Subjective Well-Being 1 This is an earlier draft of the chapter published in S. Kitayama & D. Cohen (Eds.) Handbook of Cultural Psychology. This draft may not exactly replicate the final published version. Draft: 08/04/2005 Culture and Subjective Well-Being Ed Diener University of Illinois and the Gallup Organization William Tov University of Illinois Running Head: Subjective Well-Being Send reprint requests to: Ed Diener Department of Psychology University of Illinois 603 E. Daniel Street Champaign, IL 61820 USA Email: ediener@cyrus.psych.uiuc.edu

Subjective Well-Being 2 Abstract Subjective well-being (SWB) is composed of people���s evaluations of their lives, including pleasant affect, infrequent unpleasant affect, life satisfaction (LS). We review the research literature concerning the influence of culture on SWB. We argue that some types of well-being, as well as their causes, are consistent across cultures, whereas there are also unique patterns of well-being in societies that are not comparable across cultures. Thus, well-being can be understood to some degree in universal terms, but must also be understood within the framework of each culture. We review the methodological challenges to assessing SWB in different cultures. One important question for future research is the degree to which feelings of well-being lead to the same outcomes in different cultures. The overarching theme of the paper is that there are pancultural experiences of SWB that can be compared across cultures, but that there are also culture-specific patterns that make cultures unique in their experience of well- being.

Subjective Well-Being 3 Introduction ���With great perseverance He meditates, seeking Freedom and happiness.��� -- from The Dhammapada Over two thousand years ago, the Buddha perceived suffering to be the nature of existence. But for him, the attainment of nirvana was not simply a break from this cycle of suffering, it was also a return to true bliss. Although it was not the direct purpose of meditation, happiness was certainly an important consequence, and a critical topic in Buddhist philosophy (Gaskins, 1999). Across time and cultures, generations of people have in their own way reflected upon the question of happiness. As long as it has been pondered, it may come as a surprise that the scientific study of happiness, or subjective well-being (SWB E. Diener, 1984) has advanced only recently. One of the challenges has been defining happiness in a way that enables it to be measured. Given that conceptions of happiness may vary across different societies, a number of questions arises regarding how culture influences the idea and experience of happiness. Does the structure and content of SWB differ? Do certain cultures emphasize some components more than others? Are the correlates and causes of happiness similar across cultures? Do people react differently to the experience of well-being (e.g., when they feel pleasant affect)? As it has been studied over the past two decades, SWB involves frequent pleasant emotion, infrequent unpleasant emotion, and life satisfaction (LS). The first two components are affective the last is a cognitive evaluation. These three components are not the only elements of SWB. Happiness also can be said to consist of other dimensions such as meaning and purpose in life. However, in this review we focus on LS, pleasant affect, and unpleasant affect in part because these constructs have been researched more frequently across cultures. Furthermore,

Subjective Well-Being 4 these components of SWB are major focal points that allow for a certain degree of precision in measuring the fuzzier, folk concept of happiness. Why Study SWB Across Cultures? The cross-cultural study of SWB is one indicator of the quality of life in a society. It was once considered taboo to suggest that societies could be evaluated at all (Shweder, 2000). To appraise any aspect of a culture was to ignore its worth and integrity. However, this extreme form of cultural relativism has given way to the view that, though one must be careful in comparing and evaluating societies, they may differ in variables such as health and satisfaction that are desirable in most cultures. It is true that some indicators of life quality may impose values about the good life that are not shared by all people. However, even if SWB is internally framed with respect to each culture, societies could still be evaluated in terms of how well they succeed according to these internal criteria. Culture and SWB research can also shed light on basic emotional processes. In measuring SWB across various societies, researchers have confronted issues regarding the universality of emotions, and how the representation of emotions in memory are influenced by cultural norms. The field can also add to our understanding of culture. For example, how do cultures differ in their socialization of pleasant and unpleasant affect, and how do emotions contribute to the reinforcement of cultural values and practices? These questions reflect a cultural psychological perspective. Thus, the topic is of both applied and theoretical importance. History of this Field of Inquiry Anthropologists adopted cultural relativity as a way of avoiding a Western, ethnocentric bias in observing other cultures. They made the important observation that values and practices might vary across cultures, but this need not imply that some cultures were necessarily better

