially relevant for the measurement of ‘sensitive’ questions. One proposed method so as to minimise this bias is to move these ‘sensitive’ questions near the end of the survey (Halperin & Heath, 2016). In the WVS, these politically sensitive questions do tend to come near the end of the survey. Nevertheless, it is worth being mindful of the effects this bias may bring.
3.2 Data measurement: independent variables
To measure the independent variables, the WVS poses ten statements about
hypothetical individuals, and respondents are asked how much they consider themselves like that person – along a six point scale Likert scale. These statements are in line with Schwartz’s (1992) own suggestions for how to measure these values and are widely recognised as good measures for Schwartz (1992) values. The statements asked, and how they correspond to the Schwartz (1992) values, are delineated in the table below.
Table 2:
Schwartz (1992) Values and the statements from the WVS used to measure them
Schwartz (1992) Value Question Posed in WVS Mean (Sd)
China Hong Kong
‘Openness to change’
Self-Direction
Stimulation
V70. ‘It is important to this person to think up news ideas and be creative; to do things ones own way.’
V76: ‘Adventure and taking risks are important to this person; to have an exciting life’ .45 (.23)a .53 (.27)a .37 (.27)a .48 (.21)b .60 (.26)b .35 (.28)b ‘Self-Transcendence’ Universalism Benevolence
V78. ‘Looking after the environment is important to this person; to care for nature and save life resources’.
V74. ‘It is important to this person to do something for the good of
society’.
V74B. ‘It is important to help people living nearby; to care for their needs’. .66 (.18)a .63 (.23)a .66 (.21)a .70 (.20)a .66 (.17)b .67 (.23)b .60 (.24)b .72 (.20)b
‘Conservation’
Conformity
Tradition
Security
V77. ‘It is important to this person to always behave
properly; to avoid anything people would say is wrong’.
V79. ‘Tradition is important to this person; to follow the customs handed down by one’s religion or family’.
V72. ‘Living in secure surroundings is important to this person; to avoid anything that might be dangerous’.
.62 (.18)a .58 (.25)a .62 (.27)a .66 (.23)a .62 (.18)b .61 (.25)b .57 (.28)b .69 (.23)b ‘Self Enhancement’ Power Achievement
V71. ‘It is important to this person to be rich; to have a
lot of money and expensive things’
V75. ‘Being very successful is important to this person; to have people recognize one’s
achievements’ .55 (.22)a .53 (.26)a .58 (.26)a .47 (.21)b .39 (.26)b .56 (.26)b
Hedonism V73. ‘It is important to this person to have a good time; to “spoil” oneself’
.42 (.25)a .62 (.26)b
Source: Inglehart, R., C. Haerpfer, A. Moreno, C. Welzel, K. Kizilova, J. Diez-Medrano, M. Lagos, P. Norris, E. Ponarin & B. Puranen et al. (eds.). 2014. World Values Survey: Round Six - Country-Pooled Datafile
Version: http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp. Madrid: JD Systems Institute.
However, the main point of contention is in the translation and applicability of these questions in the Chinese context. These questions are translated into Mandarin, however, in the translation, the meaning of the questions may be lost. In the PRC, individuals may not have a conception or notion of some of these values. Consequently, for some values, such as ‘hedonism’, the expected results may be noticeably low. This may suggest that the model of Schwartz (1992) values is not compatible with the Chinese context. Alternatively, it may
suggest that the statements used to measures the values are not suitably phrased for different contexts.
The statements used to measure the Schwartz (1992) values, as shown in Table 2, are measured on a six-point scale. Individuals were asked how much they identify with the statement, with 1 meaning ‘very much alike’ and 6 meaning ‘not at all alike’. Responses such as ‘don’t know’ are coded as missing values in the data set, and are not included in the
analyses. For this research, these variables were then recoded to lie on a 0-1 scale, with 0 representing no affinity with the statement, and 1 representing the strongest affinity. To capture the superordinate dimensions, these variables are then added and an average is taken to create a new variable.
