How Much Do Sources of Happiness Vary Across Countries? A Review of the Empirical Literature

This article presents a review of empirical research exploring cross-national differences in the correlates of subjective well-being (SWB). I start by giving an overview of the concept of SWB across psychological, sociological, and economic literature. Measures of SWB have good cross-cultural validity, yet there is currently little consensus regarding the cultural universality of the definition of happiness. An overview of existing empirical literature points toward robust cross-national differences in mean levels of SWB that are associated with national differences in wealth and other socioeconomic, political, and cultural factors. The degree to which individual-level variables are associated with SWB is also subject to cross-national variations. Many individuals’ characteristics contribute to happiness to the extent that they are beneficial, socially desirable, and aspired to in a particular sociocultural context. These results are discussed in light of two theoretical approaches (institutional and fit hypotheses). Directions for future research are proposed.


Introduction
The question of the nature and causes of human happiness has been a source of major interest in philosophy since antiquity, with different schools of thought offering different answers (McMahon 2006). In contrast to philosophy, it is only since the middle of the twentieth century that social scientists have discovered this research field. Since then, research on happiness, or subjective well-being (SWB), as this concept is often referred to in the social sciences, has been very fruitful, and attracted scholars from diverse disciplines, including psychology, sociology, and economics, who have published more than 80,000 articles on that topic in total over the last 30 years (according to the Web of Science).
Most of this work has been dedicated to revealing factors that contribute to a higher vs. a lower SWB. Some studies have focused on individual characteristics, such as personality traits, values, and beliefs or life circumstances. Others explored contextual or country-level predictors, such as cross-national differences in social policies, socioeconomic conditions and culture. Finally, a third group of studies combined the investigation of individual-and country-level factors exploring how they come together to affect individuals' happiness. Studies exploring individuallevel predictors probably represent the largest portion of the literature and have been the target of several review papers in psychology Lyubomirsky et al., 2005a, b). Studies exploring national-level predictors have been the major focus of sociological and economic literature on SWB and have also been reviewed before (Di Tella et al. 2003;Frey and Stutzer 2005). Yet, there have been no reviews of the third stream of empirical research, the one that combines the examination of individual-and country-level predictors of SWB. Therefore, the present review focuses on this third group of studies. Nevertheless, to put these studies into context, I also include an overview of empirical studies that examined individual-and country-level predictors independently of each other. I will start by discussing existing conceptualizations, definitions, and measurements of the concept of SWB, including the question of its cross-cultural validity (see also Cieciuch et al. 2019). I will then proceed to present an overview of empirical findings pertaining to individual-and country-level predictors, as well as their joint effects on SWB. Finally, I will discuss the results in the light of existing theories and outline directions for future research.

The Concept of Subjective Well-being
Probably due to the interdisciplinary nature of SWB research (it is an active area of research across diverse disciplines, mainly including psychology, sociology, and economics), there are currently dozens of concepts that are studied under the umbrella term of subjective well-being: life satisfaction, avowed happiness (Wilson 1967), subjective happiness (Lyubomirsky and Lepper 1999), authentic happiness (Seligman 2002), affect balance (Bradburn 1969), experienced utility, and objective happiness (Kahneman 2000), to name but a few.
Even though research on SWB is largely interdisciplinary in nature, it has long been dominated by psychologists. This is not surprising, as SWB is an individuallevel concept that deals with differences in individuals' subjective perception and evaluation of reality; and the way people see reality is the core business of psychology, a science of the human mind. In contrast, despite the rising level of interest in the concept of happiness in sociology and economics, it is still rarely discussed in the respective textbooks and journals (Veenhoven 2008). This relative lack of interest has been explained by these disciplines' main interest in understanding societal problems rooted in objective reality or conditions, rather than understanding how people feel about reality and about their conditions, and why (Veenhoven 2008). Accordingly, the most elaborate conceptualization of happiness stems from psychological literature and will be the focus in the present review. At the same time, as I explore an interplay of individual and contextual (country-level) predictors of happiness, and the latter have been almost exclusively the focus of sociology and economics (with the exception of cross-cultural research), I will provide an overview of how happiness is conceptualized in these disciplines as well.

The Concept of SWB in Psychology
The most widely accepted conceptualization of SWB in psychology was proposed in the 1980s by Ed Diener, who also coined this term. SWB was defined as "a general area of scientific interest, rather than a single specific construct" (Diener et al. 1999, p. 277), with life satisfaction and affect representing the most widely studied concepts within the field of subjective well-being. Life satisfaction represents a cognitive aspect of SWB; it refers to individuals' evaluation of how close their life is to what constitutes an ideal life in their view. Affect represents the emotional component of SWB. It is typically operationalized as frequency and intensity of positive and negative emotions experienced during the last several weeks (Diener 1984;Kahneman and Deaton 2010).
Both the cognitive and affective components of SWB are subjective and depend on individuals' personal evaluation and judgment. Also, they are both hedonic, rather than eudaimonic, in nature. Current psychological literature on well-being distinguishes between its hedonic and eudaimonic components (Ryan and Deci 2001). Eudaimonic well-being is grounded in the philosophy of Aristotle, who suggested that a life of virtue and realization of one's potential constitutes true happiness (eudaimonia). The Aristotelian conception of happiness has been further developed by humanistic and positive psychologists (Ryff 1989;Seligman 2002). Instruments measuring eudaimonic well-being typically include such concepts as meaning and purpose, personal growth and development, rather than mere satisfaction. For example, Ryff (1989) considers happiness (psychological well-being) as consisting of six dimensions: autonomy, personal growth, mastery, positive relatedness, life purpose, and self-acceptance.
In contrast, hedonic-both cognitive and affective-well-being emphasizes individuals' global evaluation and feeling regarding their lives. SWB is retrospective in nature, meaning that measures of life satisfaction or affect represent memorybased reports. Although such retrospective reports of happiness do not completely overlap with moment-to-moment happiness (e. g., measures of happiness collected during different activities and times of the day) (Fredrickson 2000;Redelmeier and Kahneman 1996), they are important as they often represent the basis of individuals' decision making. For example, it is the retrospective, stored-in-memory judgment of happiness one experienced in a particular restaurant that is likely to affect one's decision to visit this restaurant again (Stavrova 2014).

The Concept of SWB in Sociology and Economics
Sociological research has started showing an interest in the studies of happiness since the 1980s. Similar to psychology, a variety of terms referring to happiness have been used by sociologists as well. While some of these terms overlap with the ones used by psychologists, others are new and denote the particularities of the use of the concept in sociology in general. For example, while the term "happiness" and "subjective well-being" are used by both psychologists and sociologists as an umbrella term, sociologists see it as a synonym of "quality of life" and "individual and social welfare" (Veenhoven 2012). Regardless of these differences, the definition of happiness adopted in sociology is quite similar to the psychological one: happiness is defined in sociology as "the degree to which an individual judges the overall quality of his/her own life-as a whole favorably" (Veenhoven 1984). Yet, in contrast to psychologists, who differentiate between the cognitive and the affective components, sociologists see life satisfaction and happiness as a combination of a "cognitive comparison with standards of the good life (contentment) and affective information from how one feels most of the time (hedonic level of affect)" (Veenhoven 2008). In a tradition of social constructionist theories (Berger and Luckman 1966), sociologists share a constructionist view of happiness, assuming that happiness represents individuals' construction of reality, which is largely determined by comparative thinking: people are happy as long as their life meets their expectations and is not worse than a neighbor's life (Veenhoven 2008). As the main measurement instrument, sociological research typically uses a single happiness item that directly asks individuals how happy they are with their lives as a whole.
Interestingly, this very measure of happiness is also quite often used in economics (Frey and Stutzer 2002a;. The concept itself is however considered in a quite different light. In contrast to the social constructionist view shared by sociologists (and psychologists, although they would not call it that), economists use happiness as a tool to measure the desirability of objective reality, or in other words, for economists, happiness represents a measure of utility (Frey and Stutzer 2014). Economists' interest in happiness is just a result of their attempts to find the best operationalization of the utility that individuals derive from using different products, institutional or social policy changes (Di Tella et al. 2003;Di Tella and MacCulloch 2006;Frey and Stutzer 2002a, b).

