Individual values and household finances

ABSTRACT

We create individual cultural values measures for households and show that this is an important determinant of their financial behaviour. To date, personal cultural values have only been indirectly measured through religion and trust. But these are, at best, an approximation of true cultural values. Applying a holistic framework from the World Values Survey (WVS), we create individual measures of cultural values, and show that the self-expression values of this framework are positively associated with households’ financial decisions. Examining the individual cultural values that make up the WVS model, we further show that happiness, trust, and playing an active role in society, are individually important determinants of household financial decision-making. Our study shows that cultural values can be brought from a generalized national level to the individual level in order to improve our understanding of household financial decision-making.

KEYWORDS:

• Household finance
• cultural values
• religion
• World Values Survey

JEL CLASSIFICATION:

• D14
• G11
• Z1

Acknowledgements

The authors thank CentERdata for the openness with which they share their data sets. Financial support from the Business School of Dublin City University is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The decline is expected to continue, with one estimate from the Pew Research Center suggesting nearly 100 million less Christians in Europe by 2050 compared to 2010 (Hackett et al. 2015).

² The LISS Panel, or the Longitudinal Internet Studies for the Social Sciences Panel, is an EU-funded Internet-based project based in Tilburg University where a representative sample of the Dutch population are regularly interviewed on a range of socio-economic topics.

³ The survey waves are conducted approximately every 5 years in order to capture changes in values across the world.

⁴ For more technical details on the construction of the data set, see Scherpenzeel (2011).

⁵ A complete list of the many publications using LISS Panel data is available at: http://www.lissdata.nl/dataarchive/publications.

⁶ The number of observations for the investment data is limited because the question is often left blank. The survey tries to encourage people to give at least approximate answers by offering categories if the questions are left blank.

⁷ Following Inglehart and Baker (2000), we use 10 items to construct the dimensions instead of the originally proposed 22 items (Inglehart 1997). Inglehart and Baker (2000) show that the dimensions can be accurately constructed from this reduced list of 10 items. More details on the construction of the two value dimensions can be found in the Internet Appendix of Inglehart and Welzel (2005).

⁸ The variables abortion, homosexuality, and distrust are coded in a way that is counter-intuitive. Abortion, for example, reflects households’ ‘disapproval’ of abortion. Homosexuality and distrust are coded in a similar way. For more details, see Table 1.

⁹ For clarity, a finding of significance for attitude towards homosexuality would not suggest there is a link between this attitude and savings behaviour. Instead the finding would be supportive of an argument that this attitude is indicative of an underlying value system and it is this underlying belief which is related to savings behaviour. In our case, the underlying value system is captured by the overall value dimensions of the WVS.

¹⁰ It would have been interesting to test other religious adherents, particularly Muslims due to the strict guidance their religion gives on financial behaviour, but the WVS only has a weighting of 2.1% Muslims which means there are too few Muslim households in the sample.

Additional information

Funding

This work was supported by Business School of Dublin City University.