Table 2 also shows the mean and standard deviation for each of the independent Schwartz (1992) values – a value closer to 1 shows a stronger affinity to the Schwartz (1992) value. This already begins to show how the samples from the PRC and Hong Kong differ. Whilst the means for the four main superordinate dimensions are quite similar, and almost the same, the biggest difference is seen in the measurement of Hedonism. In the Chinese survey, the mean response was 0.42 with a standard deviation of 0.25. Comparably, in Hong Kong, the mean score for Hedonism was 0.62, with a standard deviation of 0.26, which is
significantly greater than the score in China. This is in line with the hypotheses that the values associated with freedom, and liberalism, would be more prevalent in Hong Kong than in China.
3.3 Data measurement: dependent variables
In studying the effects of Schwartz (1992) values on public opinion in the United States (US), Goren et al. (2016) use an array of measures to capture the change in attitudes on certain policy issue areas. For instance, to measure economic issues, they examine responses to questions about “government spending” and “federal responsibility for economic security” (Goren et al., 2016, p. 986). Whilst many of the measures in the study of Goren et al. (2016) are not available in the WVS, for instance the seven points liberal-conservative scale is not a question asked in the PRC, their work provides strong examples of questions that can capture economic, political and cultural issues.
To measure attitudes on economic issues in this research, respondents were asked whether they were in favour of private or state ownership of business, with 1 representing maximum support for private ownership, and 10 representing maximum support for
government ownership. This scale was then recoded to lie on a 0-1 scale, so that 1 reflects maximum support for private ownership, and 0 reflects maximum support for state
ownership.
For political issues, respondents were asked whether they agree with ‘having experts, not governments, make decisions according to what they think is best for the country’ and responses were originally coded so that 4 denotes a response of ‘very bad’ and 1 denotes ‘very good’. This question probes the respondents’ support for technocratic government and was also recoded to 0-1 scale so that 1 represents a great deal of support for technocratic government and 0 reflects no support for technocratic government.
In measuring cultural issues, this research decided to focus on the policy issue of immigration. The WVS asks, respondents how much they trust people of another nationality, 1 representing a great deal of trust, and 4 represent no trust at all. This was again recoded to fit a 0-1 scale, whereby 1 represents a great deal of trust, and 0 represents no trust at all.
Important to the analyses is to control for demographics, specifically age and education. The PRC’s population is divided between rural citizens in more remote western areas, and urbanised areas on the east coast and this is a big in education level. More affluent individuals in the East have better opportunities and access to education. It is therefore important to control for the education level of individuals. Age is also a significant control variable. A lot of the PRC’s older population will rely on values of tradition in opinion formation, whereas younger generations may be more self-determined.
3.4 Method
Recall that all independent variables were originally recorded on a ‘Likert’ scale – with varying degrees of affinity corresponding to the statement posed. Whilst Likert scales are by nature ordinal-level variables, there is consensus that if the scale includes at least five categories, it can be treated as a continuous variable (Grace-Martin, n.d.). Consequently, after coding the independent variables to a common 0-1 scale, there is justification enough to treat them as continuous variables. The dependent variables are also measured on a Likert scale, and can be treated as continuous. Therefore, a linear regression will be carried out to test the relationship between the independent and dependent variables. Prior to running this
regression, bivariate correlations were carried out to test the fit of the Schwartz (1992) values within the superordinate dimensions.