SWB Measurement
A variety of reliable and valid instruments have been designed to assess SWB and its different components. One of the most widely known instruments used to measure life satisfaction in psychology and sociology is the Satisfaction With Life Scale (Diener et al. 1985). It consists of five items and has a high degree of reliability and external validity (Pavot and Diener 1993). It shows a high level of self-informant consistency (Schneider and Schimmack 2010) and converges with objective measures of satisfaction (e. g. smiling behavior, use of negative emotion words in texts) (Liu et al. 2015;Settanni and Marengo 2015). It has also been shown to have high temporal stability (0.54 within 4 years; Diener et al. 2013), showing that-in contrast to what some studies have suggested (Schwarz 1990, but see Yap et al. 2016)-satisfaction with life is not a momentary assessment that is potentially subject to framing, item order, or weather effects Pavot and Diener 2008).
Most large-scale cross-national studies typically rely on a single-item scale of life satisfaction, asking participants to indicate how satisfied they are with their lives overall (a 10-point Likert scale is usually used). This single-item measure of life satisfaction has been shown to strongly correlate with the Satisfaction With Life Scale and to have good external validity as well (Cheung and Lucas 2014;Jovanović 2016). Due to its brief form and ease of use, it has been included in many large-scale cross-national surveys. As a result, most cross-national findings on life satisfaction so far are based on the data using this measure.
Regarding the affective component of SWB, a number of measures have been developed as well, with Bradburn's Affect Balance scale (Bradburn 1969) and Positive and Negative Affect Schedule (PANAS; Watson et al. 1988) being the scales used most often. These measures assess the frequency of experiencing a series of positive and negative emotions over the past 4 weeks.
Cognitive (life satisfaction) and affective measures of SWB show moderate to strong correlations with each other ), but do demonstrate distinct patterns of correlations with other constructs. For example, while higher income is positively related to life satisfaction, it does not predict more positive emotions (Kahneman and Deaton 2010). On the other hand, fulfillment of psychological needs, such as social affiliation, respect, and autonomy, is a stronger predictor of affect than it is of life satisfaction (Diener et al. 2010).
Importantly, existing SWB measures show a good level of cross-cultural validity. For example, a couple of studies showed that instruments developed to measure SWB in the West show acceptable levels of reliability, convergent and discriminant validity in other cultural contexts (Whisman and Judd 2016; see also Cieciuch et al. 2019). For example, the Satisfaction with Life Scale, which was developed for use in American samples, showed the same factor structure in samples of Serbian adolescents (Jovanović 2016), elderly Mexican people (López-Ortega et al. 2016), Korean elementary school children (Lim 2015), as well as adolescents and young adults in Portugal (Silva et al. 2015) and Italy (Di Fabio and Gori 2016). Also, reports of life satisfaction converge with more objective validity indicators, such as reports of positive vs. negative events, not only in American but in other cultures as well (Balatsky and Diener 1993).
Somewhat less consensus exists with respect to the extent of cross-cultural differences in folks theories of happiness, that is, lay beliefs about the nature and sources of happiness. On the one hand, a stream of literature in cross-cultural psychology tends to emphasize cross-cultural differences in lay (and even in dictionary, Oishi et al. 2013) definitions of happiness (Uchida and Kitayama 2009). These studies showed that North American vs. East Asian and East European participants think about happiness differently, with the former seeing it exclusively positively, and the later having a more ambivalent attitude towards happiness, including its negative aspects, such as the fact that it does not last long and may even cause envy on the part of others (Joshanloo et al. 2014). At the same time, cross-cultural commonalities in lay beliefs about happiness have been shown as well. For example, a recent exploration of lay definitions of happiness showed a certain degree of similarity, with an emphasis on inner harmony being considered central to happiness across the twelve countries studied, including the U.S., India, as well as some Eastern European and Latin American countries (Delle Fave et al. 2016).
While cognitive and affective components of SWB might be equally important, disciplines differ in what component they mostly focus on, with psychological research being interested in both, while studies conducted in sociology and economics are mostly restricted to life satisfaction or overall happiness (forgoing the distinction between cognitive and affective components altogether). As a result, most largescale cross-national datasets include a measure of life satisfaction ("Taking all things together, how satisfied you are with your life these days?") and overall happiness ("Taking all things together, would you say you are very happy, quite happy, not very happy, not at all happy?"), but no measures of affective well-being, and most findings on cross-cultural variability in mean levels and correlates of SWB, are restricted to life satisfaction and overall happiness.

