What explains the resilience of the authoritarian regime in mainland China? Why do people in China express a high level of trust toward the government even though they strongly desire democracy? Conversely, why do many people in Taiwan say they would prefer a government less democratic than the one they actually have? These findings run counter to Western scholars’ expectations about democracy and contradict current theories of politics. An understanding of prevailing cultural norms about politics in these political systems can help resolve these puzzles.

In the aftermath of the 1989 Tiananmen Square incident, most observers of Chinese politics believed that the Chinese Communist Party (CCP) had permanently lost its legitimacy and that continuation of its rule would necessitate a reliance on brute force. The collapse of the Soviet Union in 1991 reinforced this belief. Scholars argued that the CCP had only two choices: to join the “third wave” of democratization, which would bring about the party’s collapse, or to initiate serious political reforms, which would delay collapse but weaken the party’s authority. In effect, the CCP could choose between painless euthanasia or a prolonged and painful dying process.¹

Observers also believed that the survival of the CCP and successful economic reform were incompatible. In one group were scholars influenced by modernization theory, who held that successful economic reform would require concomitant political change. Because the CCP refused to introduce meaningful political reform, its economic reforms were unsustainable. The late Gordon White described the CCP’s dilemma – the democratic transition it would have to contend with after unleashing economic reform – as “riding the tiger” (a reference to the Chinese saying “If you are riding a tiger it is hard to get off”) (White Reference White1993, 2). Bruce Gilley argued that democratization was inevitable, despite the regime’s efforts to obstruct it (Gilley Reference Gilley2004).

In a second camp were scholars who doubted the long-term viability of the Chinese economic reform program. In its first stage, reform could create a win-win situation: economic gains for the majority would translate into support for the regime or at least for its reform program. However, support would depend on the regime’s continuing ability to provide economic gains. This would become increasingly difficult in a second stage, when the necessary reform of obsolete state-owned enterprises (SOEs) would jeopardize the interests of a large part of the population. The provisional peace between state and society would vanish, and a disgruntled majority would challenge the regime, leading to a crisis. If the CCP tried to avoid this crisis by refusing to move from the first stage to the second stage, the resulting financial burden of the SOEs would drag the banking system into insolvency, leading eventually to regime collapse (Chang Reference Chang2001).

Such predictions aligned with mainstream theories of comparative politics and democratic transitions. Authoritarian leaders, by definition, hold political power without the periodically renewed consent of their people, but they do allow their people certain freedoms not allowed in totalitarian societies, including the freedom to criticize the government in private settings. According to mainstream theory, the weak legitimacy and overcentralization of power built into Chinese authoritarianism – combined with a decline, brought by economic reform, in the regime’s capacity to control the economic life of its citizens and the existence of a space for citizens to criticize the government – would make a transition to democracy highly likely (see, for example, Linz Reference Linz2000).

Yet, at least as of this writing, the CCP has escaped this fate. China demonstrated a viable path for growing out of a planned economy, showing that sequencing economic and political change is possible in the transition from communism. China’s economy expanded beyond expectations, achieving an average annual growth rate, starting in 1998, of 8 to 9 percent. The government weathered the risky period of SOE reform beginning in the late 1990s, laying off some 20 million workers in the process. Although the layoffs triggered scattered protests and demonstrations, the policy did not generate the major turmoil and social unrest predicted by outside observers.

Even among workers who lost their jobs and peasants who lost their land, few joined social movements to protest against the regime (Shi and Lou Reference Shi and Lou2010a). To be sure, the frequency of protests increased, but their incidence was low in comparison to the scale of SOE layoffs.² Among citizens who chose to engage in such regime-challenging political activities, most wanted to remonstrate with rather than oppose the regime.³ Why did Chinese workers behave differently from workers in Poland’s Solidarity, who also faced regime suppression?

The scale of repression did not increase. On the contrary, surveys showed that most Chinese perceived an increase in civil liberties and political freedoms during the period of reform. Instead, the successful political navigation of SOE reform must be attributed to widespread public normative support for the regime. Survey after survey conducted by Western scholars showed levels of popular satisfaction with the regime at around 80 percent (Chen Reference Chen2004; Chu et al. Reference Chu, Diamond, Nathan and Shin2008; Li Reference Li2004; Shi Reference Shi2001).

Some scholars dismiss the ability of surveys to gain true information in China, due to the regime’s suppression. Others argue that information control in China forces people to evaluate their government based on incomplete information. Even if survey responses reflect the true feelings of people at a given time, those feelings are fragile. The removal of information barriers, these scholars believe, would result in a change of attitude toward the government. Some demonize all surveys done in China. He Qinglian, for example, asserts that all joint research projects involving Western and Chinese academics are infiltrated by the State Security Bureau and that all survey data are screened by the bureau. She alleges that scholars involved in survey research in China know the real situation but do not confess because doing so would jeopardize their credibility (He Reference He2004).

Among scholars who recognize regime resilience as genuine, many attribute the phenomenon primarily to the country’s economic performance.⁴ The problem with this argument is that even though economic performance may partially explain why people who benefited from economic reforms supported the government, it cannot explain why the victims of reform, such as laid-off SOE workers, support it. Yet a survey conducted by Wenfang Tang shows that laid-off workers in China’s rust belt supported the regime no less than other respondents did (Tang Reference Tang2005).

Another theory attributes the resilience of the regime to institutional changes. For example, the institutionalization of succession made it possible for the regime to avoid the turmoil that accompanies succession in most authoritarian regimes (Nathan Reference Nathan2003). Although this argument explains the lack of political breakdown at the top during leadership transitions, it does not explain why ordinary people express high levels of trust in their government, nor the apparent contradiction that a majority of people say they want democracy but refrain from engaging in the regime-challenging political actions that would bring it about.

