Positional income concerns and personality: evidence from Germany

ABSTRACT

Positional income concerns (PIC) strongly influence economic behaviour and life outcomes. However, very little is known about the underlying heterogeneity regarding their importance across individuals on the one hand and across different reference groups on the other. Our analysis builds on representative survey data reporting PIC vis-à-vis seven reference groups, allowing us to identify marked heterogeneity both between and within subjects. We present evidence about who (in terms of personality) compares to whom (in terms of reference group).

KEYWORDS:

• Relative income
• status
• personality
• Big Five
• SOEP

JEL CLASSIFICATION:

• D03
• D12
• D63

Acknowledgment

We thank Thomas Dohmen, Laszlo Goerke, Clemens Hetschko, Florian Hett and the participants of the 6^th Economics Workshop in Trier, the 11th German Socio-Economic Panel User Conference in Berlin, and the 26^th EALE Conference in Ljubljana for helpful comments and discussions on the topic.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Previous research usually either imposed one specific reference group (e.g. Ferrer-i-Carbonell 2005) or asked about comparisons to ‘other people’ and then inquired about the most relevant reference group (e.g. Clark and Senik 2010; Clark, Senik, and Yamada 2013). The study closest to our endeavour in this regard is Dahlin, Kapteyn, and Tassot (2014), who report survey results for different reference groups being particularly interested in whether geographically proximate peers are important. Senik (2009) relies on data from the Life-in-Transition survey for Eastern Bloc countries which deal with comparisons to different benchmarks (e.g. past living standards, parents and former schoolmates). Goerke and Pannenberg (2014; 2015) use the same data, reporting intuitive relationships between PIC and happiness, and labour supply.

² For more information about the SOEP, refer to Wagner, Frick, and Schupp (2007) or visit http://www.diw.de/en/diw_02.c.299771.en/about_soep.html.

³ Our subject restriction is due to the fact that the questions about relative income comparisons were addressed only at employed respondents in the pretest 2010.

⁴ The exact wording of the income comparison question was slightly refined over the years. However, the basic structure and their position in the questionnaire were unaffected.

⁵ Cobb-Clark and Schurer (2012) provide evidence that these measured traits are stable over time and conclude that the Big Five measures may be considered stable input into economic decisions.

⁶ Stacking all comparison intensities for each respondent into a panel data set for a regression analysis which includes personality traits scores as well as dummy variables for the reference groups, it turns out that an interaction term of agreeableness and ‘coworkers’ significantly reinforces the negative level effect of agreeableness.

⁷ The results of the previous literature regarding gender and status effects have been ambiguous (e.g. Dohmen et al. 2011).

⁸ For example, some subjects may refrain from indicating high comparison intensities and express the differences across reference groups by using only numbers less than five. Focusing on the within-subject dimension, we can address this potential issue.