The expanding class divide in happiness in the United States, 1972–2016

The expanding class divide in happiness in the United States, 1972–2016. Twenge, J. M., & Cooper, A. B. Emotion, Jul 2020. https://doi.org/10.1037/emo0000774

Is there a growing class divide in happiness? Among U.S. adults ages 30 and over in the nationally representative General Social Survey (N = 44,198), the positive correlation between socioeconomic status (SES; including income, education, and occupational prestige) and happiness grew steadily stronger between the 1970s and 2010s. Associations between income and happiness were linear, with no tapering off at higher levels of income. Between 1972 and 2016, the happiness of high-SES White adults was fairly stable, whereas the happiness of low-SES White adults steadily declined. Among Black adults, the happiness of low-SES adults was fairly stable, whereas the happiness of high-SES adults increased. Thus, the happiness advantage favoring high-SES adults has expanded over the decades. Age–period–cohort analyses based on hierarchical linear modeling demonstrate that this effect is primarily caused by time period rather than by birth cohort or age.


General Discussion
Between the early 1970s and the mid-2010s, socioeconomic status (SES) indicators such as income, education, and occupational prestige became stronger predictors of happiness among American adults over age 30. The happiness of low-SES White adults declined between 1972 and 2016 while the happiness of high-SES White adults stayed steady or declined less steeply; the happiness of low-SES Black adults generally stayed steady, while the happiness of high-SES Black adults generally increased. The result is a growing class divide in happiness, with the happiness advantage of high-SES individuals growing steadily larger over the decades.

Contrary to some previous studies, we did not find SES to be a weak or non-existent correlate of happiness (Diener, Ng, Harter, & Arora, 2010; Lykken & Tellegen, 1996; StewartBrown et al., 2015), especially in recent years. Instead, we found robust associations, particularly for income (with d = .59 between the 1st and 5th quintiles of income, and correlations as high as r = .26). The median effect for actual observed effect sizes in psychology is r = .19, and r = .29 is in the top 25% (Gignac & Szodorai, 2016); thus, the association between income and happiness is moderate to large. The difference in happiness between the lowest and highest deciles of income is also large (d = .71). Nor did we find a satiation point or a tapering off at higher levels of income as some have found using other datasets (Jebb et al., 2018; Kahneman et al., 2010); in the GSS sample, associations between income and happiness were steady and linear. Thus, at least in this sample, money and prestige do appear to buy happiness, and more is continuously better, especially in recent years.

This study was limited by several factors. First, the measure of happiness was a single item rather than a more psychometrically vetted multiple-item scale of well-being or mental health symptoms. Although we were able to use four measures of SES, all were self-reported rather than objectively verified. In some cases, n’s were too low for Black Americans to perform certain analyses. Our analysis was restricted to the U.S.; thus, we do not know if these findings extend to other nations (such as the UK) also rumored to have a growing class divide in wellbeing and other factors.

Finally, we were unable to definitively determine the causes of the growing class divide in happiness. However, the analyses controlling for marital status suggest the trend may be partially rooted in the growing class divide in rates of marriage. Although we did not directly examine marriage as a determinant of happiness, the reduced correlations with year when marital status was controlled are consistent with the general conclusion that marriage exerts a positive influence on happiness (Grover & Helliwell, 2019). However, controlling for religious service attendance did not substantially alter the correlations with year, suggesting that the increasing class divide in religious commitment is not the cause of the growing class divide in happiness.

Another possibility is that SES has become more salient as income inequality has grown; we were not able to address this possibility with this dataset, but this may be an avenue for future research. We found that the reasons for the growing class divide differed by race, with declines in happiness among low-SES Whites but little change for low-SES Blacks, and a rise in happiness among high-SES Blacks but little change for high-SES Whites. Although we can only speculate, it is possible that the increase in Black Americans’ educational attainment may play a role; increases in high school graduation rates likely lifted the prospects of non-college-educated Blacks, and increases in college graduation rates may have created a larger critical mass of college-educated Blacks, which may have had benefits for happiness. Declines in explicitly stated racial prejudice may also play a role (Marsden, 2012), especially if these declines primarily benefited high-SES Blacks.

The idea of a growing class divide was a common point of discussion after the 2016 U.S.  Presidential election (e.g., Galston & Hendrickson, 2016), and will likely continue to occupy the national conversation for years to come. If Americans without a college degree – who remain the majority of the population – are increasingly unhappy, politicians who promise change may be more attractive, and there will be growing class polarization in views of political candidates.  Growing dissatisfaction among those with less income may also make political systems such as socialism and government policies such as universal basic income more popular. Although happiness among Black Americans instead increased over time, the increases were limited to high-SES individuals, which may also prompt discussion of a class divides among Black Americans.

In a follow-up study to their findings on the growing mortality gap by social class in the U.S. (Case & Deaton, 2015), Case and Deaton (2017) concluded that much of the excess mortality among less educated White Americans was due to “deaths of despair” from drugs, alcohol, and suicide. The current results suggest that the growing class divide extends to broader measures of well-being such as general happiness. Between the 1970s and the 2010s, many of the “have nots” of the economy became increasingly unhappy, with potential downstream consequences for mental health, physical health, and the political climate of the nation.