Individual Predictors of SWB
At the individual level, personality has been shown to be one of the strongest predictors of happiness (Lyubomirsky, Sheldon and Schkade 2005). In fact, differences in K stable individual dispositions are assumed to account for about half of the variance in SWB (Lyubomirsky et al., 2005b). Among the Big Five personality traits, neuroticism has been shown to be the strongest negative predictor of life satisfaction (Steel et al. 2008). Neurotic individuals are moody, experience frequent mood swings, get upset and stressed out easily, and respond to negative stimuli with greater anxiety. It comes as no surprise that neurotic individuals tend to report lower levels of life satisfaction and happiness.
Besides neuroticism, extraversion and conscientiousness typically show an independent positive association with SWB (Hayes and Joseph 2003;Lucas et al. 2000;Steel et al. 2008;Suldo et al. 2015). Extraverted individuals feel comfortable being around other people, start conversations easily, and like social attention; they are talkative and easy going. As a result, extraverted individuals are more likely to develop a sense of belonging, have more friends and acquaintances, are more likely to report high-quality relationships with others, less likely to feel lonely and socially excluded-all these attributes being important components of a happy life (Diener and Ryan 2009). In addition, social perception studies show that extraverted individuals are also perceived as being more likeable and preferred as communication partners and friends (Feiler and Kleinbaum 2015). What we see here is that higher levels of extraversion make one's social life easier and satisfy one's need to belong (Baumeister and Leary 1995), thus, representing an important factor contributing to happiness.
Conscientious individuals like order, pay attention to details, follow schedules and plans, are dutiful and responsible. Given that all these characteristics facilitate goal achievement and success, conscientious individuals report higher SWB than their less conscientious counterparts do (Hayes and Joseph 2003;Suldo et al. 2015). Other dispositional traits reflecting one's strength of will, such as the trait of self-control, as well as one's perception of control over one's life, show consistent positive associations with SWB as well (Hofmann et al. 2014;Lachman and Agrigoroaei 2010). Importantly, even though early research has almost unanimously assumed that differences in these basic personality traits result in different levels of SWB (Diener et al. 2003), more recent studies have shown that SWB can also affect personality development (Specht et al. 2013). Using the longitudinal data from Germany (German Socio-Economic Panel Study), Specht and colleagues (2013) showed that individuals with a higher (vs. lower) baseline level of life satisfaction were more likely to become more conscientious and less neurotic over time.
Finally, individuals' feelings and beliefs regarding the "self" represent another factor influencing well-being. Specifically, higher levels of self-esteem are positively associated with life satisfaction (Cheng and Furnham 2003). Individuals with high self-esteem are less likely to experience stressful life events (Orth and Luciano 2015) and are more likely to be liked and respected by others (Reitz et al. 2016).
Individual differences in goals, values, and beliefs are associated with SWB as well. The importance of goal attainment has been acknowledged in psychology since Maslow's pyramid of needs (Maslow 1943). Goals have also been seen as a crucial component of happiness in sociological theories. For example, the Social Production Function Theory (Ormel et al. 1999) postulates that people's SWB is contingent on achieving five universal goals (stimulation, comfort, status, behavioral confirmation, and affection).
Conversely, another stream of literature suggested that it does not matter much whether or not goals are attained, but rather it is the type of goals people pursue that is crucial for SWB. For example, pursuing avoidance goals (e. g. not failing an exam) is associated with lower SWB, while endorsing approach goals (e. g. getting a great grade in an exam) is related to higher SWB (Elliot et al. 1997). Pursuing materialistic goals was shown to lead to a lower SWB (Kasser and Ryan 1993), and pursuing goals emanating from intrinsic rather than extrinsic motivations is related to an increased SWB (Sheldon and Kasser 1998). Research into values and SWB has also backed up the importance of the self-determination view (Ryan and Deci 2001). Specifically, it has been shown that self-direction, stimulation, and achievement values are positively related to positive emotions, whereas security, conformity, and tradition values are negatively associated with positive emotions (Sagiv and Schwartz 2000).
Finally, beliefs about the world have also been shown to affect happiness. For example, individuals who believe in a just world are more likely to report higher life satisfaction and more frequent positive affect than their counterparts with a lower level of belief in a just world (Correia et al. 2009;Lucas et al. 2013). Furthermore, conservative political ideology (Napier and Jost 2008), sexist ideology (Hammond and Sibley 2011), religious beliefs , and interpersonal trust (Tokuda et al. 2010) are also often cited as sources of life satisfaction. However, as most of these studies are based on correlational data, it remains unclear whether holding certain types of beliefs makes one happier and more satisfied, or whether happy and satisfied people are more likely to endorse certain types of beliefs.
Sociodemographic factors, life circumstances and events have been shown to explain individual differences in happiness as well. The findings indicate that having a job (Carroll 2007), earning good money (Luhmann et al. 2011), as well as being married or having a long-term partner (Kim and MacKenry 2002) are all positive predictors of SWB. In contrast, unemployment, divorce, widowhood, and poverty contribute to a lower SWB (Lucas et al. 2004;Williams and Dunne-Bryant 2006). However, these effects are probably bidirectional such that life satisfaction can "predispose" individuals to certain life events. For example, prospective studies have shown higher levels of baseline life satisfaction to be associated with higher chances of marriage and lower risks of divorce and unemployment (Luhmann et al. 2013). It should be noted that the explanatory power of sociodemographic factors is very small compared to that of the Big Five and other personality traits. According to Lyubomirsky et al. (2005b), while personality accounts for about 50% of variance in SWB, individual differences in sociodemographics only explain about 10%.
At a theoretical level, the relative importance of personality predictors provides support to the so-called top-down theories of happiness, according to which individuals' stable predispositions with a large genetic component shape individuals' happiness level. This view has also been expressed in the set-point theories of happiness (Headey and Wearing 1992). According to these theories, individuals have a genetically determined default level of happiness. Different life circumstances and behaviors may change it in the short run, reducing or increasing individuals' K happiness that will however ultimately return to its set point. These theories have been supported by a strong temporal stability that happiness measures are known to possess, as well as by surveys of twins. For example, one highly cited twin study supported the set-point assertion but showed that happiness levels in monozygotic twins correlate at 0.40, and in dizygotic twins at just 0.08 (Lykken and Tellegen 1996). Similarly, long-term panel studies have shown the test-retest correlation of life satisfaction to reach high values (around 0.50) across decades (Schimmack and Oishi 2005).
In contrast, the bottom-up theories assume that happiness is a result of being in a beneficial vs. threatening environment or experiencing positive as opposed to negative events. In other words, "a happy individual is happy precisely because he or she experiences many happy moments" (Brief et al. 1993). Consistent with this idea, large-scale panel studies have shown happiness to be subject to major life events (Lucas 2007), thereby questioning the postulates of the set-point theory. Not all events are capable of affecting individuals' default level of happiness, though: while periods of unemployment were shown to leave long-term "scarring" effects, the happiness brought about by getting married was short-lived (Lucas 2007;Lucas et al. 2004).
While many life events lie outside individuals' control and responsibility, there are ways in which, researchers believe, people can still shape their own happiness. Specifically, individuals can change their SWB through "effortful activities". Indeed, individual differences in everyday behavior are assumed to account for about 40% of variance in SWB (Lyubomirsky et al. 2005b). What are the activities that promote happiness?
One of these activities is prosocial behavior. The idea that prosociality is an important source of happiness harkens back to ancient philosophy, with Aristotle and Plato believing that a life of virtue is the only possible route to happiness. Recent psychological and sociological research has provided support for this idea from Antiquity, showing that altruistic behavior, volunteering and charity donations are associated with increased life satisfaction (Aknin et al. 2013;Dunn et al. 2008;Schwartz et al. 2003;Stavrova et al. 2013b). Importantly, the benefits of altruism were supported in experimental research as well. For example, Aknin et al. (2013) provided their participants with a small sum of money and instructed half of them to spend it on themselves and the other half to spend it on others. By the end of the day, the members of the latter group (prosocial spending) reported higher happiness levels than those in the former group (selfish spending).
Not only prosocial, but also merely social activities have been recognized as a source of SWB. Experience sampling studies (in these studies, participants are signaled with a smartphone application several times per day within a certain fixed time period, such as a week, and asked to fill in a small survey about their current experiences, thoughts, and feelings; this method makes it possible to study individuals' daily experiences without relying on their memory) have shown that individuals report the lowest positive affect when they are alone. In fact, people are happier when they are with their clients and bosses than just alone (Kahneman 2000). Different measures of social engagement, the presence of friends and a romantic partner were all associated with higher levels of SWB (Lucas and Dyrenforth 2006).
Finally, recent studies suggest that genetic and environmental factors (life circumstances, life events, the activities one chooses to engage in) are not independent of each other (Plomin 1994). For example, in a theory of person-situation transactions, individuals self-select into certain environments based on their personality predispositions (Caspi et al. 1989). To give an example, neurotic individuals are more likely to withdraw from social events, thus, further exacerbating their unhappiness. As a result, individuals' life outcomes (including happiness) represent a joint product of both genetic predispositions (e. g. personality) and environmental factors. This enables characteristics of one's national context to be recognized as being particularly important "environmental" factors. The following section explores existing findings on country-level predictors of SWB.