4. Analyses 4.1 The fit of Schwartz (1992) values
To first test the fit of the independent variables within the superordinate dimensions, bivariate correlation analyses were carried out to test their suitability. Tables 3 and 4 show the matrices of the correlations for the PRC and Hong Kong respectively. The separation of values into rows delineates the superordinate dimension to which they belong; for instance power and achievement share a row as they belong to the superordinate value of ‘Self-enhancement’. Whilst almost all of the correlations are statistically significant, the
coefficients themselves are quite low. Within each superordinate group of values, we would expect relatively high correlation coefficients. For example, an individual who scores high on ‘self-direction’ should be expected to score equally high on ‘stimulation’ as they both are contained with the superordinate value ‘Openness to change’ – this is, however, not the case. Table 3 gives a positive correlation coefficient r of 0.45 for the PRC, whilst table 4 gives a lower positive r value of 0.33 for Hong Kong. As a rule of thumb, a correlation of coefficient of 0.7 is considered strong, 0.5 is considered moderate and 0.3 is considered weak (Mindrila & Balentyne, n.d.). These correlation coefficients can, therefore, be considered to be weaker than what we would expect. The weakness of these correlation coefficients prompted further investigation into the reliability of the use of the Schwartz (1992) values as measures of the superordinate dimensions.
Table 3:
Bivariate correlation matrix for Schwartz values in the PRC
V70 V76 V78 V74 V74b V77 V79 V72 V71 V75 V73 V70 Self-direction V76 Stimulation .45** V78Universalism V74Benevolence (a) V74Benevolence (b) .28** .30** .25** .12** .11** .07** .53** .53** .67** V77 Conformity V79 Tradition V72 Security -.01 -.05** .233** -.09** -.17** .06** .30** .37** .40** .17** .29** .44** .20** .28** .42** .36** .27** .25** V71 Power V75 Achievement .36** .47** .32** .40** .10** .33** .13** .41** .11** .37** .07** .07** -.03 .08** .23** .26** .38** V73 Hedonism .30** .37** .10** .13** .07** .02 -.06** .18** .31** .27** * = Statistically significant at the level p<0.05; ** = statistically significant at the level p<0.01
Table 4: (overleaf)
* = Statistically significant at the level p<0.05; ** = statistically significant at the level p<0.01
Cronbach’s alpha is designed to test the reliability of multiple variables used to measure a common concept – in this case, Schwartz’s (1992) superordinate dimensions. As Tavakol and Dennick (2011) explain, the alpha test provides a “measure of internal
consistency of a test or scale” and there is consensus that a score above 0.7 is acceptable (Tavakol and Dennick, 2011, p.53; Field, 2013).
Table 5
Cronbach’s alpha test of Schwartz (1992) values in the PRC and Hong Kong
Schwartz (1992) Value Cronbach’s alpha
PRC Hong Kong Openness to change Self Enhancement Self Transcendence Conservation .62 (N=2184) .56 (N=2185) .79 (N=2173) .56 (N=2156) .49 (N=992) .48 (N=996) .68 (N=990) .46 (N=991)
As shown in table 5, all of the superordinate dimensions have relatively low alpha levels, apart from Self-transcendence, which is an acceptable level in the PRC, 0.79, and on the border of acceptability in Hong Kong, 0.68. For an alpha test, it is important to “reverse score” all variables, as was done when recording the independent variables (Field, 2013, p. 713). The fact that these alpha levels remain low suggests that the questions used to measure the Schwartz (1992) values are not reliable measures of the superordinate dimensions. Consequently, there is an either issue with the compatibility of Schwartz’s model in the PRC and Hong Kong, or the questions used to capture these higher dimensions are not compatible in these contexts. Therefore, moving forward, each independent variable will be measured separately – and not as part of the superordinate dimension.
