Income comparisons and non-cognitive skills

What makes people happy? Is it enough to be rich in order to be happy? So far the literature has found that this does not seem to be enough. Individuals are rank sensitive: People’s happiness depends on whether they are richer than others. In a Barcelona Graduate School of Economics working paper (No. 733) (November 2013), Santi Budría and Ada Ferrer-i-Carbonell examine whether the rank effect differs between individuals with different personality traits and find a considerable heterogeneity on the importance of income comparisons. The authors are the first to relax the assumption that individuals respond identically to societal information regardless of their personality.

Previous literature

The happiness literature has shown that individuals’ personality is the most important component of individual heterogeneity (Boyce 2010) and researchers have acknowledged the importance of controlling for individual personality traits when estimating life satisfaction (Ferrer-i-Carbonell and Frijters 2004).

Data and measurement

In 2005 the German Socio-Economic Panel (SOEP) included a set of questions aimed at capturing various measures of personality: a short version of the Big Five Inventory (BFI), a set of questions to assess the degree of external or internal Locus of Control (LOC), and several items to capture individuals’ negative and positive reciprocity norms. The positive and negative reciprocity items relate to an important concept in social psychology aimed at measuring individuals’ negative and positive responses to other individuals’ actions. The Big Five and the LOC measures are two alternative well-known ways to describe individuals’ personality. LOC aims at capturing the degree to which individuals believe that the course of their life is under their control or depends on external circumstances, such as luck or social conditions. The BFI is a well-accepted measure to describe the five major traits that define human personality across cultures (Costa and McCrae 1992): openness, conscientiousness, extraversion, agreeableness, and neuroticism. To test for the importance of relative income, the authors use a life satisfaction question as a proxy measure for utility.

Empirical approach

The authors construct reference groups taking into account some individual characteristics as well as introducing a geographical dimension into the analysis. In concrete, they generate reference groups by partitioning the sample into various groups using the geographical region where the household lives (West or East Germany), the gender of the respondent, the education attainment of the respondent and the age of the respondent. While the reference group is defined at the individual level, the reference income is taken as the household income. Individuals are assumed to obtain information about the others through their own reference group, i.e., the authors assume that individuals generate information by looking at those similar to them.

Estimating procedure

Life satisfaction is assumed to be a function of personal characteristics and circumstances, such as age, education, gender, household characteristics, and health. Additionally the authors control for year and federal state fixed effects. As shown in the table, the results are supportive of the prominent role of some personality features as a mediating force on the importance of relative income for life satisfaction.

[table caption=”The effect of rank income on life satisfaction by personality trait” width=”550″]
,Top 25%[attr align=”left”],Average[attr align=”left”],Bottom 25%[attr align=”left”]
Big Five[attr class=”subhead” colspan=”5″]
Neuroticism,0.117 ***,0.039,0.054
LOC[attr class=”subhead” colspan=”5″]
External LOC,-0.027,0.078***,0.105**
Reciprocity[attr class=”subhead” colspan=”5″]
Positive reciprocity,0.088*,0.052*,0.052
Negative reciprocity,0.067,0.03,0.138***

In other words, the results show heterogeneity on the effect that comparison income has on happiness or life satisfaction. While some individuals tend to be rank insensitive others are much affected by their position in the income distribution. Partitioning the sample according to personality also yields some heterogeneity in terms of income: the marginal utility of income differs across personalities.

Implications and conclusions

The authors point out two implications of their findings. First, at the theoretical level, the distinction between absolute and relative formulations of utility has proven a useful concept to rationalize many unexplained phenomena in a variety of fields, including consumption, savings, growth and financial regularities. Acknowledging the extent of individual heterogeneity surrounding relative effects would prove fruitful to bring the theory closer to data and, most probably, to account for yet unexplained phenomena.

Second, welfare analysis should take into account the diverging importance of income externalities, for example, when designing optimal income taxation or defining poverty. The poverty literature has explicitly argued that relative concerns matter for individuals and some researchers have consequently defended that the poverty line should be based on relative rather than absolute consumption. Although in practice taking personality differences into account for relative poverty measures would be very difficult, the results warn that imposing a common benchmark might be seriously misleading.

To conclude, the authors remark that they find no important personality difference in the effect of other individual socio-economic or demographic characteristics on life satisfaction. It seems that personality plays a role in shaping individual life satisfaction and preferences only through some channels and for certain personality profiles. From the Big Five personality model, only extraversion is found to be a relevant dimension defining individuals’ sensitiveness to others’ income and achievements. However, locus of control (the extent to which individuals feel they are in control of things) and reciprocity are indeed found to be important determinants of individuals’ tendency to be rank sensitive.


Boyce, C. J. (2010). Understanding fixed effects in human well-being, Journal of Economic Psychology, 31, 1-16

Costa, P.T., & R.R McCrae (1992). NEO PI-R. Professional manual. Odessa, FL: Psychological Assessment Resources, Inc.

Ferrer-i-Carbonell, A. & P. Frijters (2004). How important is methodology for the estimates of the determinants of happiness? The Economic Journal, 114, 641–659.