Country-level Predictors of SWB
Countries show substantial differences in average SWB levels. For example, analyses combining individual-and country-level predictors suggest that regional and national differences explain large amounts of variation in individuals' life satisfaction (Bonini 2008). Also, national SWB scores show a surprisingly high level of temporal stability. For example, in a 30-year period, the Japanese level of life satisfaction did not substantially deviate from a 6 on a 10-point scale, whereas that of the Danes fluctuated around an 8 (Veenhoven 1993). Which country-level characteristics explain differences in national levels of SWB?
Multiple studies converge on differences in national wealth as one of the most important country-level variables related to life satisfaction (Diener et al. 2003). Correlations between gross domestic product (GDP) and the national level of life satisfaction are typically in the range of 0.50-0.60, with individuals living in wealthier countries reporting higher life satisfaction compared to individuals in poorer ones (Di Tella et al. 2003). Interestingly, the associations between wealth and life satisfaction at the individual level are substantially smaller: individuals' wealth (e. g. personal income) correlates with their life satisfaction at approximately 0.10. One explanation of why wealth seems so important at the country level is that wealthy nations typically score higher on a wide range of other characteristics that have been shown to have beneficial consequences for their citizens' SWB as well: political freedom, civil rights, good governance, low crime rates and low social inequality (Dorn et al. 2007;Oishi et al. 2011;Ott 2011). Although there have been attempts to disentangle the effects of these different factors, they have been largely unsuccessful due to strong intercorrelations between the predictors and a small number of countries available for the analyses (Diener et al. 2003).
Besides socioeconomic and political differences, cultural dimensions have been shown to explain country-level differences in SWB as well. Among them, individualism vs. collectivism and uncertainty avoidance showed the most consistent associations with SWB. Individualism (vs. collectivism) represents one of the most widely studied and recognized dimensions of culture (Hofstede 2001;Markus and Kitayama 1991;Triandis 1995). Individualistic cultures are characterized by an emphasis on individuals' (vs. groups') needs and values, values of autonomy, and independence, whereas collectivistic cultures are defined by prioritizing groups' values and well-being over those of individuals, security, and interdependence values and a stronger differentiation between in-groups and out-groups. People in collectivistic cultures tend toward an interdependent self-concept, defining themselves primarily as members of social groups. Individualism vs. collectivism, as well as many other dimensions of culture, have been quantified. That is, nearly a hundred national cultures have been assigned an individualism-collectivism score (as well as scores on other cultural dimensions such as uncertainty avoidance), which represents nationally aggregated individuals' responses to items yielding their preferred values, norms, attitudes, and behaviors. This has allowed researchers to explore the potential associations between cultural dimensions and SWB. As a result of this exploration, residents of individualistic countries were shown to score higher on life satisfaction and positive affect than residents of collectivistic countries did Hofstede 2001). Although this effect can be partly explained by the fact that individualistic cultures tend to be wealthier, it is consistent with different psychological perspectives highlighting the importance of freedom of choice and autonomy for well-being (e. g. the Self-Determination Theory; Deci and Ryan 2000;Ryan and Deci 2000).
The cultural dimension of uncertainty avoidance describes cultural differences in the tolerance of uncertainty and ambiguity (Hofstede 2001). While people in general show an aversion towards uncertainty, this is particularly so in cultures characterized by higher levels of uncertainty avoidance. Individuals in cultures with higher scores in uncertainty avoidance are particularly uncomfortable with uncertain and ambiguous situations, and rely on strict rules and regulations as a way of coping with uncertainty. In contrast, cultures with low scores on uncertainty avoidance show greater tolerance toward uncertainty; they perceive changeable environments and an uncertain future not as a threat but as an opportunity. Given that uncertainty is inherent to virtually any aspect of our lives, residents of cultures high in uncertainty avoidance tend toward lower SWB scores than cultures that are more tolerant of uncertainty (Hofstede 2001).
It is noteworthy that even though these studies examined the effect of culture as a macro-level characteristic, it is less clear whether the mechanism of these effects operates at a macro-or a micro-(i. e. individual) level. In fact, being part of an individualistic culture can strengthen an individual's SWB either because he/she is then more likely to hold an individualistic orientation him/herself or because living in an individualistic culture gives one more autonomy from others. Similarly, a country's level of uncertainty avoidance might be negatively associated with its citizens' happiness, either because the citizens are then more likely to be uncomfortable with uncertainty themselves or because being surrounded by individuals who are intolerant of uncertainty makes one's life stressful regardless of one's own uncertainty orientation. Examining the effects of these cultural orientations at both individual and country level would foster our understanding of these processes.
Even though no studies have undertaken this endeavor with respect to the dimensions of culture discussed above, several studies have attempted to differentiate between the effects of the same constructs measured at individual and country level in other areas. For example, it has been shown that more trusting individuals report higher levels of life satisfaction than their less trusting counterparts, and that residents of more trusting cultures score higher on life satisfaction than residents of less trusting cultures do (Tokuda et al. 2010). Importantly, the effects at the individual and the country level were independent of each other: individuals living in countries with a high level of social trust are happier than individuals living in countries with a lower level, regardless of their personal trust score (Tokuda et al. 2010). Similarly, individual, national, and regional levels of neuroticism were shown to be robust negative predictors of SWB (Rentfrow et al. 2008;Steel and Ones 2002). For example, Steel and Ones (2002) demonstrated a strong negative association between neuroticism and life satisfaction aggregated at the country level. This negative relationship was replicated at the state level in the U.S., such that states with a high aggregated neuroticism score tend toward a lower SWB than states with a lower level (Rentfrow et al. 2009). Most recently, multilevel analyses of individual neuroticism and regional neuroticism in Germany showed that living in a state that was neurotic was negatively related to life satisfaction, regardless of individuals' own neuroticism score (Stavrova 2015a). Regardless of how neurotic a person is, being surrounded by neurotic individuals (that is, living in a region with a high level of neuroticism) has a detrimental effect on one's happiness.
At the same time, some individuals' characteristics seem to affect SWB only when aggregated at the macro-level. For example, a series of studies have shown that an average level of education in cities was positively related to a city-average SWB, whereas the association between education and happiness at the individual level was negligible (Florida et al. 2013). Similarly, national IQ was shown to positively predict the happiness of nations, whereas between-individual differences in intelligence are not a robust predictor of individuals' SWB (Veenhoven and Choi 2012). Taken together, these results suggest that some individual characteristics do not affect SWB directly, but only when aggregated at a higher level (city, region, country), probably via shaping societal living conditions and cultural climates.
Finally, according to a number of psychological perspectives, country-level differences in SWB might be explained by cultural differences in self-enhancement and the cultural importance of happiness. One of the most robust findings in cross-cultural psychology is the variation in self-serving and self-enhancement motives and tendencies, with individualistic cultures showing stronger self-enhancement than collectivistic ones (Heine and Hamamura 2007). Multiple psychological studies have shown that people in individualistic cultures tend to consider themselves as scoring higher on a range of positive traits and as having better future prospects than most other people (Dunning et al. 2004;Weinstein 1980). Such self-serving biases have been shown to be less common in collectivistic cultures, suggesting that crosscultural differences in the mean level of SWB might be at least partly a result of a stronger self-enhancement tendency in individualistic (vs. collectivistic) nations.
Supporting this conclusion, several studies demonstrated that self-enhancement tendencies affect autobiographical memory (Ross and Qi 2010;Ross and Wilson 2002). Study participants in individualistic cultures tended to remember events that made them proud of themselves as being subjectively more recent than events that made them feel ashamed, while no difference was observed in East Asian participants (Ross and Wilson 2002). Following up on this finding, another study compared measures of happiness in U.S. and East Asian individuals that were either retrospective ("How satisfied were you with your life this week?") vs. immediate ("How satisfied were you with your life today?"), but averaged over seven days. The usual differences in life satisfaction between East and West emerged when retrospective but not immediate measures were used. In other words, when asked to report their satisfaction within the last week from memory, U.S. Americans were more satisfied than East Asians, although daily reports revealed no differences, suggesting that cross-cultural differences in mean levels of life satisfaction might be partly explained by cultural differences in self-enhancement and autobiographical memory biases (Oishi 2002). Hence, although cross-national differences in mean levels of SWB are large, temporally stable, and show consistent correlations with countries' socioeconomic, political and cultural characteristics, it remains to be explored to what extent these differences are "true" or just reflect cultural differences in indepth psychological processes, such as memory biases.