V70 V76 V78 V74 V74b V77 V79 V72 V71 V75 V73 V70 Self-direction V76 Stimulation .33** V78 Universalism V74Benevolence (a) V74Benevolence (b) .16** .25** .23** .08** .20** .16** .34** .39** .52** V77 Conformity V79 Tradition V72 Security .11** .13** .10** .19** .09** -.04 .29** .35** .28** .16** .27** .23** .25** .24** .28** .25** .25** .18** V71 Power V75 Achievement .28** .28** .37** .40** .02 .18** .09** .37** .02 .36** .08* .27** .12** .21** .22** .22** .31** V73 Hedonism .19** .18** .21** .30** .31** .20** .18** .37** .28** .27**
4.2 Regression results: China
To test the research question, which values will have the greatest effect on citizens attitude formation across economic, political and cultural issues, linear regression was carried out for the samples in the PRC and Hong Kong. H1 predicted that collective values would be more dominant across all policy dimensions in opinion formation – regardless of context. Table 6 lends some support to this. Universalism has statically significant effects for economic and political issues. Its high effect (relative to the beta coefficients of the other independent values) of -0.096 on economic issues shows that as an individual’s affinity with Universalist values increases by 1 unit, their support for private ownership of business decreases by 0.096. Recall that the measure for economic issues was recoded, so that 1 indicates strong support for private ownership of business, and 0 represents strong support for private business. Thus, as affinity with ‘universalism’ increases, support for private
ownership of business decreases by 0.096 units. This effect on economic issues is substantiated by ‘conformity’, and ‘security’ with respective values of -0.069 and -.074 which are significant at the p<0.05 level. This lends support to H2, that collective values will draw people to be less supportive of private ownership of business. For political issues, 1 denoted a great deal of support for technocratic government, and 0 equated to a lack of support for technocratic government. There is thus a lack of support for H3. ‘Tradition’ is positively correlated with support for technocratic government with a statistically significant value of 0.057, whilst growing affinity with ‘universalism’ shows a decrease in support of -0.073 – both values significant at the p<0.05 level. Consequently, there is lack of support for H3. Only ‘power’ had a statistically significant effect on the measure for cultural issues – which was coded so that 1 indicated great trust in individuals from a different nation and 0 represented a complete lack of trust. The result shows that as individuals’ affinity with ‘power’ increases, their trust decreases by -0.088, significant at the p<0.01 level. This would suggest that there is mixed support for H1 as none of the collective values had statistically significant effects on opinions on cultural issues. There is also a lack of support for H4. Collective values did not have statistically significant or substantial effects on opinions on cultural issues. If anything, individual goals had a greater effect on influencing opinion on cultural issues.
Table 6:
Effects of Schwartz values on opinions across economic, political and cultural issues in the PRC (controlling for age and education)
Economic Issues Political Issues Cultural Issues
Self-direction .001 (.029) .017 (.026) .000 (.032) Stimulation .008 (.028) -.013 (.025) -.012 (.031) Universalism -.096** (.035)** -.073* (.032) .001 (.043) Benevolence (a) .007 (.043) -.030 (.040) -.008 (.051) Benevolence (b) .026 (.045) .005 (.042) .057 (.053) Conformity -.069* (.028)* .020 (.025) -.027 (.031) Tradition -.021 (-.020) .057* (.025)* .051 (.031) Security -.074* (.033)* -.004 (.031) -.022 (-.038) Power .021 (.028) -.004 (.026) -.088** (.031)** Achievement -.018 (.031) .057* (.029)* .031 (.034) Hedonism -.012 (.027) .025 (.025) .008 (.030) R2 Adj R2 N .049 .042 1926 .014 .006 1609 .016 .005 1166
Results of multiple linear regression. First row shows unstandardised b coefficients. Second row shows standard error, in brackets.* = Statistically significant at the level p<0.05, ** = Statistically significant at the level p<0.01, ***= Statically significant at the P<0.001
4.2 Hong Kong
The regression analysis for Hong Kong provides support for H5, especially for
opinions on political issues. For political issues, ‘stimulation’, ‘power’ and ‘achievement’ (all individual oriented values) were positively correlated with opinions on political issues and statically significant at the p<0.01 level for stimulation and power, and at the p<0.05 level for achievement. Thus, as individuals’ affinity for these values increases, their support for
technocratic government increased. Moreover, stimulation had a statistically significant effect on opinions on cultural issues. As an individual’s affinity with stimulation increased, their trust in individuals from other nation’s increased by 0.129 units, significant at the p<0.001 level. Therefore, whilst not all individually oriented goals were statistically significant, there is evidence to suggest that they played a greater role in affecting opinion in Hong Kong than in the PRC, lending support to H5. There is also support for H2 – though less than in the PRC. On economic issues, support for private or state ownership business, the second measure of ‘benevolence’ (b) was strongly negatively correlated with private ownership of business, significant at the p<0.001 level. Thus, whilst not all collective values had significant, or relevant, effects, there is still support that these collective goals play a great role in opinion formation in the PRC and Hong Kong.