If neither economic performance nor institutional change explains authoritarian resilience, how did the CCP achieve such a miracle? The question has application beyond the single case of China. The histories of Japan, South Korea, Singapore, and Taiwan reveal the existence of authoritarian regimes that enjoyed genuine support from their people for extended periods of time even though their policies caused significant economic difficulties for large segments of their populations. Before their democratic transitions, people living in these countries were found to tolerate their governments more than people living in authoritarian societies outside of East Asia. This led Chalmers Johnson to argue that rulers in East Asia possessed “a particular kind of legitimacy that allows them to be much more experimental and undoctrinaire than the typical authoritarian regime” (Johnson Reference Johnson1982).⁵

To be sure, authoritarianism eventually yielded peacefully to democracy in three of these cases. But this was not necessarily due to a massive collapse of support for the predemocratic regimes. The United States played a critical role – an undeniable outside factor – in supporting democratic transitions in these countries, all of which faced similar security threats that forced them to rely on U.S. support and respond to U.S. concerns. Moreover, the Asian Barometer Survey I (2001–2003) found that the level of popular commitment to democracy in many newly democratized East Asian societies was lower than that on other continents (Chu et al. Reference Chu, Diamond, Nathan and Shin2008). In Southern Europe more than three-quarters of the mass public believed that democracy was preferable to authoritarianism under all circumstances, with overwhelming levels of public support demonstrated by multiple surveys. By contrast, less than half the public in South Korea and Taiwan thought that democracy was the best form of government, and more than half of those countries’ citizens supported possible authoritarian alternatives. When respondents were asked to rate both the level of democracy that was suitable for their political system and the level that it had achieved on a scale of 1 to 10, residents of Taiwan rated their system as being more democratic than they wanted it to be: that is, democratic “supply” was higher than democratic “demand.” Even in Japan, the region’s oldest democracy, citizens showed low enthusiasm for its democratic political system and little desire for more democracy. Yet despite the low levels of popular support for democracy in Asian societies, their democratic regimes remain stable.

The study of political culture offers a way to explain these puzzles found in mainland China, Taiwan, and other Asian countries. To understand people’s responses to political regimes, we need to understand the normative rationality of each population. Yet culture has been misunderstood in recent studies of political behavior. Most scholars conceptualize political culture as simply another kind of resource, arguing that culture can either increase the benefits or reduce the costs associated with various political activities.⁶ This definition removes the essence of political culture – assigning meanings to political action – from political culture studies. I argue instead that culture defines for political actors what constitutes a cost and what constitutes a benefit and in that way shapes the logic underlying an actor’s choices. To understand the puzzles identified in mainland China and Taiwan, we need to study the specifics of political culture in these societies.

Intellectual Origins of Modern Culture Studies

Theories of political culture constitute a response to a fundamental problem of the social sciences, that is, “how a genuinely positive social science would differ from the other positive sciences that had developed before it” (Eckstein Reference Eckstein1996, 471). In the words of Max Weber, “The distinctive (not sole) task of social science [is]: trying to ‘understand’ the meanings, or motivations, that underlie actions.”⁷ Culturally defined meaning plays a critical role in human behavior. Meaning systems not only relate the self to situations that confront actors and in which they are forced to act but also orient people to situations, similar to the way that maps orient travelers.

Following Weber, Talcott Parsons claimed that social action is influenced by the “normative orientation” of actors. When actors choose the means to pursue their ends, their normative orientation plays two critical roles insofar as the situation allows for alternatives: (1) it shapes the ends that actors decide to pursue; and (2) it constrains the means that actors consider available to them to pursue such ends (Parsons Reference Parsons1968 [1937], 44, 74–75). Parsons and Edward Shils argued further that normative orientations play major roles in three distinct but related components of decision making:

1. 1. Cognitive decoding: since the same situation can be cognitively decoded in a variety of ways, actors must interpret the situation in a certain way before choosing an action. People with different cognitive orientations may interpret the same event in fundamentally different ways;

2. 2. Affective⁸ encoding: cognitively decoded situations can have positive or negative emotional meaning for actors, in turn supplying energy for action;

3. 3. Evaluative encoding: alternative actions left open by cognition plus affects must be assigned positive or negative valuations in light of one’s normative system. Such normative systems may sanction or prohibit certain actions. Although in some areas, human actions can be straight responses to feelings, in most cases, cognition plus affect plus evaluation together shape people’s choices of action (Parsons and Shils Reference Parsons and Shils1951, 53).

After decoding the meaning of an event, actors need to decide whether to respond and, if so, in what way. An actor responding to a political event with which he or she disagrees may choose to protest, write to newspaper editors; donate money to a cause supporting his or her view; support a particular political party or candidate in the next election; block the entrances to government buildings; join secret societies or, in an extreme case, become a terrorist. Each actor’s evaluative and affective orientations help him/her decide on the proper goals and the proper means for pursuing such goals.

Those subdimensions of orientation work together to produce a single response to a single stimulus (Parsons Reference Parsons1968). People with different cognitive orientations may interpret the same event in fundamentally different ways. Affective orientation can cause an actor to sample information selectively so as to avoid cognitive dissonance.⁹ Different evaluative orientations will lead actors to different views about the rightness or wrongness of a given act. How people decode an event is shaped by the yardsticks they use to measure the event.

Evaluative and affective orientations may cause actors not only to use different standards to gauge the meaning of an event but also to weigh the importance of various pieces of information differently. As a result, a single piece of information can trump all other conflicting information for a particular actor. From this perspective, differences in actors’ interpretations of situations are often due not to differences in the information they receive, as argued by rational choice theorists, but to differences in the ways in which affective and evaluative orientations lead them to decode the same information. This implies that actors from different cultural backgrounds will differ systematically in the meanings that they derive from the same information. Unfortunately, contemporary students of political culture usually interpret action theory in a different way, one that constitutes a barrier to the advance of studies of political culture.

