Status perception, residential context and subjective wellbeing among South Korean adults: A multilevel analysis of primary survey data

ABSTRACT

This study examines the multilevel associations between perceived socioeconomic status (SES) and subjective wellbeing (SWB) among community residing South Korean adults. Based on primary survey data, this research provides evidence on the contingent ways in which status perception and wellbeing are associated across individual and contextual (residential-community) levels. Three related but distinct questions guide the analysis. First, net of objective SES indicators (income, education, employment status, homeownership and financial assets), is perceived status related to SWB? Second, is the community-level status, net of individual-level status, associated with wellbeing? Lastly, how do residential characteristics moderate the linkage between status perception and SWB? Results from fitting hierarchical linear models reveal that perceived status is significantly and positively related to SWB. Adjusting for individual-level controls, significant contextual effect also emerges, i.e., residence in a higher-status community is associated with greater wellbeing. Finally, the relationship between status perception and SWB is stronger in lower-quality communities measured in terms of aggregate perceived SES, level of business commercialization (e.g., density of factories), suicide prevalence, and proportion of government welfare recipients.

KEYWORDS:

• Perceived status
• Subjective SES
• Wellbeing
• Happiness
• Neighborhood
• Effects

View correction statement:

Correction

Notes

¹ According to a large number of studies, such as those reviewed in Euteneuer (2014) and Tang et al. (2016), there is indeed a robust relationship between SSS (subjective social status) and a wide range of health outcomes and biological risk factors. And in many cases, SSS are shown to be a much more reliable predictor vis-à-vis the more traditional counterparts, namely objective socioeconomic measures. As has been pointed out, one critical issue in the current scholarship is to what extent negative affect may drive the statistical association between SSS and health. The question is whether the relationship remains significant after adjusting for negative affect. In recognition of this potential confounding effect, the current study includes depression as a control variable for a more stringent test.

² A study by Präg et al. (2016), for example, analyzes cross-national data to investigate the subjective SES-health association. It finds that the slope magnitude is greater for higher-income countries. While noteworthy for its methodological approach, the study is limited in that it does not consider how and the extent to which subjective status at individual and contextual (country) levels may independently and interactively relate to psychological wellbeing.

³ According to Alderson and Katz-Gerro (2017), all things equal, Americans who rate their respective income more highly are happier and the reference group salient for social comparison conditions this income-wellbeing relationship. Also based on the US data, Firebaugh and Schroeder (2009) offer a complimentary argument, noting that neighbors’ income plays a critical role in subjective wellbeing. Specifically, living in richer neighborhoods within poorer counties is associated with greater happiness. While both studies highlight the importance of relative income and the importance of context, neither examines the issue from a multilevel analytic perspective. The current research does so by shifting the focus to hierarchically nested data of individual residents clustered in different communities and how community characteristics partly account for the variance in the dependent variable (happiness) across individuals, as well as the degree to which those characteristics moderate the relationship between relative status position and psychological wellbeing.

⁴ Properly conceptualizing and operationalizing “neighborhoods” has been a subject of much debate (Sharkey and Faber 2014). Despite the lack of clear consensus, there is nevertheless general agreement that “the place matters” (Sampson, 2011) when it comes to analyzing health outcomes, as well as other issues of inequality (see also Sharkey, 2013).

⁵ According to the Easterlin, i.e., “income-happiness,” Paradox (Eaterlin, McVey, Switek, Sawangfa, & Zweg, 2010), a richer person is happier than a poor person. Yet, a nation’s average happiness does not rise, at least in the long run, as the country grows in wealth (e.g., per capita GDP). Clearly, while Korea’s economy grew at an impressive rate in the past, it has not been followed by a corresponding increase in its citizens’ overall subjective wellbeing. We can add to this observation that a higher-income country is not necessarily happier than a lower-income country, as showcased by the above comparison concerning Korea.

⁶ According to a reviewer, another possibility or pathway is that relative status is related to depression via subjective wellbeing. To test this, an additional multilevel model was estimated using (ln) depression as the dependent variable. The key result, summarized in Table A3, was not statistically significant. A further recommendation was to see whether relative status was associated with self-rated or physical health. Once again, the result was shown to be statistically not significant. Table A4 provides the evidence for this.

⁷ Bivariate correlations among the area-level variables suggest the possible need for data reduction due to collinearity problems (for detail, see Table A1). As a precautionary measure, Principal Component Analysis (PCA) was conducted to see whether the variables could be reduced, given the relatively small number of the level-2 sample size of 77. Results from PCA using the orthogonal (Varimax) rotation reveal three distinct constructs. Kaiser–Meyer–Olkin value exceeds the conventional limit of 0.5, and the Barlett’s Test of Sphericity is also statistically significant (Chi-Square = 95.49, df = 15, p < 0.001). Factor loading scores for the three components range between 0.78 and 0.91, indicating relatively high internal validity. Together, they explain 78% of the total variance. Table A2 summarizes the main PCA results. Scree Plot also shows three factors with the Eigenvalue greater than 1. Multilevel regression models (not shown) were re-estimated using the three composite scores based on PCA. Findings concerning the relationship between perceived status and subjective wellbeing at individual and contextual levels, for the main purposes of this study, are similar to those reported below in the main text.

⁸ As a reviewer of this paper poignantly states, subjective social status may capture certain elements not measured by objective SES variables, such as informal sources of income from relatives. Consequently, statistical results based on an aggregated version of this variable in multilevel modeling may reflect some of the unobserved characteristics rather than the pure casual effect of perceived or relative status. This is a legitimate concern and the findings reported above should be interpreted with caution. Despite this limitation, prior research amply demonstrates that people’s general wellbeing, both physical and mental, is significantly shaped by where or how they place themselves on the hierarchy of status attainment (e.g., income ladder). That is, social comparison, along with the related concept of relative deprivation, plays a critical role in determining one’s happiness. Of course, we cannot rule out the possibility that perceived status is a proxy for something other than relative deprivation derived from social comparison. However, a substantial interdisciplinary literature provides empirical evidence that is consistent with main findings of the current research: The status of neighbors and one’s relative position among them have both a direct and indirect effect on subjective wellbeing. Future studies would benefit from incorporating potential sources of omitted variables that might bias the parameter estimates under discussion.