Introduction

In addition to reducing risk factors for illness, there has been increasing interest in identifying and fostering strengths for promoting well-being1. One positive attribute that may enhance well-being is expressing one’s character, for example, living in accord with one’s ethical values and beliefs. Character has been considered by historically prominent philosophers as a necessary condition for attaining a fulfilled life and complete well-being2,3.

Psychologists Park, Peterson, and Seligman proposed a family of 24 character strengths (e.g., kindness, perseverance, fairness, honesty, gratitude, etc.), which reflect core ethical values that are ubiquitous across cultures4,5. Character strengths can be largely divided into “strengths of the head” (self-oriented, e.g., curiosity, hope) and “strengths of the heart” (interpersonal-oriented, e.g., kindness, gratitude)6. Measures to assess each character strength on a continuum have been developed and widely used in research (e.g., the values in actions [VIA] classification of strengths)4. In addition, survey questions that assess one’s global character—a general disposition to promote good of oneself and others—have also been developed1. Individuals with a greater disposition toward promoting good are more likely to express specific character strengths across contexts, thus disposition to promote good may be considered as a higher-order assessment of character and an upstream factor that precedes the actions to express character strengths (e.g., volunteering, civic engagement, service to others), what one might refer to as “moral perseverance”7. An example item for assessing the global character of promoting good is “I always act to promote good in all circumstances, even in difficult and challenging situations”1,8. Past research has suggested that a greater orientation to promote good is associated with better health and well-being outcomes, such as greater life satisfaction8, social well-being8, mental health9,10, self-rated physical health8, and overall flourishing11. Some specific aspects of promoting good such as benevolence have also been associated with a lower risk of mortality12. Experimental studies further suggest that character is potentially modifiable via strength-based interventions that aim to help people fulfill their potential13,14,15.

Everyone has the potential to live according to their ethical beliefs. Through identifying a set of 24 character strengths nearly universally accepted across cultures, Park et al. speculated that “Characters are grounded in biology through an evolutionary process that selected for these predispositions toward moral excellence as means of solving the important tasks necessary for survival of the species”16. Culture may, however, shape the specific ways in which character strengths are operationalized. Take the character strength of perseverance as an example: individualism-oriented cultures emphasize autonomy and self-set goals, whereas collectivist cultures prioritize interpersonal harmony and pursuit of group goals17. Therefore, the understanding of perseverance in Western countries may be more centered around personal achievement, whereas in collectivist cultures perseverance may more demonstrate one’s commitment to the group’s success that contributes to group harmony18. Further, structural factors may also shape the expression of character. For instance, the prevalence of “heart strengths” (e.g., kindness, fairness) in a city is associated with city-level contextual features such as population density, cost of living, and proportion of families with children, possibly via affecting the extent of opportunities for individuals to develop meaningful social interactions6. Although the specific ways to express character strengths may vary across societies, analyses of traditions across countries and cultures also indicate that there is a great deal that is shared19. Prior evidence suggests that a general disposition to promote good is likely universally valued, even in challenging and difficult circumstances (e.g., in countries with a recent history of conflicts)9.

Within each society, character may be patterned by sociodemographic factors, as evidenced by prior research on disposition to promote good, specific character strengths, and prosocial behaviors more broadly. First, prior evidence suggests that character strengths tend to increase with age. For instance, a recent survey using a representative sample of American adults revealed a monotonic positive correlation between age and orientation to promote good20. Several studies in Europe also found that most character strengths increase with age, with the strongest correlations for curiosity, self-regulation, forgiveness, and gratitude5,21. Second, evidence on gender difference in character remains mixed. For instance, some studies among American adults found that women vs. men report a greater orientation to promote good and are more likely to participate in prosocial activities (e.g., volunteering)22,23, whereas other studies found little evidence of gender differences or the differences might vary by the particular character strength being assessed24,25,26. Third, social integration may enhance expression of character. For instance, marriage (as a form of social integration) has been associated with a greater likelihood of volunteering27. Fourth, the extent of opportunities to express character may be patterned by socioeconomic status (SES). Psychological research (often conducted in experimental settings) has sometimes found an inverse association between indicators of SES (e.g., education, income, occupational prestige) and prosocial behaviors (e.g., charitable giving)28. Some researchers posited that individuals with lower SES have fewer economic resources and experience more environmental threats, and may thus develop heighted sensitivity to the environment and others, leading to greater empathy29. In contrast, research from other disciplines (often used large samples in observational studies) found either positive or u-shaped associations28. The contrary findings may be explained by methodological differences28. It is also possible that the associations are modified by cultural (e.g., interdependence/independence)30 and contextual factors (e.g., societal economic inequality)31. We speculate that prosocial acts involving sharing material resources may be more dependent on SES, whereas one’s general disposition to promote good of both oneself and others may be less contingent on one’s economic resources or social identities (e.g., race and ethnicity, immigration status, etc.). Next, many religions consider character building as a central component of their teaching. An important avenue that Park et al. took to identify the 24 character strengths was via reviewing religious scripts4. While empirical research has seldom directly studied religion and one’s global disposition to promote good, religious participation has been linked with higher levels of specific character strengths (e.g., kindness, gratitude, forgiveness)32 and greater engagement in prosocial acts (e.g., volunteering, civic engagement)33.