Current Theories of Political Culture: A Critique

Scholars of political culture agree that people’s orientations should be studied in order to understand political decisions. But they disagree on other issues, such as which specific orientations to focus on, at what level, and how such orientations impact decision making. This section briefly reviews the major approaches to political culture studies and identifies the problems associated with each.¹⁰

The social character approach borrows concepts from anthropological studies of personality and from the study of national character in political science (Adorno et al. Reference Adorno, Frenkel-Brunswik, Levinson and Sanford1950; Almond and Verba Reference Almond and Verba1963; Banfield Reference Banfield1958; Bell Reference Bell1996; Goldhagen Reference Goldhagen1996; Inkeles Reference Inkeles1983; Inkeles and Smith Reference Inkeles and Smith1974; Metzger Reference Metzger1977; Pye Reference Pye1985, Reference Pye1992; Sun et al. Reference Sun, Yuan, Payne and Zhong2005). The conceptualization of culture in this approach is exactly right: culture is defined as a complex of mental properties that assign meanings to social actions and that vary from one society to another. Cultures differ, among other ways, in how people view authority relations, in their members’ commitment to particular religious or ideological views, and in the content and salience of their historical memories.

But the social character approach suffers from three problems. First, although defining culture as a collective property, it normally examines culture empirically at the level of the individual, ignoring the fact that culture may also influence individuals’ choices at a group level, through social interaction (Aronoff Reference Aronoff1988, xv–xvi). In this way the social character approach in practice becomes a study of individual psychology. Second, when the approach was developed in the 1950s, scholars pursuing it did not possess the empirical and statistical tools that would allow them properly to identify culture and test its impacts on people’s behavior. Third, the theory does not clearly specify or test the causal mechanisms attributed to culture (Dickson Reference Dickson1992).

The civic culture school is the most influential approach to the study of political culture, and it explicitly incorporates the orientations proposed by Parsons and Shils. Unfortunately, the conceptualization veered off course when Almond and Verba tried to decompose culture into separate cognitive, affective, and evaluative domains. Rather than viewing these orientations as parts of a unified mental process that jointly determines actors’ responses to situations, The Civic Culture treated them as if they play separate roles in a sequence of mental processes that lead to different political behaviors. This confusion created several problems for political culture studies.

First, the approach moved political culture theory away from its focus on the meaning of social action toward the idea of knowledge of information. Almond and Verba defined cognitive orientation as the knowledge political actors have (1) of their nation and its political system, (2) of the structures and roles of their government, and (3) of the downward flow of policies from government. This definition transforms the concept of cognitive orientation into cognitive ability. The former refers to the predispositions that influence people’s decoding of information; the latter refers to actors’ knowledge and understanding of the political arena. The idea of culture as a distinctive normative rationality disappeared from their conception of political culture.

Second, in studying the impact of culture on political processes, Almond and Verba confined their inquiry to its effects on political participation. They tried to explore how an individual’s psychological orientation influences his or her decisions regarding whether and how to participate in politics. Followers of the civic culture approach tend to ignore the impact of culture on other aspects of politics.

Third, the civic culture school fails to deal properly with the problems of level and unit of analysis. In Parsonian action theory, orientation is conceptualized as an individual phenomenon with social attributes. Almond and Verba used the national society as the unit of analysis and defined political culture as the distribution of alternative orientations in a country. This approach implicitly assumes that the political process in a society is determined by the orientation held by the majority of people and that those of minorities can be neglected. This commits, in effect, the “ecological fallacy” of inferring individuals’ characteristics from the aggregate statistics of the group to which they belong. It causes scholars to neglect the political struggles generated by and around culture.

A third school of political culture studies focuses on the study of symbols, rituals, and myths. It explores how the socially constructed nature of knowledge shapes people’s behavior. This approach defines culture as a set of collective meanings that groups create, share, and symbolically express. Although this definition of culture follows the Weberian concern with the meanings of social action, the way its adherents employ the idea generates several problems.

First, since symbols, rituals, and myths may be interpreted in different ways, it is difficult for scholars to identify clearly a determinate “meaning of action” behind a given cultural icon or to specify how such a symbol assigns meaning to people’s political behavior. The causal linkage between culture and politics offered by this school is usually quite weak. Second, the school tends to confuse social objects that express culture with culture itself; this makes it difficult for scholars to identify the causal mechanisms by which culture influences politics. Third, given that there is usually more than one set of symbols and myths in a society, scholars taking this approach have difficulty choosing which symbols to study when trying to analyze a given society’s culture.

The culture theory approach developed by Mary Douglas and Aaron Wildavsky defines culture as “social bias.” In Risk and Culture, Douglas and Wildavsky argue, “Since an individual cannot look at all directions at once, social life demands organization of bias. People order their universe through social bias. By bringing these biases out into the open, we will understand better which policy differences can be reconciled and which cannot.” To these scholars, the “organization of bias” is socially constructed. Social biases consist of both affective and evaluative components. These biases influence people’s perceptions of risks, which, in turn, influence their choices of action (Douglas Reference Durkheim1982, Reference Douglas1973; Douglas and Wildavsky Reference Douglas and Wildavsky1982; Ellis Reference Ellis1993; Wildavsky Reference Wildavsky1987; Wilson Reference Wilson2000, 252).

Although this approach preserves the idea of meanings of action, the theory deduces culture from people’s behavioral patterns, a method that creates several problems. First, although culture should indeed influence behavioral patterns, and we may therefore infer the nature of cultural attributes from our observation of individual and social behaviors, as a matter of theory culture should not be equated with behavioral patterns. Meanings cannot be assigned to actions by patterns of behavior; they can be assigned only by a culture that lies behind patterns of behavior. This confusion about the relationship between culture and behavior prevents scholars of this school from identifying the causal mechanisms by which culture influences social behavior.