Subjective Well-Being 5 than others. In particular, we should avoid judging other cultures by the standards of our own. However, taken to extremes, cultural relativism would prevent one from saying that Nazi Germany, or Cambodia under the Khmer Rouge, were in many respects undesirable cultures (Edgerton, 1992). This level of extreme value relativity would make cultural psychology irrelevant to public discourse. According to Edgerton (1992), not all practices in a culture are adaptive some may even be harmful. He defined maladaptive cultures as those in which there was rampant dissatisfaction or impaired physical and mental health. Thus, there are certain criteria by which we can judge the success of a culture. As one such criterion, SWB is important because a society functions poorly when a majority of its people are discontent and depressed. It should be noted that very little quantitative work has examined the well-being of small cultures (e.g., Biswas-Diener, Vitters��, & Diener, 2003), although a number of international surveys of SWB in modern nations have been conducted (e.g., Cantril, 1965 Inglehart, 1990 see Table 1). Only recently has research examined the structure and causes of SWB in different cultures. In 1995, for example, E. Diener and M.L. Diener found that self-esteem correlated more strongly with LS in individualist than in collectivist cultures, and that financial satisfaction more strongly predicted LS in poor than in rich nations. Since then, there has been a rapid growth in the field of culture and well-being, and both universal and unique correlates of SWB have been documented. We foresee further growth in this research area in the decade to come. General Approaches to Cross-Cultural Comparisons of SWB The comparisons that researchers make across cultures are guided by their assumptions about the interplay between culture and SWB. We review some of these approaches below.

Subjective Well-Being 6 Dimensional Approach Some theorists hold that the causes of well-being are fundamentally the same for all people. Ryff and Singer (1998) posited that purpose in life, quality relationships, self-regard, and a sense of mastery were universal features of well-being. Self-determination theorists (Deci & Ryan, 1985 Ryan & Deci, 2000) maintain that well-being hinges on the fulfillment of innate psychological needs such as autonomy, competence, and relatedness. If these sources of well- being are universal, they provide dimensions along which we can compare societies. Cultures should differ in SWB to the extent that they provide people with different levels autonomy, meaning, and relationships. A related perspective is the universalist position on emotions. Drawing on diverse findings, some researchers propose that there are discrete, basic emotions that appear in all cultures (Ekman & Friesen, 1971 Izard & Malatesta, 1987 Plutchik, 1980 Tomkins, 1962, 1963). For example, facial expressions of anger, sadness, and joy appear early in infancy (Izard & Malatesta, 1987), and are easily recognized in many different cultures (Ekman & Friesen, 1971 Ekman et al., 1987). Facial expressions of laughing and crying among congenitally blind infants (Thompson, 1941) suggest that there may be genetic programs directing the expression of emotions. The possibility of biologically based, basic emotions is important for it implies that we can compare people across societies on these emotions (however, see Ortony & Turner, 1990 for a critique of the basic emotions concept). Uniqueness Approach In contrast to the universalist approach, some ethnographers emphasize emotions as social constructions. According to these researchers, the very concept of emotion may differ across cultures. Lutz (1988) noted that Western ethnopsychologies often view emotions as

Subjective Well-Being 7 hidden and private. In contrast, her work in Micronesia revealed that Ifalukian concepts of emotions were more public and relational. Cultures may also differ in their labeling of specific feelings. For example, according to Wierzbicka (1986) there is no word for disgust in Polish. Extreme versions of the uniqueness approach hold that emotions are purely a Western idea, and that internal experiences can be represented in countless ways across cultures. More moderate formulations, on the other hand, maintain that biologically based emotions may be universal, but that culture can significantly alter their development and labelling. Thus, although sadness is often considered a basic emotion with recognizable antecedents, the Tahitians do not appear to have such a label for it (Levy, 1982). Instead, they often refer to feelings of sickness or exhaustion, for which the causes are nonspecific. Although the uniqueness approach does not preclude the possibility of making comparisons across cultures (e.g., Wierzbicka, 1986), it takes as its starting point the culturally patterned subtleties of emotional experience. Identity Approach Another perspective on universality is that regardless of the specific elements, all cultures enjoy identical levels of SWB. Cultures may differ in their values and in the needs they fulfill, but people eventually adapt leading all societies to be relatively happy. The identity approach likens well-being to a ���hedonic treadmill��� upon which people run but never change position. Only when cultures are severely disrupted or experiencing trauma (e.g., warfare or famine) will adaptation be impossible, resulting in widespread unhappiness. This position may sound absurd, but in Table 2, diverse groups appear enjoy somewhat comparable levels of LS. For instance, the Amish, Inughuit, and Maasai all report LS that is not significantly different from the richest Americans, suggesting that material luxury is not necessary for well-being. All these groups may be meeting needs such as social relationships, which are critical for SWB. Thus, there may