A Joint Examination of Individual-and Country-level Predictors of SWB
Countries might not only differ in mean levels of SWB, but also in its correlates. Are sources of happiness culturally universal, or do they vary across countries? And are there any systematic patterns in these variations? In contrast to sociological research tradition, psychological research often tends to assume universality in its theory and findings (Oishi et al. 2009). This tendency is evident in the recently criticized propensity of psychological researchers to draw conclusions about human nature from data obtained from WEIRD (Western, Educated, Industrialized, Rich, Democratic) samples (Henrich et al. 2010), as well as in theories of universal human needs and motivation. For example, Maslow advanced the idea of universal needs as far back as in the 1940s-needs inherent to human nature (Maslow 1943). Striving for universality can be observed in later theoretical work as well. For example, Ryff's psychological model of well-being postulates the existence of six universal needs (autonomy, growth, relationships, purpose in life, environmental mastery, and self-acceptance), the fulfillment of which leads to wellbeing (Ryff 1989), whereas Deci and Ryans' self-determination theory advances the existence of three such universal needs (autonomy, competence, relatedness) .
While intuitively appealing, such a universalistic approach has been only partially supported by empirical research. Even though positive relationships, self-esteem and other presumably universal sources of SWB typically show positive associations with life satisfaction across cultures, the magnitude of such associations often differs (Diener and Diener 1995). Although substantial efforts have been directed at uncovering systematic patterns in these variations, theory building has lagged behind. I give below an overview of empirical findings and theoretical approaches aimed at explaining cross-national variations in correlates of SWB in sociology, economics, and psychology.
Using the search terms "happiness or life satisfaction or subjective well-being" and "cross-national or cross-cultural" resulted in 1519 articles in academic journals in PsychINFO, PsychARTICLES, EconLIT and an additional 470 in Sociological Abstracts. A brief look at the listed articles showed that most of them do not explore cross-national variability in individual-level associations, but rather explore either predictors of SWB at the individual level or national differences in the mean level of SWB. Therefore, the search was refined by adding the term "multilevel" (as exploring cross-level interactions typically requires a multilevel analysis). This modification resulted in 54 articles in academic journals in PsychINFO, PsychAR-TICLES, EconLIT and an additional 12 in Sociological Abstracts. For this review, I selected empirical papers that explicitly indicated having explored both individual-and country-level predictors of SWB in the abstract. Further relevant empirical papers were detected via studying the literature cited in these articles as well as the literature that cited them. Finally, a number of relevant empirical papers-that were not detected using the search terms described above but that I was familiar with through my work-were considered as well. The list of reviewed papers is presented in the Appendix (Table 1). While this search strategy does not render the present review comprehensive, it nonetheless makes it possible to detect major theoretical and empirical trends in the literature.
An examination of the empirical work showed that most of the existing empirical results can be categorized as adopting one of two approaches: I describe the first one as the "institutional hypothesis", while the second one is typically referred to in the literature as the "fit hypothesis".

The Institutional Hypothesis
The theoretical reasoning behind most sociological and economic research exploring cross-national variability in correlates of SWB can be summarized as follows: individuals' characteristics contribute to happiness to the extent that macro-level conditions are favorable to individuals with these characteristics-what I refer to as the "institutional hypothesis". Below, I summarize research findings supporting (and refuting) the institutional hypotheses across different domains, including employment, family life, and health.

Employment
The adverse consequences of job loss for happiness and life satisfaction have been largely acknowledged in the literature (Carroll 2007). Cross-national studies in sociology and economics additionally explored whether labor market policies, including unemployment benefits and employment protection regulations, could mitigate the negative effect of unemployment. For example, Voßemer et al. (2017) studied the role of cross-national differences in labor market policies in shaping the well-being and health of unemployed individuals across 26 European countries. Consistent with the institutional hypothesis, they found that unemployed individuals were better off in countries with more (vs. less) generous unemployment benefits. This finding was further refined by Ochsen and Welsch (2012), who distinguished between the effects of employment protection policies and unemployment benefits, using the data from nearly 400,000 individuals in ten European countries, from 1975 to 2002. Their analyses showed that employment protection and a higher benefit replacement rate were positively associated with the life satisfaction of everyone, with employment protection being particularly appreciated by the employed but higher replacement rates-by the unemployed. Having said that, not all the studies supported the institutional hypothesis. For example, Eichhorn's (2014) analyses of the data from the European Values Study including 28 countries showed that the effect of unemployment on life-satisfaction was not moderated by unemployment benefits. In another study, a generous social policy regarding unemployment benefits has been shown to mitigate the negative effect of financial hardship on the SWB of self-employed individuals across 31 European countries (Annink et al. 2016), suggesting that the beneficial effect of supportive institutions might be more specific than previously assumed, at least with respect to unemployment benefits.

Family Life
Another type of social benefits that has been explored as a factor potentially contributing to the well-being of social groups in need of support are family and parental benefits. Most studies examining the effect of parenthood on SWB have shown that parents typically report lower SWB levels than childless individuals (e. g. Luhmann et al. 2012). Can the burden of parenthood be alleviated by welfare state support? Glass et al. (2016) showed that the negative effect of parenthood on happiness was weaker in countries with more (vs. less) generous welfare policies supporting families. It has been suggested that welfare state support directed at parents can compensate for the stress of parenthood, reducing the disparities in happiness between parents and nonparents (see also Hank and Steinbach 2019). Generous welfare state support has also been shown to mitigate the negative effect of being childless in old age. Specifically, Neuberger and Preisner (2017) used the data of the Survey of Health, Ageing and Retirement in Europe (including 19 countries) and the English Longitudinal Study of Ageing-both large-scale surveys including respondents aged 50 and older-to show that generous social benefits are associated with a higher SWB for childless elderly individuals.

Health Care and Aging
Another line of research explored the role of health care policies in mitigating the effect of ill-health on SWB (see also Pförtner et al. 2019). For example, Kööts-Ausmees and Realo (2015) used the data from 32 countries (using the data of the European Social Survey) and showed that even though the association between selfreported health status and life satisfaction is positive across countries, it was weaker in countries with higher (vs. lower) government spending on health care programs. That is, investing in health care appears to reduce the negative effect of ill-health on life satisfaction: becoming ill in a country with meager health care spending might result in accumulating personal debt or refusing medical treatment altogether, ultimately resulting in further health deterioration.
The generosity of the welfare state was shown not only to compensate for individuals' health problems, but also to make life easier for the elderly. For example, Moor, de Graaf and Komter (2013) showed that older individuals are more satisfied with life in countries with better welfare services targeted at the elderly. Specifically, they detected a positive association between the share of the elderly in institutional care homes (used as an indicator of the welfare generosity directed at the elderly) and the life satisfaction of elderly individuals in general. Similarly, the negative consequences of financial hardship among the elderly are buffered by the generosity of the welfare state. Specifically, Niedzwiedz et al. (2015), using the data of elderly individuals (Survey of Health, Ageing, and Retirement in Europe) living in Southern, Scandinavian, Postcommunist, and Bismarckian welfare regimes, demonstrated that the negative effect on SWB exerted by low socioeconomic status was attenuated in Scandinavian regimes (which are known for their generous social spending).