Prior studies have significantly advanced our understanding of character, but some knowledge gaps remain. First, most prior research studied specific character strengths (kindness, gratitude, etc.) and prosocial acts (e.g., volunteering, charity giving, civic engagement), whereas one’s general disposition to promote good remains less understood. Disposition to promote good provides a higher-order measure of character, which has great potential to be used in large-scale surveys to further understand character across contexts and cultures1. Second, prior studies often used small and/or non-representative samples, which limits the external validity of findings. Third, most prior evidence was from Western, Educated, Industrial, Rich and Democratic (WEIRD) countries. However, the expression of character may be shaped by various cultural (e.g., collectivism/individualism)17 and contextual factors (e.g., economic development, income inequality, political system)6, thus we should not assume the findings from WEIRD countries can be applied to other countries. Knowledge on social distribution of promoting good across diverse populations will help inform culturally relevant strength-based programs that aim at enhancing people’s capacities to express character and identify social groups for whom such programs may be most meaningful.

To begin addressing these knowledge gaps, using data from 22 diverse countries (with the samples weighted to be nationally representative within each country), this study examined population mean levels of orientation to promote good across countries and its distribution within each country by key sociodemographic factors. We hypothesized that (1) the mean levels of promoting good would vary considerably across countries, (2) the means of promoting good would exhibit variations across sociodemographic categories on average across countries, and (3) the differences in means of promoting good by sociodemographic categories would also vary by country.

Methods

The description of the methods below has been adapted from VanderWeele et al.34. Further methodological detail is available elsewhere35,36,37,38,39,40,41.

Study population

This study used data from the global flourishing study (GFS), a longitudinal study that enrolled 202,898 adults (age range: 18–99 years) from 22 culturally and geographically diverse countries, with the samples weighted to be nationally representative within each country. Survey items included aspects of well-being such as happiness, health, having a sense of meaning and purpose, character, social relationships, and financial stability, along with various demographic, socioeconomic, political, religious, personality, childhood, community, and behavior factors. Data collection was carried out by Gallup via a combination of modes (e.g., in-person, phone, web) that varied across countries38. Data for Wave 1 were collected principally during 2023, while some countries began data collection in 202235. The following countries were included in wave 1 data collection: Argentina, Australia, Brazil, Egypt, Germany, Hong Kong (Special Administrative Region of China), India, Indonesia, Israel, Japan, Kenya, Mexico, Nigeria, the Philippines, Poland, South Africa, Spain, Sweden, Tanzania, Turkey, United Kingdom, and the United States. The precise sampling design to ensure nationally representative samples varied by country35. During the survey translation process, Gallup adhered to TRAPD model (translation, review, adjudication, pretesting, and documentation) for cross-cultural survey research (ccsg.isr.umich.edu/chapters/translation/overview)40. The data that supports this research are publicly available via the Center for Open Science (https://www.cos.io/gfs). Details about the GFS study methodology and survey development are reported elsewhere35.

The present study used data from all participants in Wave 1 of GFS (N = 202,898). Poststratification and nonresponse adjustments were performed to ensure the sample was representative of the adult population in each country35,38. Ethical approval was granted by the institutional review boards at Baylor University and Gallup, and all participants provided informed consent. All methods were carried out in accordance with relevant ethical guidelines and regulations.