Second and relatedly, if behaviors are used as the measure of culture, it becomes impossible empirically to assess the impact of culture on social behavior because a single measure cannot be used as both an independent and a dependent variable in the same equation. Third, the Douglas-Wildavsky approach believes that different behavioral patterns can be arranged along a single cultural dimension.¹¹ But culture is more properly seen as consisting of different dimensions that govern different kinds of social interactions, as I will try to show in the empirical chapters of this book.

Social constructivism in international relations, the most recent approach in studies of political culture, has brought normative issues back to the study of international politics. The behavioral revolution in the middle of the last century shifted researchers’ attention away from norms, mainly due to methodological reasons – normative and ideational phenomena were notoriously difficult to measure. But persistent puzzles that could not be resolved by the analysis of social-structural and institutional variables prompted international relations scholars to return to the study of norms and ideation in the 1980s.¹² Those scholars made the central claim that the behavior of transnational actors is influenced by ideas (Keohane and Nye Reference Keohane and Nye1977). As states are embedded in dense networks of international social relations, state actors’ perceptions of the world and of their own roles are socialized by international society (Finnemore Reference Finnemore1996). For this reason, scholarship on international regimes in the 1980s and theorizing about social constructivism in the 1990s directed scholars back to the role of norms in international relations (Lapid Reference Lapid and Kratochwil1997).

The focus on norms highlighted two points of importance for the study of culture. First, unlike other kinds of ideas, which may be held privately, norms are collectively held ideas about behavior: they must be shared by a group if they are to function as norms. They are not just subjective but intersubjective; they must be understood as social. Second, while values and ideas may or may not have behavioral implications, the existence of behavioral consequences is part of the definition of norms: norms that have no impact on behavior can be considered aspirant or defunct norms, but not functioning norms.¹³

In trying to explain state behavior, social constructivism explores the normative orientations of subnational units and individuals within the state. Rather than assuming that the normative orientation held by the majority determines the impact of culture, social constructivists examine the political struggles generated by alternative orientations within a society on both the behavior of individuals and the political processes of society. Social constructivists are also more methodologically rigorous than other cultural theorists. They clearly specify causal mechanisms and chains when making arguments about norms, culture, and ideas. They consider the micro foundations on which their theoretical claims rest, and they evaluate various claims in the context of carefully designed research.¹⁴

As evident from the foregoing review, the debate among scholars of political culture focuses on three issues. First, although all culturalists agree that the object of cultural study is people’s orientations, they disagree about the structure of orientations, how different orientations are linked to the outside world, and about what specific orientations they should focus upon. Second, the approaches are divided over the proper level and unit of analysis to be used in cultural studies. Third, culturalists do not concur on the mechanisms through which culture influences the political behavior of individuals and political processes in a society. The following sections seek to address these areas of disagreement and to provide some resolutions.

On Which Orientations Should Students of Culture Focus?

For Weber, the distinctiveness of social action is that human beings “take a deliberate attitude towards the world to lend it significance” (Weber Reference Weber1947, 72). He reminded social scientists to explore not only the instrumental rationality associated with actions but also the normative understanding of their own purposes that are held by individuals and/or groups. Thus, for Weber, the primary objective of cultural study was to explore the normative rationality of social action. To do this, researchers need to focus on orientations that assign “meaning” and “purpose” to the political actions of individuals and to the political processes of society. The most important of these orientations are norms.

Political culture has often been seen as a unified system of mental attributes that includes values, norms, attitudes, and beliefs. But these mental attributes are not all the same. Research by social psychologists suggests that values and norms, on the one hand, and attitudes and beliefs on the other are different psychological dispositions that have separate origins and roles. Values and norms are internal standards that specify proper behavior. Attitudes and beliefs are orientations toward or convictions about particular objects.¹⁵ The norm that a person should respect his or her father, for example, defines the proper way for people to treat their fathers. The attitude of a person toward his father, however, is shaped by the interaction between the norm that person subscribes to and the actual behavior of the father.

A person’s psychological orientation may be thought of as consisting of three layers. At the core lie values. Values are desired or preferred states of the world.¹⁶ The middle layer represents norms, which are shaped by values. Norms assign meaning to actions and guide people’s choices of behavior.¹⁷ The outer layer consists of attitudes toward specific objects produced by the interaction between values, norms, and the behavior of the objects themselves.

Values have some key characteristics:

1. 1. They are mental constructs;

2. 2. They provide actors with criteria for assessing what is right or wrong with various possible choices of ends. In helping people to select the ends of their actions, they define a core element of what actors regard as rational – that is, of people’s subjective rationality;

3. 3. As a primary or core element of thinking, values influence other orientations;

4. 4. Values do not directly produce attitudinal or behavioral outputs but function by shaping norms, which are then applied to features of the actor’s real-world environment to produce attitudes and behaviors. Because the effects of values are mediated, the values themselves must be inferred from attitudes or behavior;

5. 5. Values may be either an individual or a social property;

6. 6. Values are relatively few, thus unspecific in their guidance. The same values can be activated differently in different situations and lead to contradictory behaviors. For this reason, it is difficult, if not impossible, for students of political culture to establish empirical linkages between values and behavior.

Attitudes are important psychological properties, but they have improperly occupied a central position in studies of political culture. Almond and Verba defined political culture as “attitudes toward the political system and its various parts, and attitudes toward the role of the self in the system” (Almond and Verba Reference Almond and Verba1963, 13). In response, Eckstein pointed out that “orientations are not ‘attitudes’: the latter are specific, and the former are general dispositions” (Eckstein Reference Eckstein1988, 789). Located in the outer layer of one’s mental system, attitudes differ from values and norms in several important ways. Attitudes always involve specific objects, but values and norms may not. Values set general principles for how actors should engage with the outside world and thus govern a broad range of issues. Norms have a larger boundary than attitudes, as they usually govern a category of events or a group of people. Attitudes, however, have specific targets. The target of regime-based trust, for example, is one’s government, and the targets of incumbent-based trust are government officials. Attitudes are not only narrower in scope but less emotion-laden than values and norms (Abramson Reference Abramson1983, 35).