Migration
In light of the ongoing migration and refugee crisis in the Europe, several studies examined the role of social policies in migrants' satisfaction (see also Careja 2019). Kogan et al. (2017) showed that the life satisfaction of immigrants in Europe (18 EU countries were analyzed) depends much more on natives' attitudes toward migration than it does on legal immigration regulations and policies. A welcoming social climate including positive migration attitudes was shown to make a happy migrant. Heizmann and Böhnke (2018) distinguished between migrants from the EU and other countries, and found that Kogan and colleagues' findings primarily apply to migrants from the EU, whereas the life satisfaction of migrants from outside the EU does benefit from inclusive integration policies.

Gender
Besides specific policies directed at improving the living conditions of one or another group, cross-national sociological research considered inequalities between different social groups. For example, Başlevent and Kirmanoglu (2017) explored the happiness gap between working women and housewives across 29 European countries. They showed that working women were happier than housewives and that this happiness gap increased with increasing country-level gender equality in economic participation, educational attainment, and political empowerment. Crossnational differences in gender equality were also shown to explain between-country variation in gender differences in happiness, with women being more satisfied in more gender-equal countries (Tesch-Roemer et al. 2008). However, this finding was not replicated among adolescents. Looze et al. (2017) used individual-level data of over 150,000 adolescents from 34 European and North American countries. Their analyses indicated that the national-level of gender equality (e. g. women's economic participation and decision-making power in politics and business) showed a positive association with life satisfaction in both boys and girls to the same extent. Similarly, despite some support for a positive effect of gender equality on women's happiness, researchers have also noticed that the happiness of American women has been in decline since the 1970s, despite growing gender equality (Stevenson and Wolfers 2009). Potentially, increases in gender equality resulted in a higher role ambiguity and role conflict in women who are now supposed to succeed in both family life and at work (see also Grunow 2019). These findings point toward a limitation of the institutional approach, showing that an objective improvement in living conditions might do little to enhance happiness, unless it is followed by the requisite changes in the sociocultural climate.

Miscellaneous
Although most research testing the "institutional hypothesis" explored the role of social policies, institutional and structural differences can more generally foster or undermine the well-being of one or another group as well. For example, Tay et al. (2014), who analyzed individual-level data from over 100 countries, showed that the positive association between individuals' income and SWB is enhanced in countries with higher national corruption levels (probably as it allows one to bribe more officials-an ability crucial for survival in corrupt countries). Sortheix and Schwartz (2017) examined the effect of personal values from Schwartz's values theory on life satisfaction across 25 European countries and showed that while the effects of some values did not show meaningful cross-cultural variations (e. g. high benevolence and hedonism values were related to higher life satisfaction, whereas high power and security values were related to lower life satisfaction), the effects of other values were less universal 1 . For example, individuals scoring high on achievement value were more satisfied with life if they lived in countries with lower scores in the Human Development Index, whereas individuals scoring lower on achievement values were happier in countries with higher scores in the Human Development Index. A strong achievement orientation might potentially foster prosperity, which is more important for well-being in developing than in developed countries.

