Macau study links relative income to life satisfaction through three perceptions

Macau residents who place their household higher on a 10-point income ladder report greater life satisfaction, according to a cross-sectional analysis of 1,023 adults in the seventh wave of the World Values Survey. The association remained significant after researchers adjusted for age, gender, education, marital status and whether a respondent was born in Macau. Statistical mediation analysis identified three potential indirect pathways: greater acceptance of income inequality, higher generalized trust and a stronger sense of control over life. Perceived control accounted for the largest estimated proportion of the total relative-income and life-satisfaction association, although the observational design cannot establish that any of these attitudes caused higher satisfaction.

The survey data were collected in Macau in 2019 using a multistage probability sample of residents aged 18 or older. Life satisfaction was measured on a 10-point scale, with respondents asked how satisfied they were with their life as a whole. Relative income was assessed by asking participants to place their household on a scale from the lowest income group at 1 to the highest at 10, taking account of wages, salaries, pensions and other earnings. The core ordinary least-squares regression estimated a coefficient of 0.279 between relative income and life satisfaction before controls were introduced and 0.275 after adjustment, both with p-values below 0.001. Because the study was cross-sectional, these coefficients describe associations in the survey population rather than changes in satisfaction caused by income.

Three attitudinal variables provided a more detailed account of how economic position might be connected with well-being. Acceptance of inequality was measured on a scale running from support for making income more equal to support for greater incentives for individual effort. Generalized trust was captured by a binary question about whether respondents trusted most people. Control perceptions were measured with a 10-point scale indicating how much choice and control people believed they had over the course of their lives. In the mediation models, each of these measures was positively associated with both relative income and life satisfaction, producing the statistical pattern needed for indirect pathways. The measures were not randomized interventions, however, and the analysis did not establish a causal sequence among income, attitudes and satisfaction.

Not all three pathways were equally strong. The indirect association running through acceptance of income inequality was statistically significant, with a Sobel test statistic of 3.466 and p < 0.001. The pathway through control perceptions was also significant, with Z = 5.052 and p < 0.001. By comparison, the indirect association through generalized trust was only marginally significant at Z = 1.932 and p = 0.053. The researchers therefore described acceptance of inequality and control perceptions as supported mediators, while treating trust as only partially supported. The direct relative-income coefficient remained significant after the mediators were introduced, which is consistent with partial rather than complete statistical mediation in the fitted models.

Estimated proportions of the total association attributed to the three indirect pathways were 10.0% for acceptance of income inequality, 4.5% for generalized trust and 36.9% for control perceptions. These percentages should not be read as causal shares of an effect, nor as proportions of the indirect association alone. The authors specifically note that they are descriptive summaries of indirect associations in observational data. Within the fitted statistical model, perceived personal control was more closely connected to the relative-income and life-satisfaction relationship than the two social attitudes were, but the result does not show that increasing control would necessarily increase satisfaction.

Several robustness checks produced the same broad result. The association persisted in complete-case analysis, which excluded observations containing missing values. Researchers also replaced the income-ladder measure with subjective social class and found a significant positive association with life satisfaction at p < 0.001. Substitution of the outcome measure produced similar findings: relative income was positively associated with self-reported happiness and self-rated health, both at p < 0.001. Ordered logit and ordered probit specifications, used to account for the ordinal nature of the 10-point satisfaction scale, also retained a positive relative-income relationship at p < 0.001. The study’s authors cautioned that income placement and subjective social class may overlap conceptually, so the agreement between those measures does not amount to fully independent validation.

Missing information was addressed with 20 imputed datasets generated through chained equations. Generalized trust had 968 valid observations, corresponding to 5.38% missing data, while age had the largest gap, with 813 valid responses and a 20.53% missing rate. Most other observed variables had less than 2.5% missingness. A Harman single-factor check was also performed within the multiple-imputation framework; the first factor accounted for 36.3% to 36.7% of total variance, below the study’s 50% threshold for evidence of severe common-method bias. The analyses were conducted in Stata 15.0, using ordinary least-squares regressions for the main models, complete-case and alternative-model checks for robustness, Sobel tests for the indirect associations, and interaction models to examine demographic differences.

The income and life-satisfaction relationship was similar across demographic groups. Separate estimates were positive and highly significant for men and women, for respondents aged 18 to 35 and those aged 36 or older, for people with lower or higher education levels, for married and unmarried respondents, and for Macau-born residents and migrants. Coefficients ranged from 0.206 to 0.309, with every subgroup p-value below 0.001. The point estimates differed modestly: the coefficient was 0.278 for men and 0.273 for women, 0.239 for respondents aged 36 or older and 0.206 for those aged 18 to 35, 0.309 for people with post-secondary education or higher and 0.274 for those with less education, 0.268 for married and 0.281 for unmarried respondents, and 0.295 for Macau-born residents and 0.255 for migrants. None of the five interaction terms between relative income and gender, age, education, marital status or immigration status was statistically significant, providing no empirical support for demographic variation in the association.

Macau’s distinctive social setting makes the findings noteworthy. The special administrative region combines rapid socioeconomic development, relatively extensive welfare provision, cultural diversity and longstanding interaction between Chinese mainland and international visitors. The study does not show that relative income matters more in Macau than elsewhere, because it did not conduct a cross-country comparison of effect sizes. It does show that perceived economic standing is positively associated with satisfaction in this population. In the adjusted basic model, age and education were also positively associated with life satisfaction, while immigration status had a significant negative association, indicating that migrants reported lower satisfaction than locally born respondents. The model also found no statistically significant relationships for gender or marital status in that specification.

The central limitation is the study’s cross-sectional design, which prevents the researchers from determining temporal order or causal direction. The survey was not designed as a longitudinal experiment, so it cannot show whether economic position, attitudes and life satisfaction changed in a particular sequence or whether changes in one preceded changes in another. Self-reported measures introduce the possibility of response bias, including social-desirability effects, while single-item indicators for life satisfaction, relative income, inequality acceptance, trust and control limit reliability and conceptual depth. The analysis therefore identifies statistical associations and partial mediation in the Macau sample, rather than causal mechanisms. Its clearest result is that higher relative income was consistently linked with greater life satisfaction, with perceived personal control showing the strongest statistical indirect pathway among the three candidates tested.

Subject of Research: Cross-sectional study of relative income, life satisfaction and three possible mediators among Macau residents

Article Title: Social comparison and well-being in Macau: the mediating roles of income inequality acceptance, generalized trust, and control perceptions

Article References: Zheng, A. J. (2026). Social comparison and well-being in Macau: the mediating roles of income inequality acceptance, generalized trust, and control perceptions. International Review of Economics, 73(2), Article 41. https://doi.org/10.1007/s12232-026-00557-7

Image Credits: AI Generated

DOI: 10.1007/s12232-026-00557-7

Keywords: Macau, relative income, life satisfaction, income inequality, generalized trust, perceived control, social comparison, World Values Survey, subjective well-being, cross-sectional study, Social, comparison