Second, attitudes – unlike values and norms, which are shaped primarily through socialization – are shaped by the interaction of norms and the behavior of the attitude’s target. For example, a student’s attitude toward his or her teacher is shaped jointly by social norms that regulate student-teacher relationships and by the behavior of the teacher. Even if a student acquires the social norm requiring respect for a teacher, if a teacher does not behave properly over time, the student may develop a negative attitude toward that teacher.

Third, social structure and political institutions have different impacts on values and norms, on the one hand, and attitudes on the other. Since values and norms interact with the outside world via attitudes, attitudes help cushion values and norms, making them more sustainable in the face of structural and institutional changes. Attitudes, however, directly interact with political objects and are thus more sensitive to environmental changes than values and norms. When the behavior of a political object causes an attitudinal change, the change may not quickly, or ever, lead to a concomitant change in the inner layers of one’s mental system.

Fourth, the causal pathway through which attitudes influence behavior is different from that of norms. Attitudes influence choices through a “logic of expected consequences” that helps actors to assess how various objects in their environment will respond to actions that they take.¹⁸ Norms influence action through a “logic of appropriateness” that tells actors whether certain actions are socially legitimate (Finnemore and Sikkink Reference Finnemore and Sikkink1998; March and Olsen Reference March and Olsen1998). While individuals’ attitudes may influence their understanding of outside stimuli, shape their response toward such events, and constrain the means they choose to pursue their goals, the attitude of a particular individual may not be shared by others in a community. Thus, attitudes cannot legitimate an individual’s choices at the societal level. A man who hates his father, for example, may refuse to ever see him, but his hatred does not legitimate his decision in the eyes of society. Only norms can assign meaning to actions and determine the social legitimacy of actions.

Similarly, policy makers may choose actions based on their expectation of the utilities associated with alternative courses of action. But this is only part of the story. Ethics and virtues can also shape their choices. In many situations, actors may attend less to material benefits than to ethical consequences. Rather than being motivated by instrumental considerations – getting what they want – actors may choose to do something because they believe such behavior is good, desirable, or appropriate.

The comparative advantage of political culture studies, as compared to studies using various versions of the theory of instrumental or rational choice, lies in its focus on the logic of appropriateness and this logic’s influence on behavior. As this logic is defined by values and norms and directly guides people’s behavior through norms, political culture studies should focus on how norms affect political behavior in different ways in different societies.

Norms play a more critical role than either values or attitudes in shaping behavior, yet they are largely ignored by scholars of political culture (with the exception of social constructivists). Parsons defined a norm as a “verbal description of the concrete course of action that is regarded as desirable combined with an injunction to make certain future actions conform to this course” (Parsons Reference Parsons1968, 75). More recently, international relations constructivists defined norms as “shared expectations about appropriate behavior held by a community of actors” (Finnemore Reference Finnemore1996, 22; see also Katzenstein Reference Katzenstein1996, 5; Klotz Reference Klotz1995, 22). Jack Gibbs noted that all definitions of norms have three common attributes. A norm is: (1) a collective evaluation of behavior in terms of what it ought to be; (2) a collective expectation as to what behavior will be; and (3) a set of reactions to behavior, including attempts to apply sanctions or otherwise induce a particular kind of conduct (Gibbs Reference Gibbs1965).

Norms and values both provide actors with criteria for evaluating social and political phenomena. However, two differences between them make norms the better focal point for cultural studies. First, whereas values identify generically desired end-states, norms prescribe or proscribe specific behaviors for specific actors in specific situations (cf. Coleman Reference Coleman1990, 246–47). This gives them a direct impact on behavior that makes them particularly suited to be a focus of the study of political culture.¹⁹

Second, norms are social phenomena. As individual mental constructs, values reflect individual psychology. An individual’s mental construct can be considered a norm only when it is accepted by a group. James Coleman pointed out that “a norm is a property of a social system, not of an actor within it.” Norms are:

This is exactly what we expect culture to be. Political culture exerts its influence on individuals’ choices at two levels rather than one. At the individual level, actors’ normative orientations guide their behavior.²⁰ But their normative orientations alone may not have the social characteristics that can define them as culture. Only when normative orientations are shared within a certain social group can they render shared behavioral guidance in a society. “The concept of a norm,” argued Coleman, “existing at the macro social level and governing the behavior of individuals at a micro social level, provides a convenient device for explaining individual behavior” (Coleman Reference Coleman1990, 241).

In light of the above, political culture will be understood in this book as a body of norms that set standards of appropriate behavior for a group or category of people and that distinguish them from other groups or categories of people. It constitutes socially shared guidance to accepted and expected patterns of conduct for a particular social group. For its carriers, it defines the appropriate means to achieving desired ends – what I call their normative rationality.²¹

This definition departs from previous definitions in several ways. First, it sets aside attitudes, beliefs, and values in order to focus on norms. Because attitudes and beliefs are shaped by the interaction between values acquired in one’s early life and the performance of political objects, they can be influenced by changes in institutions. This leads scholars who define culture as attitudes and beliefs to treat culture as an intervening variable between institutional change and behavioral change (Barry Reference Barry1970; Jackman and Miller Reference Jackman and Miller1996a, Reference Jackman and Miller1996b). The independent causal impact of culture is thereby obscured. This difficulty is avoided if one focuses on values and norms, which change less readily. Since values influence behavior only by way of norms, the most rewarding focus for political culture studies is norms.