Assessment of orientation to promote good

An item from VanderWeele’s Flourishing Index1 was used to assess one’s general orientation to promote good: “I always act to promote good in all circumstances, even in difficult and challenging situations”. The response was rated on a Likert scale ranging from 0 (not true of me at all) to 10 (completely true of me). The response was analyzed as a continuous score, with a higher score indicating a greater orientation to promote good.

Assessment of sociodemographic characteristics

Sociodemographic factors were assessed through participant self-reports. These included age (in years, responses grouped into “18–24’, “25–29”, “30–39”, “40–49”, “50–59”, “60–69”, “70–79”, and “80+”), gender (male, female, or other), marital status (never married, married, separated, divorced, widowed, or domestic partner), employment status (employed by an outside organization, self-employed, retired, student, homemaker, unemployed and searching, or other), educational attainment (up to 8 years, 9–15 years, or 16+ years), religious service attendance (> once/week, once/week, one-to-three times/month, a few times/year, or never), and immigration status (born in this country, born in another country). Religious affiliation was also assessed, and response options included 15 major religions [e.g., Christianity, Islam, Hinduism, Buddhism, Judaism, etc.], “some other religion”, or “No religion/Atheist/Agnostic”; response categories varied by country42. Race/ethnicity was assessed in most countries (except in Germany, Japan, Spain, and Sweden, due to restrictions in collecting such data), with response options varied as appropriate for each country.

Statistical analyses

The descriptive analyses showed participant sociodemographic characteristics in individual countries and in the total sample, weighted to be nationally representative within each country. The population mean levels of orientation to promote good (adjusted for complex survey design and weighted to be nationally representative) were estimated for each country separately along with 95% confidence intervals, standard deviations, and for presentation purposes, we ordered the mean estimates from highest to lowest. Additionally, we calculated Gini coefficient43 to evaluate the population inequality in orientation to promote good within each country. The Gini coefficient, ranging from 0 (perfect equality) to 1(highest inequality), was originally used to assess population income inequality and has been increasingly used in measuring disparities in health and well-being44. Variation in mean orientation to promote good by sociodemographic categories was estimated, with all analyses initially conducted by country (reported in the online supplement). A global test was conducted (with a global p-value calculated) to assess whether the mean levels of promoting good varied statistically significantly across categories of each sociodemographic variable within each country.

In the primary analyses we conducted random effects meta-analysis to pool mean values of orientation to promote good in each specific sociodemographic category from the country-specific analyses45,46, along with 95% confidence intervals, standard errors, lower and upper limits of 95% prediction intervals, heterogeneity (τ), and I2 to estimate the extent to which the levels of promoting good in a given sociodemographic category vary across countries47. Forest plots of estimates are shown in the online supplement. We examined religious affiliation and race/ethnicity (when available) in country-specific analyses but not with the meta-analyses, because the observed response categories for these two demographic factors varied by country. We reported a pooled global p-value48 across countries to evaluate whether the levels of promoting good varied statistically significantly by each sociodemographic variable within at least one country. Bonferroni corrected p-value thresholds are provided in the meta-analytic results based on the number of sociodemographic variables49,50.

Meta-analyses were conducted in R51 using the metafor package52, and country-specific analyses were performed in SAS 9.4 (SAS Institute, Inc). As a supplementary analysis, population weighted meta-analysis was performed to pool results from country-specific analyses. This study was pre-registered with Center for Open Science prior to data access (https://doi.org/10.17605/OSF.IO/CYQXH); all code to reproduce analyses are openly available in an online repository39.

Missing data

In the combined cross-national sample, 0.33% of the participants had missing data on orientation to promote good. Missing data on sociodemographic factors ranged from 0.01 to 1.01% (except for race/ethnicity, which was not measured in some countries). We conducted multivariate imputation by chained equations (five imputed datasets created) to impute missing data on all variables in each country separately53,54,55. We included sampling weights in the imputation models to account for specific-variable missingness that may have been related to probability of inclusion in the study36.

Accounting for complex sampling design

The GFS used different sampling schemes across countries based on availability of existing panels and recruitment needs35. All analyses in the present study accounted for the complex survey design components by including weights, primary sampling units, and strata. Additional methodological details are reported elsewhere36,38.

Post Hoc analysis (not pre-registered)

We additionally conducted exploratory pairwise comparison of the means of promoting good between categories of each sociodemographic factor. Pairwise differences were estimated separately by country, and then pooled for a cross-national average difference using random effects meta-analyses.