Table 7:
Effects of Schwartz values on opinions across economic, political and cultural issues in Hong Kong (controlling for age and education)
Economic Issues Political Issues Cultural Issues
Self-direction .020 (.034) -.040 (.036) -.002 (.033) Stimulation .055 (.034) .085** (.036) .129*** (.033)*** Universalism .002 (.041) .007 (.042) .051 (.039) Benevolence (a) .017 (.041) .047 (.040) .042 (.039) Benevolence (b) -.184*** (.050)*** -.030 (.052) -.004 (.048) Conformity -.005 .051 .022
(.034) (.036) (.033) Tradition -.007 (.031) -.043 (.032) .060* (.030)* Security -.009 (.040) -.020 (.042) -.006 (.039) Power .029 (.035) .103** (.037) .011 (.034) Achievement -.024 (.037) .087* .038 -.049 (.036) Hedonism .013 (.035) .014 (.037) .007 (.034) R2 Adj R2 N .040 .027 955 .054 .041 955 .062 .049 959
Results of multiple linear regression. First row shows unstandardised b coefficients. Second row shows standard error, in brackets.* = Statistically significant at the level p<0.05, ** = Statistically significant at the level p<0.01, ***= Statically significant at the P<0.001
5. Conclusion
The findings of this paper point towards two separate conclusions. Firstly, the testing of Schwartz (1992) values in the PRC and Hong Kong has shown that they do not map reliably into the superordinate dimensions. This means that the questions in the WVS used to measure the Schwartz (1992) values do not serve as reliable measures for the higher value dimensions in the PRC and Hong Kong. Consequently, this suggests that Schwartz’s (1994) circular model of values does not fit the Chinese and Hong Kong contexts, with the current measures employed, and that a different model would be better suited to map the values across the ten motivational domains. The notion of some of these values may not exist in the Chinese and Hong Kong contexts, and the ability to map them to higher dimensions is reduced.
The superordinate dimensions are amalgamations of the various ten motivational value types, thus no conclusions can be drawn about these dimensions from the regression analyses carried out on the individual values alone – apart from that which is mentioned above. Nevertheless, this research showed tentative support for H1, suggesting that collective
goals played a dominant role in effecting opinion preferences in the PRC and Hong Kong. There was also support for H5, there were clear differences in the respective effects of individual oriented values in Hong Kong and in the PRC.
The findings point towards the need to better understand the applicability of SVT in the Chinese and Hong Kong contexts. In particular, it calls for the construction of surveys that better capture Schwartz (1992) values in these contexts. It also suggests the need for reconceptualisation of the dominant values in China.
This being said, the results in this paper are by no means unanimous. Regression coefficients were, in most cases, substantively small – even in those cases in which the result was statistically significant. Analysing third party survey data from a country with a vastly different political system to that of the researcher can lead to spurious conclusions. The discussion of the PRC’s limits on social space would suggest that survey biases may be of greater consideration than in liberal democracies. Respondents may be weary of potential government observation of survey data and may have thus adjusted their responses to be more in line with party ideology.
References:
Brewer, P. (2003). Values, political knowledge, and public opinion about gay rights: A framing-based account. Public opinion quarterly, 67(2), 173-201.
Cieciuch, J. (2011). Integration of Schwartz’s value theory and Scheler’s concept of value in research on the development of the structure of values during adolescence. Polish psychological bulletin, 42(4), 205-214.
