Subjective well-being, peer comparisons and optimal income taxation

David Ulph, Sean Edward Slack

    Research output: Working paperDiscussion paper

    2 Downloads (Pure)


    Empirical evidence suggests that an important determinant of subjective well-being is how an individual’s consumption compares with that of their immediate peers. We introduce peer comparisons into the standard optimal tax framework and demonstrate that the optimal linear tax expression is adjusted in three key ways, the latter two of which are novel to this paper and act to lower the tax rate. First, the dependence of well-being on peer income introduces an externality that distorts labour supply above that which individuals would choose were they to recognise the interplay between their own choices and the Nash equilibrium level of peer consumption. The optimal tax rate is adjusted upwards to (partially) correct this distortion. Second, if individual labour supply is a function of peer consumption, there are ‘Keeping up with the Joneses’ multiplier effects that raise the Nash compensated labour supply elasticity above the individual labour supply elasticity. This implies a lower tax rate on efficiency grounds. Third, Nash indirect well-being is decreasing in the wage rate for workers with wages close to the reservation wage. To the extent that this lowers the covariance between gross earnings and the net social marginal value of income, this will act to lower the optimal tax rate.
    Original languageEnglish
    Place of PublicationSt Andrews
    PublisherUniversity of St Andrews
    Number of pages36
    Publication statusPublished - 23 Nov 2016

    Publication series

    NameSchool of Economics and Finance Discussion Paper
    PublisherUniversity of St Andrews
    ISSN (Print)0962-4031
    ISSN (Electronic)2055-303X


    • Optimal taxation
    • Relative consumption
    • Subjective well-being


    Dive into the research topics of 'Subjective well-being, peer comparisons and optimal income taxation'. Together they form a unique fingerprint.

    Cite this