Subjective Well-Being 8 be important conditions for happiness that are met in non-industrial societies such as the Maasai. In contrast, the LS of the homeless indicates that not all groups are happy, and that people do not fully adapt to all conditions. The Middle Path In this chapter, we take a middle path. We argue that there are some universals, such as the tendency for people to be slightly happy unless they are exposed to harsh conditions. Some variables influence SWB in all cultures such as temperament and positive relationships. There may also be common goals, such as the need for respect, that characterize people in all cultures. Furthermore, because cultural influences often permeate national boundaries, cultures are not completely independent of one another. However, each culture also retains unique qualities and should not be compared with others in a careless way. Not all comparisons of SWB are meaningful because the value placed on certain subjective states and the labels for them, often differ. The patterning of well-being may also vary across cultures, making it dangerous to compare variables at a high level of abstraction. Thus, although comparisons are possible, they should only be made with due care to the unique factors present in various societies. Cultural differences in SWB can be likened to differences between individuals. People can be compared on certain universal features such as height and weight. They can also be compared on factors such as health, but health is made up of many lower-order concepts that may relate to each other differently across individuals. Although societies can be compared on longevity, the patterns of illness differ across cultures. In a similar way, cultures can be compared on SWB, but there are also unique facets of well-being in each society that are best captured by specific descriptions of the local culture.

Subjective Well-Being 9 Overview We will cover several major topics in culture and SWB research. We begin with the issue of patterning and structure, examining how the elements of SWB cohere across societies. Next, we consider whether or not cultures differ in mean levels of SWB where the structures can be compared, and what factors might contribute to these differences. We then review various correlates and causes of SWB, showing both similarities and differences in cultural recipes for happiness. Following this discussion, we ask whether SWB leads to the same outcomes in different cultures, or whether there are unique effects that depend on the role of emotions in a culture. Finally, we assess the various challenges involved in measuring SWB across cultures, and the impact that measurement artifacts may have on the findings. Patterning and Structure The validity of cross-cultural comparisons of SWB depends on how it is structured in different societies. If there are both universal and culture-specific emotions, are aggregates like pleasant and unpleasant affect applicable in all cultures? Is the concept of LS understood by people in all societies? Also, do the three components of SWB relate to each other similarly across cultures? We review the research bearing on these issues below. Levels of Analyses As discussed earlier, the existence of universal emotions has been debated for some time. Researchers have used a number of methodologies to answer the question of universality including ethnography, facial expression recognition, and emotion taxonomies. After conducting cross-cultural research on facial expression recognition, Ekman and his colleagues (Ekman & Friesen, 1971 Ekman et al., 1987) suggested that happiness, anger, fear, sadness, and disgust

Subjective Well-Being 10 were universal. However, there are also emotions that appear in some cultures, but not others. Some appear to be labeling of specific situation-outcome pairings in relation to feelings. In Japan, for example, the term kanashii refers specifically to sadness arising from personal loss (Mesquita & Fridja, 1992). Other indigenous emotions seem to be complex blends such as aviman in India, which has been described as ���prideful, loving anger��� (Scollon, Diener, Oishi, & Biswas-Diener, 2004). According to Mesquita, Fridja, and Scherer (1997), the debate over universality has hindered culture and emotion research by focusing on the mere presence of certain emotions in a culture rather than on how emotions are ���practiced.��� They argued that emotional experience is a process that includes appraisal of a situation, physiological reactions, overt behaviors and other components. What distinguishes one emotion from another is the pattern of components. At a general level, universal patterns of emotional experience may exist, due to innate, neurophysiological programs. For example, joy may inherently feel pleasant and evoke the urge to laugh or smile. However, at the level of specific components, cultural differences may abound. The type of events that elicit joy, or attempts to regulate it may vary across societies. The perspective provided by Mesquita et al. resonates with several lines of research on well-being. In assessing the cross-cultural applicability of pleasant and unpleasant affect, SWB researchers have not only been interested in which emotions are present, but also in how frequently they are experienced, how they are patterned, and how norms can shape the structure and composition of pleasant and unpleasant affect. In short, the field of culture and SWB has been concerned as much with the ecology or practice of emotions (Mesquita et al., 1997), as it has with the comparability of SWB across cultures. We will see that the distinction between