Results

Descriptive analyses

The weighted cross-national sample had a balanced representation of different age groups, except that fewer participants were over 70 years old (Table 1). The weighted sample had 51% female participants, 49% male, and a small proportion who identified as other gender (0.3%). Most participants were married (53%), employed by an outside organization or self-employed (57%), had 9–15 years of education (57%) and were native-born (94%), with a smaller proportion reporting they never attended religious services (37%). The sample size for each country ranged from 1,473 (Turkey) to 38,312 (United States). See Supplementary Tables S1A to S22A for participant characteristics for each individual country.

Table 1 Distribution of sociodemographic factors in the overall sample combined across 22 countries (weighted to be nationally representative within each country, N = 202,898).
Full size table

Ordered means of orientation to promote good by country

Of all countries, Indonesia (mean = 8.63, 95% confidence interval [CI] 8.58, 8.69) and Argentina (mean = 8.63, 95% CI 8.57, 8.68) had the highest mean score of promoting good, and Japan was the only country with a mean score below 7.00 (mean = 6.13, 95% CI 6.09, 6.16) (Table 2). The countries with the highest means of promoting good were predominantly countries that are considered more collective56 (e.g., Indonesia, Argentina, Mexico, Philippines, Nigeria). The countries with the lowest means included several individualistic and high-income countries57,58, such as Sweden, the United Kingdom, and Germany. Standard deviation of the mean score was above 2.00 in all African countries and in several Asian countries (e.g., India and Japan). Furthermore, the countries with lower means of promoting good tended to have greater inequality in population distribution of promoting good, as indicated by a larger Gini coefficient.

Table 2 Ordered means of orientation to promote good with standard deviations and gini coefficients by country.
Full size table

Sociodemographic variation in promoting good: pooled estimates across countries

The random effects meta-analyses which pooled results across countries found that the mean scores of orientation to promote good varied by all sociodemographic factors that we examined (Table 3), although in many cases differences in demographic categories when the results were pooled across countries were relatively small, e.g. 0.2 or 0.3 (though see below for country-specific differences which were often much larger). The mean scores of promoting good also differed across categories for all sociodemographic factors in at least one of the 22 countries (all global p-values were below the Bonferroni corrected significance level of p < .007). For instance, the youngest age group reported the lowest mean score of promoting good among all age groups (e.g., meanage 18−24 =7.81 [95% CI 7.53, 8.10] vs. meanage 70−79=8.06 [95% CI 7.83, 8.30]). The mean score was similar between male (mean = 7.99, 95% CI 7.74, 8.23) and female participants (mean = 8.03, 95% CI 7.79, 8.27), but it was considerably lower in those who identified as other gender (mean = 7.48, 95% CI 6.99, 7.96). However, the sample size of other gender identities was very small (0.3%), thus the estimates may involve considerable statistical uncertainty. Participants who were married (mean = 8.12, 95% CI 7.89, 8.34) reported a higher mean score than those in all other marital statuses (e.g., meansingle, never married=7.82, 95% CI 7.49, 8.15). Moreover, participants who were unemployed and looking for a job reported a lower disposition to promote good (mean = 7.72, 95% CI 7.34, 8.10) than those employed by an outside organization (mean = 8.04, 95% CI 7.78, 8.30), the self-employed (mean = 8.14, 95% CI 7.92, 8.35), or the retired (mean = 8.06, 95% CI 7.83, 8.29). Similarly, the means of promoting good increased incrementally with years of education (e.g., meanup to 8 years=7.91 [95% CI 7.61, 8.21] vs. mean16+years=8.18 [95% CI 7.92, 8.44]). Next, the means of promoting good also increased monotonically with frequency of religious service attendance (meannever attendance=7.76, [95% CI 7.52, 8.01] vs. mean> 1/week attendance=8.47 [95% CI 8.29, 8.65]), which constituted the largest difference across demographic groups. Lastly, the mean score of promoting good was slightly higher among the native-born (mean = 8.01, 95% CI 7.77, 8.25) than immigrants (mean = 7.82, 95% CI 7.58, 8.05).