Dirlik, A. (2015). Born in Translation: “China” in the making of “Zhongguo”. Retrieved from: http://www.boundary2.org/2015/07/born-in-translation-china-in-the-making-of-zhongguo/ Accessed 25 November 2019.
Feldman, S. (1988). Structure and consistency in public opinion: The role of core beliefs and values. American journal of political science, 32, 416-440.
Field, A. (2013). Discovering statistics using IBM SPSS statistics. (4thed.). SAGE publications ltd.
Goren, P., Schoen, H., Reifler, J., Scotto, T., & Chittick, W. (2016). A Unified Theory of Value-Based Reasoning and U.S. Public Opinion. Political Behavior, 38(4), 977-997.
Grace-Martin, K. (n.d.). Can Likert Scale Data ever be Continuous?
Retrieved from: https://www.theanalysisfactor.com/can-likert-scale-data-ever-be-continuous/ Accessed 19th December 2019.
Halperin, S., & Heath, O. (2017). Research methods in political science (2nded.). Oxford University Press.
Heywood, A. (2012). Political ideologies: an introduction (5th ed.). Hampshire: Palgrave Macmillan.
Inglehart, R., C. Haerpfer, A. Moreno, C. Welzel, K. Kizilova, J. Diez-Medrano, M. Lagos, P. Norris, E. Ponarin & B. Puranen et al. (eds.). (2014). World Values Survey: Round Six - Country-Pooled Datafile
Version: http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp. Madrid: JD Systems Institute.
Jacoby, W., (2006). Value Choices and American Public Opinion. American Journal of Political Science, 50(3), pp.706–723.
Kissinger, H. (2012). On China (2nd Ed.). London: Penguin.
Mackerras, C. (2008). China in transformation, 1900-1949 (2nd ed.). London: Routledge.
Malka, A., Lelkes, Y., & Soto, C. J. (2019). Are Cultural and Economic Conservatism
Positively Correlated? A Large-Scale Cross-National Test. British Journal of Political Science, 49(3), 1045-69.
Mindrila, D., & Balentyne, P. (n.d.). Scatterplots and Correlation. Retrieved from:
https://www.westga.edu/academics/research/vrc/assets/docs/scatterplots_and_correlat ion_notes.pdf
Nathan, A. (2015). The problem with the China model. Retrieved from:
http://www.chinafile.com/reporting-opinion/viewpoint/problem-china-model Accessed 21 October 2019.
Ng, N., & Fulda, A. (2017). The religious dimension of Hong Kong’s umbrella movement. Journal of Church and State 60 (3), 377-397.
O’Neil, P. H., Fields, K., & Share, D. (2018). Cases in comparative politics (6thEd.). New York: W. W. Norton & Company.
Qing. J., & Bell., D.A. (2012). A Confucian constitution for China. Retrieved from:
https://www.nytimes.com/2012/07/11/opinion/a-confucian-constitution-in-china.html Accessed 21 October 2019.
Schneider, F. (2019). ‘Creative Reinventions of China’s “Harmonious Society”’. Paper presented at the academic conference China is Creative, 29-31 August 2019, Amsterdam: University of Amsterdam.
Schwartz, S., & Bilsky, W. (1990). Toward a Theory of the Universal Content and Structure of Values: Extensions and Cross-Cultural Replications. Journal of Personality and Social Psychology, 58(5), 878-891.
Schwartz, S. (1992). Universals in the Content and Structure of Values: Theoretical Advances and Empirical Tests in 20 Countries. Advances in Experimental Social Psychology, 25, 1–65.
Schwartz, S. (1994). Are there universal aspects in the structure and contents of human values?. Journal of social issues, 50(4), 19–45.
Schwartz, S., Caprara, G., & Vecchione, M. (2010). Basic Personal Values, Core Political Values, and Voting: A Longitudinal Analysis. Political Psychology, 31(3), 421-452.
Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International journal of medical education, 2, 53-55.