Subjective Well-Being 11 pleasant and unpleasant affect can be made at a general level, and that there are both similarities and differences in the specific aspects of these emotions. Structural Evidence In an early study, Watson, Clark, and Tellegen (1984) found that the mood structure of Japanese participants formed two factors identifiable as positive and negative affect. This two- factor structure was very similar to that of American participants. Hierarchical cluster analyses of emotion words from the U.S., Italy, and China also revealed superordinate groupings of positive and negative emotions (Shaver, Wu, Schwartz, 1992). Pleasant and unpleasant emotion clusters were also observed in experience sampling data provided by Japanese, Indian, and two American samples (Scollon et al., 2004). Moreover, indigenous emotions that were included in the Japanese and Indian samples did not form separate clusters, but grouped together with the pleasant and unpleasant emotions. M.L. Diener, Fujita, Kim-Prieto, and E. Diener (2004) studied the frequency of twelve emotions and found that they formed positive and negative clusters in seven regions of the world (Africa, Latin America, East Asia, Southeast Asia, West Asia, Eastern Europe, and Western Europe). Moreover, in virtually all of these regions, a core group of emotions consistently loaded onto either positive or negative clusters. That is, positive emotions consisted of pleasant, cheerful, and happy, whereas negative emotions consisted of unpleasant, sad, and angry. Similarly, Shaver et al. (1992) found that one positive (joy) and three negative emotions (anger, sadness, and fear) formed basic level categories in all three cultures they studied. Finally, Kuppens, Ceulemans, Timmerman, Diener & Kim-Prieto (in press) found that positive affect and negative affect emerged as strong universal intracultural dimensions, as well as smaller but significant nation-level dimensions on which nations could be discriminated.

Subjective Well-Being 12 Thus, when speaking of emotion aggregrates, there is compelling evidence that pleasant and unpleasant affect are perceived in all cultures. There is also support for the universality of particular emotions such as joy, anger, and sadness. However, cultural differences may arise regarding more specific emotions. For instance, outside of the core emotions, M.L. Diener et al. (2004) observed differences in how other emotions clustered. Pride clustered with positive emotions in Latin America, Western Europe, and East Asia, but with the negative emotions in Africa, Southeast Asia, Eastern Europe, and West Asia. Pride also aligned with the negative emotions among smaller samples in India and Italy (Scollon et al., 2004 Shaver et al., 1992). These findings should be interpreted cautiously. The simple fact that pride clusters with negative emotions in a culture does not necessarily mean that it is experienced as a negative emotion. In the case of M.L. Diener et al.���s data, the cluster analyses were based on the frequency of experience and included weights for means, standard deviations, and correlations���any of which could have affected how emotion terms clustered. In those regions where pride was experienced less frequently, it clustered with the negative emotions, which were generally experienced less often than the positive emotions. In contrast, worry and stress clustered with the positive emotions in Western Europe and East Asia primarily because they were frequently experienced in those areas. Thus, emotional experience may be universal in some ways, but culturally varied in other ways. Recently, Kuppens, Ceulemans, Timmerman, Heymans, Diener & Kim-Prieto (in press) found that although positive affect and negative affect emerged as strong universal intracultural dimensions, there were also smaller but significant nation-level dimensions of emotional experience on which nations could be discriminated.