The Fit Hypothesis
The role of sociocultural climate has been the focus of psychological research exploring cross-cultural differences in correlates of SWB. This line of research advanced what can be called a "fit hypothesis", according to which individuals' characteristics contribute to SWB to the degree to which these characteristics are widespread and, consequently, socially desirable in a particular culture (Stavrova 2014;Stavrova et al. 2013a, b). This theoretical perspective is grounded in decades of psychological research on normative conformity, social sanctions, social identity and person-environment fit (Abrams et al. 2002;Cialdini and Goldstein 2004;Kristof-Brown et al. 2005).
Studies on the psychology of social processes showed that people like similar others more than dissimilar others (Byrne 1961) and judge members of their in-group more positively than members of their out-group (Tajfel et al. 1971). Similarly, peo-ple judge group members holding opinions or views that deviate from the group's average as socially unattractive and unlikeable; they are more likely to ostracize and exclude them from the group (Abrams et al. 2002;Ouwerkerk et al. 2005). This idea is supported by the studies on conformity and norm violations, showing that people whose behavior or opinions deviate from social norms are often subject to informal social sanctions, such as disapproval and exclusion (Christensen et al. 2004;Pool et al. 1998). In fact, as shown by the research on the backlash effect, behaviors typically related to social approval and respect might backfire and result in disrespect when executed by those for whom they represent a violation of social norms (Rudman and Fairchild 2004). For example, volubility, expressions of anger and assertive negotiation techniques have overwhelmingly positive consequences for men, but negative ones for women (Brescoll 2011;Brescoll and Uhlmann 2008). People are often aware of the backlash that being different than the majority can elicit. This is illustrated by the so called Spiral of Silence phenomenon that describes the hesitancy with which minority (vs. majority) opinion holders express their opinions (Noelle-Neumann 1974). For example, individuals who believe that their opinion is discrepant from (vs. congruent with) the majority's opinion are slower (that is, more hesitant) when it comes to publicly sharing their opinion (Bassili 2003).
Finally, studies in organizational psychology (Kristof-Brown et al. 2005) have demonstrated the importance of fit between individuals and their immediate work contexts. These studies have shown that individuals whose values, goals, and personality are similar to those in their organization or team are more likely to have higher job satisfaction, commitment and performance than individuals who are dissimilar to their immediate work environment (Kristof-Brown et al. 2005).
The idea of fit has transcended cross-cultural literature on correlates of SWB as well. Cross-cultural studies in psychology traditionally focus on comparing the predictive validity of the same characteristic across several countries with different scores in important cultural dimensions, such as individualism vs. collectivism. Individualistic cultures value positive emotions, autonomy and self-concern, whereas collectivistic cultures place emphasis on relationships concern, interdependence and normative conformity (Oyserman et al. 2002). Consistent with the fit perspective, relationship concern was shown to be a stronger predictor of SWB in collectivistic cultures, whereas self-concern was a stronger predictor of SWB in individualistic ones (Mesquita and Karasawa 2002). Similarly, satisfaction with one's freedom, experiences of positive emotions and valuing pleasure (hedonism) were more strongly associated with life satisfaction in individualistic countries than in collectivistic ones (Joshanloo and Jarden 2016;Oishi et al. 1999;Suh et al. 1998). Taken together, these findings suggest that people are happy and satisfied with life to the extent that their characteristics are in harmony with cultural values.
Although cultural differences along the individualism vs. collectivism dimension are well established, they might not represent the optimal way to explore the fit hypothesis. Individualism-collectivism (as well as any other cultural dimension) describes cultural differences not just in one but in multiple values, needs, traditions, customs and the like, and thus lack the specificity required for a straightforward test of the fit idea.
K A most recent stream of research therefore explored the fit hypothesis more directly by investigating the interactions between individuals' characteristics and the degree to which these characteristics are widespread and socially desirable in a given cultural context. These two aspects of normative behaviors-being common and socially desirable-represent core attributes of social norms in social psychological literature (Cialdini et al. 1990). They are referred to as descriptive (as they describe behaviors that people commonly exercise under certain circumstances) and injunctive (as they describe behaviors that people believe one should exercise under certain circumstances). It should be noted that this approach differs from the dominant view in the sociological literature that only recognizes the injunctive (but not the descriptive) component as part of the definition of social norms (Coleman 1990). Using the social norms approach to explain cross-national differences produced promising results by demonstrating the importance of fit for SWB in different life domains. For example, the happiness gap between employed and unemployed individuals was shown to vary as a function of a country-level social norm to work, with the unemployed being especially dissatisfied with their lives in countries with a strong social work ethic (Stavrova et al. 2011).
The person-culture fit pattern also emerged with respect to other areas of human life, such as marital status and parenthood. For example, while cohabiting women are often shown to report lower SWB than their married counterparts, this effect was shown to be restricted to countries with a strong norm for women to get married (Stavrova et al. 2012). Similarly, the misery of single parents and parents raising a child out of wedlock (cohabiting) could be substantially alleviated in cultural contexts with tolerant norms regarding childbearing practices. In countries where childbearing in cohabiting unions and by single parents is common and normatively accepted, there was a smaller gap in life satisfaction between married and cohabiting or single parents than in countries where these alternative family models were not accepted Fetchenhauer 2015a, 2015b).
Interestingly, support for the fit hypothesis was also demonstrated with respect to less visible individual characteristics such as beliefs, worldviews, and ideologies. A series of studies showed that the positive effect of religious beliefs on SWB was restricted to religious countries (Gebauer et al. 2017;Stavrova et al. 2013). Similarly, holding specific secular beliefs, such as a belief in scientific-technological progress, was associated with a higher SWB to the extent that this belief was common in a particular cultural context (Stavrova et al. 2016). Conservative political ideology was a positive predictor of SWB only in above-average conservative contexts (Stavrova and Luhmann 2016a). Finally, even the effect of characteristics that are sometimes described as representing universal sources of happiness, such as virtue and prosociality (Aknin et al. 2013) has been shown to vary across countries, following a fit pattern. For example, helping others was related to higher life satisfaction in countries where helping was common than in countries where helping others was rare (Oarga et al. 2015). Similarly, the positive effect of civic virtue on life satisfaction turned negative in countries where above-average virtuous behaviors in economic games was subject to sanctions (Stavrova et al. 2013b).
What mechanisms account for the fit effect? On the one hand, individuals who deviate from the majority behavior (or in other words who violate social norms) are likely to be sanctioned by their fellow group members. Indeed, multiple studies in social psychology and sociology have shown that individuals expressing a deviant opinion are more likely to be disliked by others and even ostracized (Christensen et al. 2004;Wood et al. 1997). In turn, social exclusion represents one of the most important factors that negatively affect psychological functioning (Williams 2007;Williams et al. 2000). Similarly, loneliness has been shown to hinder psychological well-being, resulting in poor health and even increased mortality risks (Cacioppo and Cacioppo 2014;Cacioppo et al. 2010). It follows that pursuing a lifestyle that significantly deviates from the social norms in one's surroundings is likely to give rise to negative behaviors from others, resulting in social isolation and therefore damaging one's sense of well-being. In brief, this explanation assumes that the reason behind the unhappiness of norm-deviant individuals are social sanctions-the toll of disrespect and the pain of social exclusion.
At the same time, social norms work not only because people fear punishment (social sanctions) in case of noncompliance, but also because they are often internalized and become part of individuals' selves. In this case, a lower SWB of normdeviant individuals might be explained by their self-disappointment and a feeling of guilt associated with failing to live up to one's own standards. Psychological studies have shown that people tend to make an effort to minimize the discrepancy between their actual and ideal self-concept (Crocker and Knight 2005). This mechanism is most likely to be at work in the case of involuntary nonconformity, such as involuntary unemployment (in a country with a strong work ethic), cohabitation (in a country with strong traditional family norms) or single parenthood (in a country with a strong two-parent family norm). Yet, nonconformity is sometimes more voluntary: People usually have some degree of freedom to decide whether they want to support a minority political ideology or become an atheist in a religious country. The negative consequences of nonconformity are therefore hard to explain in such cases by individuals' self-disappointment or guilt associated with falling short of reaching their ideal selves. After all, they could become their ideal selves if they just changed their views.
Several papers have attempted to explore which of the two mechanisms discussed above-deviance from social vs. personal norms-can best explain the negative effect of a lack of fit. In a cross-cultural study of unemployed individuals' well-being, Stavrova et al. (2011) included a measure of the norm to work at both individual and national level (the latter was obtained by aggregating individuals' responses to the norm to work scale). If individuals' deviation from their personal norm of work and the resulting feeling of self-discrepancy and guilt represent the underlying mechanisms of the fit effect, then the interaction between individuals' employment status and the national norm to work should vanish once researchers control for individual differences in the personal norm to work. Yet this did not take place: the stronger the country-level norm to work was, the less happy the unemployed were, regardless of how strong or weak their personal norm to work was. Similar results were obtained in a couple of other studies that explored the role of social vs. personal norms. For example, in a study of 43 European countries, Stavrova and Fetchenhauer (2015b) showed that single parents were particularly dissatisfied with life in countries with a strong two-parent family norm. Importantly, this effect was present even among single parents who themselves did not hold a two-parent family norm, pointing to the working of the social sanction mechanism of the fit effect. The same conclusions were reached in a study of cohabiting parents' happiness: cohabiting (vs. married) parents were less happy in countries with strong traditional family norms, regardless of whether they themselves supported these norms or not (Stavrova and Fetchenhauer 2015a). Taken together, these findings provide support for the social sanctions (rather than norm internalization) explanation of the fit effect.
To summarize, the literature reviewed points to substantial cross-cultural variability in the degree to which individuals' characteristics are associated with SWB and thus challenges the assumption of the universality of human happiness. Existing literature seems to converge on the idea that multiple individuals' characteristics contribute to happiness to the extent that they are common and socially desirable in a particular sociocultural context. The cross-national variability in effect sizes can be considered as not only statistically but also practically significant. In fact, in some domains, the particularities of a national context do not only make an individual-level association smaller, but actually make it vanish or even reverse. Knowing what country-level characteristics switch the predictors of SWB on and off and why represents an important step in assessing the cross-national generalizability of SWB findings.