Second, I do not define political culture as a separate body of norms specialized for interactions with politics and government. Instead, I assume that the norms that shape people’s social behavior also shape their political behavior. Even when people hold several conflicting values, such as freedom and respect for authority, it is not because they hold one value for the social sphere and another for the political sphere. Rather, the contradiction encompasses both spheres. Artificially dividing the norms that people apply to social issues from those they apply to political issues prevents researchers from properly understanding the mechanisms by which culture influences choices made in the course of political action. For this reason, my empirical strategy for identifying respondents’ norms in survey research avoids the use of words like “politics” and “government.” I ask about general social norms and then explore how people’s commitment to such norms affects their political behavior.

Level and Unit of Analysis Problems in Cultural Studies

One of the major debates in political culture studies centers on the location of culture. Both Weber and Durkheim held that certain obligatory norms exist in societies to guide people’s behavior, but they disagreed on where those obligatory norms are located. The difference comes from their theoretical orientation. Weber was an action theorist and Durkheim an interaction theorist (Eckstein Reference Eckstein1996, 485). Weber argued that all developed societies were built around expectations that actors have of other actors. When they become reliable and fixed, these expectations serve as the nervous system of societies. As a methodological individualist who believed all social phenomena were reducible to individual atoms, Weber claimed that culture or expectations existed only in the minds of individuals. He therefore chose individuals as the unit of analysis in his study of culture.

Durkheim, by contrast, viewed culture as a collective property.²² He believed that there exists a collective consciousness that originates outside the individual and exists independently from individuals. Realizing that collective consciousness may vary across and within societies, he chose both whole societies and sub-populations as units of analysis in his studies of culture (Eckstein Reference Eckstein1996, 482).

The level-of-analysis problem continues to puzzle students of political culture. Most scholars agree that culture is a social phenomenon; the challenge is how to operationalize this collective property. Some empirical scholars have equated culture with the orientations of individuals, using it to explain individual behavior.²³ The problem with this approach is that when culture becomes individualized, it is no longer a collective property; it becomes individual psychology. Other scholars have treated the distribution of individual orientations in a society as an aggregate quality called “political culture” and have used this aggregate to explain both individual behavior and political processes (Hofstede Reference Hofstede2001, 15). The civic culture approach used the differing distributions of certain orientations in different societies to explain the differing degrees of stability of democratic regimes (Hofstede Reference Hofstede2001). Inglehart used the distribution of what he labeled post-materialist values in different societies to explain both the prevalence of unconventional forms of political participation and the political transformation of societies as wholes (Inglehart Reference Inglehart1997). But there are several problems with treating culture as an individual attribute or a societal distribution of individual attributes.

First, these approaches make it difficult, if not impossible, to say how the cultures of different societies differ. How do we know that the culture in country A is different from that in country B? How much difference in the distribution of particular orientations is necessary before we can claim one culture differs from another? Finding the cutoff point is an impossible mission. There is no convincing theoretical argument to specify for researchers what percentage of difference is necessary to delimit cultures. There is no reason to claim, for example, that a society in which 55 percent of the people have a democratic orientation is more likely to transition to a democracy than a society in which 45 percent of the population have such an orientation.

Second, political actors are nested in social environments and regularly interact with other people in their social groups. When culture is defined as the distribution of orientations in a society, researchers are limited to using the dominant orientations to represent a society’s culture (as is done, for example, by Inglehart Reference Inglehart1977, Reference Inglehart1990, Reference Inglehart1997; Inglehart and Welzel Reference Inglehart and Welzel2005). This assumes that each person in society has an equal influence over politics and that only the culture of the majority influences the behavior of individuals and the political processes in a society; in effect, the culture of the minority can influence neither the behavior of others nor the society’s politics as a whole.

Yet the operations of culture are more complicated than this. For example, let us assume that an actor with a high level of political efficacy (the belief in his own ability to influence politics) lives in a country where most people have high efficacy. The culture at both levels prompts him to participate in politics. If he now moves to country B where most people are apathetic, the culture at the social level may lower his level of participation even if his personal sense of efficacy does not change. This suggests that a combination of orientations at different levels can affect people’s behavior. Unfortunately, the current approach to culture allows researchers to concentrate their inquiry on only one level. To understand fully the impacts of culture, researchers need to study them at both levels and to explore how cultural norms at different levels interact with one another to shape political behavior.

Because culture defines what is right and wrong, we may assume that whatever clashing norms exist in a society compete for influence. There is no reason to assume that the norms that stand in a majority at a particular point in time will continue to dominate indefinitely. For example, studies of democratic transition show that the democratic culture held by a minority can significantly influence a society’s political evolution. To understand how cultural norms held by a minority can prevail, we need to study how different norms in a society interact with one another to shape the political behavior of individuals and the political processes of the society. This raises another problem for students of political culture: What unit of analysis should be used in cultural studies?

People are usually embedded in different layers of social environments. An actor may be nested in the normative environments of successively larger geographically based social units such as a village, township, province, and country as well as in social groups such as a church, temple, or classmate association and in functional organizations, such as a factory or company. We cannot assume that the normative environment at the national level always has a greater impact on an individual than the normative environment in his or her local organization or community. Indeed, we should assume that the closer the normative environment to an actor, the greater its impact on the individual. Since political actors are in daily contact with other members of their groups and communities, the collective consciousness of their smaller and more local associations should impact them more strongly than the cultural environment of their society at large.²⁴

When the degree of cultural homogeneity is high in a society, we may identify culture at the national level. But when the degree of heterogeneity is high, researchers should examine the impacts of cultural environments in relevant subunit social groups (potentially including geographic regions, professions, gender groups, age groups, and others) to identify the sources of struggles over culture and the influences of such struggles on behavior. Even to study the impacts of culture at the national level, one may need to analyze the impacts of cultural environments in subnational units.

This analysis suggests that the primary collective unit of analysis for studying a culture’s impacts on the behavior of individuals should be an actor’s immediate social environment. It further implies that at any level of analysis scholars should not assume that the majority’s norms fully determine the cultural environment. Instead, analysts should examine how competing orientations within a community interact with one another and how these interactions influence individuals’ choices.²⁵Finally, political culture scholars need to study the simultaneous impact of cultural norms at different nested levels of the society.