The heterogeneity estimate τ (i.e. the estimated standard deviation of the mean scores across countries) suggests some variation in the means across countries, with τ ranging between 0.50 and 0.80 for most sociodemographic categories (Table 3). The variation was particularly large among the group of unemployed and looking for a job (τ = 0.90), and particularly small for the categories of religious service attendance > 1/week (τ = 0.42) and 1/week (τ = 0.38). The I2 values were mostly above 92 across sociodemographic categories, suggesting that variability in mean orientation to promote good across countries may be mainly attributed to heterogeneity rather than sampling variability. The estimated prediction intervals were also wide across sociodemographic categories, adding further evidence on heterogeneity across countries.

Table 3 Random effects meta-analysis of the mean scores of orientation to promote good by sociodemographic category.
Full size table

Sociodemographic variation in promoting good: country-specific analyses

In country-specific analyses (Supplementary Table S1B to S22B and Supplementary Figures S1 to S34), the mean scores of promoting good varied by almost all sociodemographic factors (with the global p-value < 0.05, except for immigration status) in most countries. First, there was a strong age patterning in almost all countries (except Egypt, Israel, Turkey, and United Kingdom), although the pattern varied across countries. In most high-income countries (e.g., Australia, Japan, Spain, Sweden, United States), mean orientation to promote good increased with age in a nearly monotonic fashion. In contrast, in several economically less developed countries (e.g., India, Philippines, Tanzania), the mean score generally decreased with age. Second, the means of promoting good differed by gender in around half of the countries. Among these countries, female participants often reported a slightly higher mean score than male participants (except in Brazil, Hong Kong, Israel, and Kenya), whereas the mean score was considerably lower among participants who identified as other gender in most countries (e.g., Australia, Indonesia, Japan, Sweden), though again there was considerable sampling uncertainty in this group. Third, orientation to promote good varied by marital status in most countries. The never married often reported a lower mean score of promoting good than other marital statuses (e.g., Argentina, Australia, Brazil, Japan, Mexico, Spain, Sweden). However, in several countries with a low national divorce rate (e.g., India, Kenya, and Philippine), the divorced had the lowest mean score of promoting good among all marital statuses. Fourth, promoting good differed by employment status and years of education in most countries. Specifically, the mean score was lower among the unemployed and looking for a job than other employment statuses in many countries (e.g., Hong Kong, India, Germany, Japan, Sweden, United Kingdom). In several countries (e.g., Argentina, Germany, Spain), students also reported a low mean value of promoting good, which may of course also be related to age. In more than half of the countries (e.g., Germany, India, Poland, Spain, Tanzania, United Kingdom, United States), the means of promoting good also increased with years of education in a nearly monotonic pattern. Fifth, orientation to promote good increased with higher frequency of religious service attendance in almost all countries. The only countries with a global p-value > 0.05 for religious service attendance are Egypt, India, and South Africa.

The means of promoting good differed by immigration status (with a global p-value < 0.05) in only a handful of countries/territories. Specifically, the mean score was higher among the native-born than immigrants in Hong Kong and India, whereas the pattern was reversed in Spain, Sweden, the United Kingdom, and the United States. The mean scores of promoting good by religious affiliation and race/ethnicity in each individual country are also reported in the online supplement (Supplementary Table S1B to S22B).

The population weighted meta-analysis yielded similar results to the random effects meta-analysis, with only two exceptions (Supplementary Table S23). First, in population-weighted meta-analysis the mean score of promoting good was lowest in the oldest age group (i.e., 80 years+, followed by the youngest group of 18–24 years), whereas the mean score was lowest in the youngest group in the random effects meta-analysis. Second, the divorced reported a lower mean score of promoting good than all other marital statuses in the population weighted meta-analysis, whereas the never married reported the lowest score in the random effects meta-analysis. These differences are substantially driven by the large weight given to Indian in the population weighted meta-analyses. Results of the post hoc analyses on pairwise comparison of the means of promoting good between categories of each sociodemographic factor are reported in Supplementary Figures S35 to S115.

Discussion

Living in accordance with one’s ethical values helps individuals identify meaningful life pursuits and lead a fulfilling life5. For many people this will involve an orientation to promote what is good for oneself and for others. Fostering character and cultivating strengths facilitate individuals’ quest to achieve fulfillment, realize potential, and contribute to society4. Past empirical studies have focused on studying specific character traits. This study adds novel evidence on global character (specifically, a general disposition to promote the good of oneself and others) in a range of culturally diverse countries, with the samples weighted to be nationally representative within each country.