Directions for Future Research
Although cross-national comparative research on SWB has been greatly facilitated by the inclusion of measures of life satisfaction in most cross-national large-scale survey programs, it is still in its infancy. The vast majority of cross-cultural studies in psychology still involve comparing findings across two or three countries or cultural contexts. While this approach usually uses highly valid instruments and sophisticated (experimental or longitudinal) study designs, it does not permit any definitive conclusions to be drawn about the role of country-level characteristics in explaining between-country differences in the results. To give an example, observing a positive association between self-esteem and happiness in individualistic Canada and a zero correlation in collectivistic Japan is informative, yet it tells us little about the role of individualism in driving these differences.
In this sense, studies that rely on large-scale cross-national survey datasets and statistically test the effect of contextual characteristics on within-country relationships (e. g. using a multilevel analysis) represent an important methodological advancement. The availability of basic SWB measures in multiple large-scale crossnational studies currently allows researchers to explore the variations in correlates of life satisfaction across more than 100 countries. Yet, this research is restricted by the measures available in such datasets. While personality characteristics (e. g. the Big Five personality traits) represent the most important predictors of SWB, there is currently not a single publicly available cross-national dataset that includes reliable measures of personality. On the contrary, such datasets usually include measures of sociodemographic characteristics, values, attitudes, and behaviors-aspects that were shown to explain relatively little variance in SWB (Lyubomirsky et al. 2005b). As a result, we know quite a lot about the cross-cultural variability of factors whose overall impact on SWB is small, but we know very little about the crosscultural variability of the most important predictors of SWB, such as personality. As SWB is not only a topic of interest to psychologists, but also to sociologists and economists, it is highly recommended for cross-national large-scale survey projects to include measures of basic dimensions of personality to fill in this knowledge and data infrastructure gap.
Another limitation that is associated with the nature of available cross-national data is that such data are usually cross-sectional and, thus, do not allow causal inferences. Nevertheless, most studies using cross-national data tend to interpret their results in causal terms, typically implying a causal effect of individuals' characteristics on their happiness (see also Schmidt-Catran et al. 2019). Indeed, most theoretical models in sociology and economics consider happiness as an outcome, not as a predictor. However, recent longitudinal studies have shown that individual differences that usually serve to predict SWB might well be its outcomes. These studies have shown that life satisfaction and happiness contribute to better health and longevity (Danner et al. 2001;Diener and Chan 2011;Stavrova 2019), as well as career success (Cropanzano and Wright 1999;Rose and Stavrova 2019), and can even trigger important life events such as marriage or separation (Luhmann et al. 2013;Stavrova and Luhmann 2016b). These findings make the interpretation of established patterns of cross-cultural variations in correlates of SWB problematic. For example, the fit perspective assumes that endorsing culturally shared beliefs contributes to SWB. Alternatively, it is plausible that (1) SWB contributes to the endorsement of culturally shared beliefs or (2) that both SWB and the endorsement of culturally shared beliefs are driven by further unassessed factors (e. g. high selfcontrol might result in both higher SWB and conformity). It is possible that not only the strength but also the causal direction of the associations between individual characteristics and SWB varies across countries. To conclude, extending cross-national studies on correlates of SWB to include longitudinal or experimental (when appropriate) data in select countries will become crucial in advancing this research field.
While most empirical work exploring cross-national variability in correlates of SWB falls under one of the two theoretical approaches described here-institutional and fit hypotheses-there are multiple further macro-level indicators reflecting different aspects of individuals' living conditions (wealth, social inequality, political freedom, etc.) that can be used to explain cross-national differences in the correlates of SWB. The most prominent example is the role of national wealth (GDP per capita) in the association between individuals' income and their SWB. A couple of studies have shown that the positive association between individuals' income and their SWB is stronger in poorer countries than in wealthier ones (Diener et al. 2010;Schyns 2002), suggesting that money buys happiness when it ensures that basic needs are met (access to clean water, medical care, etc.). An exploration of how country-level differences in further objectively measurable characteristics such as income inequality, human development or particularities of the political regime might be a worthwhile endeavor for future cross-national comparative research.
It should be acknowledged that cross-national comparative research on hedonic components of well-being has been flourishing despite these limitations. In contrast, however, cross-cultural studies on eudaimonic well-being are virtually nonexistent. Yet, eudaimonic well-being-in particular its currently most often studied component, meaning in life-is of increasing importance for both individuals and society to function (Hill and Turiano 2014;Stavrova and Luhmann 2016b). In fact, monitoring and seeking to improve citizens' sense of meaning in life has been discussed as a potential matter of public policy (Steger 2014). Hence, exploring cross-cultural differences in mean levels and correlates of meaning in life might represent an important direction for future studies.
On a related note, the theoretical developments in the area of cross-national SWB research might be used to explain cross-national differences in predictors of physical health. Indeed, a series of recent studies have detected that cross-national variations in the associations between religiosity and physical health show a fit pattern as well. Across nearly 60 countries, religious individuals only enjoyed better physical health than nonreligious individuals in countries where religiosity was common and socially desirable. In contrast, in secular countries, being religious was not associated with any physical health advantage. This pattern was detected across the U.S. census regions and was even extended to mortality risks. Religiosity was only related to lower mortality risks in highly religious regions (e. g. the Bible Belt; Stavrova 2015b).
It is noteworthy that certain statistical techniques, such as multilevel regression or multilevel structural equation modeling, made testing both the institutional and the fit hypotheses possible. Nevertheless, a substantial number of countries is typically required for this technique to lead to reliable conclusions, and most of the studies reviewed here barely satisfied these requirements (Hox 2002). Besides including more countries, one way to address this limitation is to move to a lower level of analysis (e. g. studying regions within one or several countries). Adopting this method might be particularly promising when trying to disentangle the institutional and the fit hypotheses: while institutional effects most often exist at national level, the effect of social norms and culture can be tested at lower levels as well.
Finally, understanding what drives cross-cultural differences in correlates of SWB might represent a tool for studying the mechanisms behind the associations between individual characteristics and happiness. For example, research on the positive effect of religiosity on life satisfaction has considered multiple potential mechanisms, including increased social networks (Lim and Putnam 2009), healthy behaviors (Wallace and Forman 1998), as well as social approval and respect (Gebauer et al. 2017), etc. The fact that the positive effect of religiosity is restricted to religious countries can be interpreted as providing support to the social approval mechanism. In other words, religiosity is only related to increased life satisfaction in contexts where it represents a source of higher social esteem, approval and recognition, suggesting that it is these social benefits (rather than healthy behaviors or other factors) that represent the underlying mechanism of the effect.

Conclusions
Why study happiness? Besides feeling good, happiness is associated with many positive life outcomes. Happy people benefit from better health and longevity (Danner et al. 2001), career success (Cropanzano and Wright 1999), and positive relationships (Stavrova and Luhmann 2016b) than their less happy counterparts (for a review, see Lyubomirsky et al. 2005a). Happiness promotes helping and prosociality (Dulin and Hill 2003), broadens individuals' cognitive repertoire (Fredrickson and Branigan 2005) and involvement with approach goals (Elliot and Thrash 2002). Hence, the study of happiness and factors promoting individuals' happiness appears an important endeavor with the potential to improve the human condition.
Cross-national comparative research on SWB has enjoyed fast-growing interest from a number of social science disciplines. These studies made substantial contributions to our understanding of cultural differences in mean levels and correlates of SWB. Further extending empirical studies to include longitudinal and experimental designs and promoting theory building in the area might represent the most immediate steps that future research should undertake.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The positive effect of employment protection on SWB is stronger in employed persons of intermediate age (compared to women/ housewives and older people). The effect of generous unemployment insurance on SWB is stronger in women/housewives, older people and the unemployed The positive association between financial satisfaction and SWB is stronger in poorer (vs. wealthier) nations. The positive association between the satisfaction of esteem needs and SWB is stronger in individualistic (vs. collectivistic) nations Suh et al. (1998)

Emotions
Individualism 61 Emotions are a stronger predictor of SWB in individualistic (vs. collectivistic) countries The negative effect of unemployment on SWB was alleviated in counties with a weak social work ethic; national unemployment rates did not moderate the effect of unemployment on SWB Stavrova et al. (2012) Gender and partnership arrangement (marriage vs. cohabitation)

Gender role norms 30
Cohabiting women report lower SWB than married women do, especially in countries with conservative gender role norms; gender role norms had no effect on cohabiting vs. married men's SWB Stavrova et al. Religiosity is positively associated with SWB, and this relationship is especially strong in countries with a stronger (vs. weaker) norm of religiosity Stavrova et al. (2013b) Civic virtue Country-level civic virtue; country-level antisocial punishment (punishment of high contributors in public goods games) rates 73/13 Civic virtue is positively associated with SWB, but less so in countries with a weak countrylevel civic virtue and high countrylevel antisocial punishment rates Oarga et al. (2015) Prosocial behavior Cohabiting parents report lower SWB than married parents do, but only in countries with a strong norm proscribing childbearing in cohabiting unions The negative effect of unemployment on SWB is stronger in countries with a strong work ethic (especially for men and the longterm unemployed) Van de Velde et al. (2017) Religious service attendance; frequency of prayer Country-level religiosity 29 The negative association between service attendance and depression is weaker in less (vs. more) religious countries. The positive association between frequency of prayer and depression is stronger in less (vs. more) religious countries