Subjective Well-Being 13 Differences in the frequency of emotions may be related to cultural norms. For example, cultural norms might make some situations more common than others. Thus, the American cultural environment might afford more opportunities for self-enhancement (and the experience of pride), whereas the Japanese cultural environment might be more conducive to self-criticism (Kitayama, Markus, Matsumoto, & Norasakkunkit, 1997). According to Markus and Kitayama (1994), normative social behavior and cultural models of the self might also shape the desirability of certain emotions. In individualist cultures, pride is an enjoyable emotion that highlights individual achievement as well as success in meeting the cultural goals of autonomy and independence. However, in collectivist cultures, emotions resulting from sympathy and humility may feel good because they are consistent with the cultural goals of interdependence. Emotions that conflict with these norms may be de-emphasized and less frequently experienced. Thus, pride may not be as valued in some collectivist Asian cultures because it is self-focusing and separates the individual from the group (Kitayama & Markus, 2000 Markus & Kitayama, 1994 Scollon et al., 2004). In a similar way, the Oriyas in India devalue anger because it is regarded as socially destructive (Menon & Shweder, 1994). On the other hand, shame1 is viewed as a good emotion for women to have because it is integral to sustaining the patriarchal order of society. The Oriya case draws attention to intracultural variation in emotion norms. That is, norms may not apply or be uniformly perceived across all individuals within a culture. Eid and Diener (2001) investigated this issue by examining the desirability and appropriateness of pleasant and unpleasant affect in the U.S., Australia, China, and Taiwan. They found that norms for pleasant emotions (e.g., joy, affection, pride, and contentment) were more heterogeneous in China and Taiwan than in the U.S. and Australia. For instance, the vast majority (83%) of

Subjective Well-Being 14 Australians and Americans regarded all four pleasant emotions as appropriate. In contrast, only 9% of Chinese and 32% of Taiwanese felt this way. A majority of the Taiwanese (57%) had mixed feelings about pride, although joy, affection, and contentment were appropriate. A plurality of the Chinese (32%) felt that joy and affection were appropriate, but that pride was clearly inappropriate. Another class of individuals found only among the Chinese (16%), regarded all pleasant emotions as inappropriate. These findings suggest that culture may influence emotion norms in two ways. First, cultures may foster unique normative patterns, as was observed in the Chinese sample. Second, some patterns may be pancultural, but their relative frequency within cultures may differ. All pleasant emotions are clearly favored in the U.S. and Australia. The ambivalence towards pride in China and Taiwan is consistent with previous research on collectivist Asian cultures. However, the relation between emotion norms and emotional experience may not always be direct. Recent work by Tsai, Knutson, and Fung (2005) suggests that the emotions that people value (ideal affect) are not necessarily the ones they experience most frequently (real affect), although the correlations are moderate. These researchers found that although cultural values predicted the preference for high versus low arousal pleasant emotions, the reported frequency of these emotions was better predicted by personality traits. Furthermore, norms may influence some emotions more than others. M.L. Diener et al. (2004) found that the correlation between the appropriateness and frequency of an emotion was larger for ���secondary��� emotions like pride, guilt, gratitude, and jealousy, than it was for the core emotions. That is, norms appear to predict the experience of secondary emotions more than the experience of core emotions. Indeed, the main cultural differences in structure were due to how the secondary emotions clustered, and the various geopolitical regions diverged most in the frequency of these emotions. For example,

Subjective Well-Being 15 people from Southeast Asia reported more frequent experience of guilt and shame, whereas those from Latin America registered more pride than people from other areas. Also, norms for pride and guilt were more variable across cultures than norms for other emotions (Eid & Diener, 2001). Differences in the experience of peripheral emotions such as pride may reflect cultural ideologies regarding attribution styles, such as whether success should be attributed to the self or to the situation (Heine, Lehman, Markus, & Kitayama, 1999). In contrast, a core emotion like happiness is much broader and may tend to follow from outcomes that are considered good in each culture, so that valuing general happiness is likely to be more common across cultures. In addition to emotions, there is also support for similarity in the structure of LS across cultures. Vitters��, R��ysamb, and Diener (2002) carried out confirmatory factor analyses on the five items of the Satisfaction With Life Scale (SWLS E. Diener, Emmons, Larsen, & Griffin, 1985) and found that a one-factor model fit the data reasonably well in forty-one nations. In all nations, the comparative fit index was above .90. This finding suggests that the SWLS measures a single construct and that the concept of ���life satisfaction��� may be similarly understood across a wide range of cultures. That is not to say that the criteria for LS are universal, but rather that people in a number of diverse cultures appear to react to queries about LS in a consistent way. The Relation Between Emotions and Life Satisfaction (LS) Although the structure of emotions is somewhat consistent across cultures, and the items of the SWLS also seem to cohere reliably, the relation between emotions and LS may vary across cultures (Schimmack, Radhakrishnan, Oishi, Dzokoto, & Ahadi, 2002 Suh, Diener, Oishi, & Triandis, 1998). Suh et al. (1998) examined the relation between LS and affect balance (the difference in frequency of pleasant and unpleasant affect). They found that LS and affect balance correlated positively across forty nations thus, experiencing more pleasant than