To our knowledge, this is the first study that examines the ordered means of people’s general disposition to promote good across countries. Similar to prior evidence on specific character strengths16,59, this study found that the mean score of people’s general disposition to promote good was above the middle point of the scale in all countries (all countries reported a mean value above 6.00 on a 0–10 scale; Japan was the only country with a mean below 7.00), suggesting that character may be universally valued across societies16,60. This study, however, also found evidence that is somewhat contrary to some prior findings. For instance, earlier studies (mostly conducted prior to 2015) found that the rank ordering of the 24 specific character strengths is similar across countries, and there is little evidence of patterning by the country’s economic development level16,59. In comparison, in this sample the countries with highest means in promoting good are mostly those that are considered more collective (e.g., Indonesia, Philippines, Argentina, Tanzania), whereas the countries with lowest means are predominantly high-income and individualistic countries/territories57,58 (e.g., German, United Kingdom, Sweden, Hong Kong; all had a mean score nearly 1 point lower than Indonesia/Argentina). Individualistic countries may prioritize personal achievement and autonomy, whereas collectivistic countries may more greatly value group harmony and community well-being, and therefore may naturally emphasize mindsets and actions that benefit the collective61. Data for this study was collected recently during 2022–2023, thus we speculate that these results may also be related to the global surge in prosocial activities during the COVID-19 pandemic, which was particularly evident in developing countries62. There is also prior evidence suggesting a decline in prosocial activities (e.g., formal volunteering) in some wealthy countries prior to COVID-1963, which might be attributed to the increased social disparity, decreased religious participation, and weakened social capital in those societies over recent years64. However, we acknowledge that these findings on country means need to be interpreted with caution, because measurement invariance across countries cannot be assumed and the mean scores may not be directly comparable between countries due to methodological and interpretative complexities in cross-country research. Differences across countries may be influenced by matters of translation, different modes of assessment, cultural norms, interpretation of items and also of response scales, and seasonal effects in the time of assessment65. It is important for future studies to further investigate to what extent the scores of promoting good are comparable across diverse populations and societies.

This is also the first study that calculated Gini coefficients to evaluate inequality in the population distribution of orientation to promote good across countries. In this sample, the countries with more equal distribution of the disposition to promote good are mostly those that are considered collective and less economically developed, whereas countries that showed the highest inequality in promoting good are predominantly high-income individualistic countries. Such findings suggest that cultural values and societal structural factors play important roles in shaping the equality of character distribution within populations. Future studies may consider using more nuanced measures of inequality66 to enhance our understandings of the population distribution of orientation to promote good across cultures, societies, and contexts.

This study also adds novel evidence on the age patterning of individuals’ orientation to promote good. The pooled estimates across countries suggest that older vs. younger adults, on average, reported a higher tendency to promote good. This finding is congruent with prior research that found positive associations between age and specific character strengths (including character strengths that are more directly relevant for the promotion of good such as: kindness, perseverance, fairness). For instance, a recent meta-analysis a of 47 samples (N = 1,098,748 individuals) found an overall positive correlation between age and 23 (out of 24) character strengths, although the effect sizes were generally small67. The Socioemotional Selectivity Theory posits that as individuals age they are increasingly aware of their limited lifetime, thus tend to prioritize goals and activities that lead to meaning and fulfillment68. Promoting good is one of such activities that may confer a sense of generativity, purpose, and fulfillment69. The country-specific analyses, however, suggest variations in the dynamics across countries—in this sample the disposition to promote good generally increased with age in high-income countries, whereas the pattern was reversed in some less economically developed countries. We speculate that the different age patterning may be related to different societal trends and structural factors across societies. For instance, in some high-income countries there has been a mental health crisis and declined sense of purpose in young people over recent years20, which may have impeded their capacities of expressing character oriented towards the good of oneself or towards others. In less developed countries, there are sometimes weak social protection system for older adults (limited or non-existent: pensions, social security benefits, access to affordable healthcare, and elder care services)70, which may have led to structural barriers for their participation in generative pro-social or self-oriented activities. It will be worthwhile to further investigate societal factors and structural conditions that may enhance or impede people’s capacities to promote good across different national contexts in future studies.