In short, to capture how culture affects individual behavior and political processes, researchers need to identify: (1) how norms influence choices made by actors via individual psychology; (2) how norms exert influence through the local social environment; and (3) how norms at these two levels – individual and community – interact to shape individual behavior and collective political processes.

Causal Mechanisms in Cultural Studies

The effects of cultural norms on social action take place in five steps:

(Step 1) Cultural norms influence how actors interpret the meaning of social action. Facing an outside stimulus such as a governmental decision, an actor must interpret the decision and comprehend its meaning before he or she decides how to respond. This is the process identified by Parsons and Shils (Reference Parsons and Shils1951) as “cognitive decoding.” A psychological device known as “framing” offers perspective and manipulates the relevance and importance of information to influence the judgment of political actors. Cultural norms may influence framing through three mechanisms: they provide standards for evaluating others’ behavior; they allow people to interpret other actors’ motivations; and they provide guidance for how to seek, filter, and arrange information to simplify information processing. When operating under the condition of incomplete information, a person may assume that another actor shares and is following the same norms. Thus a person’s cultural norms may affect how she or he codes others’ behavior.

People may also try to seek out and evaluate additional information to help them decode the meaning of an event. Their decisions (usually unconscious) on what and how much information to seek are shaped by cultural norms. For many, the primary purpose is not to find alternative facts but to seek evidence to validate their predispositions in order to avoid the unpleasant state of cognitive dissonance.

(Step 2) Cultural norms provide standards against which people evaluate the behavior of other actors. As established earlier, because these standards vary, two people may reach different conclusions about the same behavior by a political actor. An observer whose norms are satisfied by the behavior of a political actor will evaluate that behavior as norm-compliant, whereas another observer may evaluate the same behavior as noncompliant.

(Step 3) Cultural norms provide sources of affect toward other actors. Just as different norms require people to use varying standards to evaluate political actors and actions, the same object may give rise to different emotional responses as well. When an event damages a person’s private interests, cultural norms do not automatically authorize the individual to feel resentment or antagonism. Likewise, norms may not mandate gratitude or loyalty in return for a positive action. An individual who receives a benefit from someone else but learns it was acquired by means he or she disapproves of may not feel positive affect toward the person who helped provide the benefit; an individual who is harmed for reasons he or she regards as legitimate similarly may not feel resentment.

(Step 4) Cultural norms define the goals actors can legitimately pursue in a particular circumstance. In a given situation, persons holding different cultural norms may feel authorized to pursue dissimilar goals. For example, when the normative relationship between individuals and authority is defined as reciprocal, people are likely to oppose their government when the government fails to provide them with a desired service. When the citizen-government relationship is normatively modeled on familial relations, the same government failure may not constitute a legitimate reason to withdraw support. Instead, if the government’s intentions are perceived as good (that is, its failure to deliver is not perceived as an act of malice), the norm of hierarchy authorizes people to remonstrate with rather than to oppose their government. Similarly, different cultural norms define different ideal political systems. Although “democracy” has been used to describe all ideal forms of political system, the ways in which people in different societies define democracy are shaped by their cultural norms.

(Step 5) Cultural norms constrain the means actors can legitimately use to pursue their goals. Studies of political culture have focused primarily on the impact of political culture on the frequency of political participation while neglecting the fact that culture also shapes the forms of political action that actors deem permissible. Although all cultural norms allow people to engage in some disruptive political activities (often labeled “unconventional participation”) under certain circumstances, different cultural norms may not define the same thresholds for doing so.

At each of these last four steps, norms exert a causal effect because they are enforced by sanctions at both the individual and the social levels. These sanctions take the form either of rewards for carrying out actions regarded as correct or of punishments for carrying out actions regarded as incorrect (Coleman Reference Coleman1990, 242).²⁶ The rewards and punishments are both internal and external. When a norm is internalized, a sense of shame develops that punishes norm subscribers who either violate the norm or fail to carry out the actions that it prescribes. At the same time, individuals who internalize a norm are accepting the right of others to punish them if they violate that norm.

Studies of self-regulation and self-esteem in social psychology help explain how norms control norm holders’ behavior. E. Tory Higgins found that people have three selves: (1) the “actual self,” which refers to the attributes a person thinks he actually possesses; (2) the “ideal self,” or mental representation of the attributes a person would ideally like to possess; and (3) the “ought self,” the attributes the person believes he should or ought to possess (Higgins Reference Higgins and Berkowitz1989). The ideal self is defined by norms held by people connected to the actor, while the ought self is defined by the actor’s own internalized norms.

According to Higgins, people in their daily lives are motivated to reach a state in which the actual self matches both the ideal self and the ought self. This is because self-regulation, in which people seek to match their performance to the standards held in the community and to their own expectations, has emotional consequences: the closer the match among their three self-concepts, the better people feel about themselves (see also Pelham Reference Pelham1991). Thus, both the ideal self and the ought self serve as behavioral guides. When there is a good match between the actual self and ideal self, people experience feelings of security. This mechanism can be called the external policing system of norms. When there is a good match between the actual self and ought self, people experience feelings of self-esteem. This mechanism can be called the internal policing system of norms.²⁷

Discrepancies among the three selves lead to negative emotions. To reduce the pain caused by negative emotions, people strive to close the gap between the actual self and the other self-guides. Thus, the desire to avoid negative feelings can deter actors from violating the guidance of norms. A person might be deterred from stealing from a store, for example, not necessarily by the probability of being caught, but by the psychological pain associated with norm violation. Since the level of commitment to a given norm varies among actors, the pain created by mismatches among the above three selves varies for different people with respect to a given norm.