This study finds strong evidence suggesting that orientation to promote good varies by SES - the mean scores of promoting good were generally higher among the employed vs. the unemployed, and the means of promoting good also increased with years of education in this sample. This is congruent with most of the prior findings from disciplines outside of Psychology, which suggested positive associations between SES and prosocial behaviors28. Prior researchers posited that individuals with higher SES have more material resources and a lower cost of sharing those resources, and the likelihood of prosocial acts increases when the cost of such behaviors is low28, which may in turn reflect self-perceptions on orientation toward the good. To our knowledge, whether the global character of promoting good is patterned by SES has seldom been studied. This study adds such evidence and suggests that individuals with higher SES may have greater resources for expressing character broadly across contexts, and that this pattern is present across countries at various economic development statuses (e.g., Tanzania, Brazil, India, Germany, Japan, United States).

The present study also expands the literature on religious involvement and character. In this sample, the means of promoting good increased with frequency of religious service attendance in almost all countries, even in some of the most secular societies71 (e.g., Japan, Spain, Sweden). Indeed, religious service attendance manifested the largest differences in means of promoting good across groups of all demographic variables examined. These differences are consistent with prior studies suggesting positive longitudinal associations of religious service attendance with specific character strengths32 and prosocial acts33. Attending services may strengthen adherents’ beliefs in the importance of character in achieving personal fulfillment, honoring spiritual and/or religious commitments, and attaining community well-being. Participation in structured rituals and institutions may also help foster specific character strengths (e.g., perseverance, self-regulation) and connect individuals to a larger group with shared values and purpose that help reinforce one’s ethical beliefs and service to others67. Further studies are needed to provide more nuanced understanding of the dynamics between religious involvement and character across religious denominations and across different national religious contexts.

This study has several limitations. First, while the item for measuring disposition to promote good was taken from a previously validated index1 for assessing multi-dimensional well-being, the GFS cognitive interview indicates that some participants had difficulty in understanding the question/translation40. Therefore, measurement errors may need to be considered. Second, what is perceived as “good” probably varies across individuals, societies, and cultures. As a single-item higher-order measure of character, the self-reported survey item on promoting good in this study is not taking a stance on what “goodness” is, but rather assessing to what extent individuals report acting in accordance with what they perceive as good. Even though people may understand “goodness” differently, the measure can provide initial evidence on the extent to which people perceive themselves as being able to act for good. Future studies that further understand what is perceived as “good” across diverse populations will enhance our conceptual understanding of character and help inform the refinement of the measure of promoting good. Third, the findings on country ordered means of promoting good need to be interpreted with caution, given the methodological complexities in cross-cultural research (e.g., challenges in survey question translation, different response styles across cultures, etc.). Fourth, all analyses in this study are cross-sectional and descriptive, thus causality cannot be inferred. Fifth, because we do not assume measurement invariance in promoting good across countries, we used random effects meta-analysis to pool country-specific estimates in our primary analysis. It would be worthwhile for future studies to validate the measure of promoting good in diverse populations and use other modelling strategies (e.g., multi-level regression analysis) to further understand individual and contextual level factors that shape people’s orientation to promote good across societies. Sixth, some important contextual factors such as neighborhood characteristics and community support systems that may also shape individuals’ orientation to promote good were not examined in this study due to lack of data. These limitations are, however, balanced by important contributions of this study. This is the first study that examined the variation in people’s general disposition to promote good across countries and across sociodemographic groups within each country, using data from a wide range of diverse countries. This coverage expands beyond the Western, Educated, Industrialized, Rich, and Democratic (WEIRD) samples that have been typically evaluated in this area of research72.

Character is often considered as an essential component of well-being1. Character helps promote a sense of purpose, resilience, meaningful social interactions, and the fulfillment of one’s potential4, and is also of importance in its own right3. Expressing character oriented towards the good may also contribute to the trust and support within communities that in turn strengthens the overall social fabric4. This study provides novel evidence on variations in the means score of disposition to promote good across sociodemographic groups and countries. It is worth noting that such evidence does not indicate certain sociodemographic groups inherently have stronger or weaker character. Rather, it suggests the importance of investigating structural and societal factors that may differentially impede or enhance people’s capacities to express character across different sociodemographic groups. Further investigations on this topic will help identify potential sociodemographic groups for whom programs for reducing structural barriers may be most meaningful, which in turn will enhance population equity in capacities to express character oriented to the good. Further research will also enhance our understanding of culturally appropriate approaches for fostering character and the roles of character in attaining human flourishing.