The internal policing system is effective to varying degrees in controlling individual behavior. The more deeply an actor internalizes a norm (even unconsciously), the more painful it is to violate that norm. This does not mean that norms always trump material cost-benefit calculations: rather, the strength of a norm determines the threshold at which an actor will abandon that norm for a contradictory material interest. The deeper the internalization, the more likely a person is to be willing to sacrifice material interests to abide by the norm because the emotional consequences of a violation will outweigh the pleasure brought about by material gain.²⁸ Conversely, if a norm is only weakly internalized, the actor may take norms into consideration but allow material interests to outweigh principles.

A given person may internalize multiple norms. Since norms are situation-specific, the choice among relevant norms to apply to a given situation is defined by the internal rank order of norms. For example, two widely accepted norms in China are that one should be loyal to the country and one should be loyal to the family. How will an actor choose between country and family in case of a conflict, for example, if the state requires him to join the army and go to war when his father is dying? For most Chinese today, the rank order of the norms requires loyalty to country to take precedence over loyalty to the family.

Norms are also enforced by the external policing system. James Coleman found that when someone spits on the street or fails to mow their lawn, neighbors or people passing by may confront and blame the offender on behalf of the community. Coleman calls such an act a “heroic sanction” (Coleman Reference Coleman1990, 279). Imposing a heroic sanction requires the sanctioner to pay costs (albeit usually low ones) that will not be fully compensated by the direct personal benefits of the act. And if the norm violator reacts defensively, unpleasantness will ensue for the sanctioner. This might lead us to expect community members to “free ride” – to stand by and let others take on the costs of imposing community sanctions. Despite these risks, the opposite of free riding – an excess of zeal – does occur frequently. Coleman identified the puzzle that “there are many empirical situations in which just the opposite of free-rider activities [seem] to occur, even though the circumstances are those in which free riders would be predicted to abound” (Coleman Reference Coleman1990, 273).

The answer, he suggested, is that sanctioners paradoxically often enjoy implicit support from norm violators. When a norm becomes social, sanctioners can reasonably assume that the norm violator knows that his behavior is wrong, and that he will acknowledge the right of others to impose sanctions on him. A norm violator’s feelings of guilt will reduce the costs incurred by those who impose sanctions on him.

In addition, sanctioners may be rewarded by a community norm that disapproves of shirking and approves of working toward common goals. When people jointly pursue certain outcomes, each has an incentive to reward others for working toward the mutual goal. The heroic sanctioner may be rewarded from two sources. First, demonstrating one’s good qualities to others, by helping the community for example, may generate self-esteem, thereby increasing the internal benefits associated with heroic action. Second, when a norm becomes social, observers may reward heroic sanctioners with esteem – an external reward.²⁹ The combination of these two rewards can outweigh the costs of heroic sanctions.³⁰ Because praise can come only from people with the same normative rationality as the sanctioner, homogeneity of norms in a community increases the likelihood that heroic sanctions will occur.³¹

Norm enforcement is not an isolated one-time event but unfolds through a gradual chain of incremental sanctions that can lead the community collectively to impose more powerful punishments of norm violators, such as social isolation. The process of incremental sanctions often involves three phases. The first is the circulation of information about an event or action. The second is the formation of some consensus about the moral meaning of that event – how it is to be interpreted and which rules are to be applied. The third phase is the implementation of the consensus, the transformation of shared opinions into some form of action. The action can range from individual acts of snubbing to collective decisions to expel (Merry Reference Merry and Black1984, 279).

Gossip plays an important role, helping first to form a consensus that a certain norm is applicable to the action in question and later to forge consensus on a collective decision either to expel the offender and cut off communications with him or her or to support heroic sanctions applied by other individuals (Zablocki Reference Zablocki1980). Böckenholt and van der Heijden found that the decision to comply was influenced not only by one’s personal beliefs about complying with the law but also by the expected evaluative reactions of friends and family, that is, social sanctions (Böckenholt and van der Heijden Reference Böckenholt and Heijden2007). In a review article, Robert Cialdini and others point out, “what’s surprising, given the unambiguity and strength of the evidence, is how little note people take of this potent form of influence” (Cialdini, Petty, and Cacioppo Reference Cialdini, Petty and Cacioppo1981).

The cumulative effect of incremental sanctions on violators can be significant (Coleman Reference Coleman1990, 278). The potential for such unpleasantness can dissuade norm violators from continually engaging in proscribed acts and lead them to modify their behavior. A popular Chinese saying, zhongnu nanfan (“It is risky to offend the community”), alludes to this situation.

Conclusion

Self-regulation and social regulation interact. When culture in a society is homogeneous, sanctions from the internal and external levels push an actor in the same direction, encouraging certain choices and prohibiting others. Since the pressures from the two levels reinforce one another, the effects of self-regulation and social regulation are cumulative: the internal policing system provides actors with guidance on what choices are right, and the external policing system constitutes a feedback loop that rewards prescribed behavior and punishes proscribed behavior. In this way, norms come to have a powerful impact on behavior.

Observing this phenomenon, Lu Xun, the famous Chinese writer of the early twentieth century, wrote in his short story “A Madman’s Diary” that Confucian doctrine (li), the powerful source of China’s enduring traditional norm system, killed many people. Speaking through the voice of the Madman, he claimed that the history of Confucianism could be summarized in one phrase: chiren (eat people). Many of the best-selling novels in early twentieth-century China carried a common theme of battles between young protagonists who engaged in heroic crusades against the traditional norms defined by the dominant, traditional culture. Most of the protagonists initially ignored pressure (heroic sanctions) from their parents, extended family, and acquaintances in order to pursue liberty and happiness. Yet nearly every story ends with the failure of the main character in the face of mounting social pressure (incremental sanctions). Even without the action of formal legal institutions backed up by material force, cultural norms have the powerful mechanism of social sanctions to enforce their rules. The foregoing discussion of how norms work to influence social behavior lays the groundwork for understanding the theory of political culture proposed in the following chapters.