Conspicuous Consumption, Conspicuous Health, and Optimal Taxation

Is there a health-status race in Australia whereby people get joy from being healthier and fitter than others? And what are the general implications for public policy if there is? My PhD student Redzo Mujcic and myself brought out a new working paper recently on how a health status race can be good for the public purse and furthermore reduces the case for taxing work in order to give people an incentive to take more leisure time.

The abstract:

We present a simple model of status-seeking over multiple socioeconomic domains by introducing the concept of conspicuous health as an argument in the utility function, in addition to the well-established conspicuous consumption term. We explore the implications of such a utility function for optimal income taxation, where we show an increase in concerns for conspicuous health to have an opposite effect on the marginal tax rate, compared to an increase in concerns for conspicuous consumption. Using life satisfaction panel data from Australia, along with an improved measure of exogenous reference groups (that accounts for the time-era of respondents), we find evidence of a comparison health effect.

Author: paulfrijters

Professor of Wellbeing and Economics at the London School of Economics, Centre for Economic Performance

3 thoughts on “Conspicuous Consumption, Conspicuous Health, and Optimal Taxation”

  1. It looks like you put this here to get some comment about it, but no one is biting.

    I have read the paper. Obviously I agree that people compare their standard of living to others and that relative status is important. However this paper doesn’t seem to add much to the debate.

    Let’s just think about the utility function you describe. It predicts that if my wife’s health deteriorates dramatically I get a utility boost. And if my friends and other family also get fat, catch a disease etc, then my utility shoots up (ok, there are lots of other things that enter the utility function when close family get ill, but you get my point).

    It also suggest to me that the reference group is a choice – I gain utility from choosing to compare to less healthy individuals. Wouldn’t this lead to healthy people to seek out unhealthy people to befriend in order to feel better? Isn’t the reference group itself a choice? I can imagine the cliche situation of the pretty girl befriending the ugly girl at high school to make herself feel better. But isn’t the general finding that birds of a feather flock together? Healthy people are clustered together in the social network? Is that what this utility function predicts?

    You seem to be presenting an ‘inverse Schelling model of segregation’, where people have a slight preference for others to be less healthy. The equilibrium outcome would be very mixed rather than clustered, so that there are no gain from shifting to a new reference group.

    The optimal taxation section is strange to me – what reality is this trying to represent? Do you really think people make decisions about labour effort in response to relative health? To ask the question you like to ask, what are the policy implications?

    I’m suspicious that the negative coefficient for the reference group income and health is merely a statistical artefact capturing that healthier people are more likely to be above average health in their reference group (and vice-versa). Meaning even if life satisfaction was only dependent on own health, you would get this result by construction of the data (a negative relationship between own health and reference group health).

    Doesn’t Table A2 essentially show it doesn’t really matter who the reference group is?


    1. Hi Cameron,

      thanks for the effort. Yes, the utility function is a work-horse version and hence would have bizarre implications in various circumstances we dont use the model for. We for instance dont really identify it off people losing their partners, but rather as to what happens to people like themselves! Endogenous partnering in this case also wont matter for our results, though the assumption of a particular form of reference group (the cell means) is quite important for the results and not ideal, but that is a general problem for lots of studies into reference groups. A case of best-one-can-do-at-this-moment I am afraid. Can you do better?


      1. “Can you do better?”

        Well, I’m not sure that there is a way to do better using this methodological approach. Aggregation of utility functions into expected patterns in aggregate data is always a problem – much more so once we introduce relative concerns in utility function.

        In a more general sense the idea of relative concerns in any metric introduces the possibility of an arms-race-type process, which itself is an interesting phenomena. However it doesn’t seem conducive to analysis by comparative statics.

        I guess my approach would be to start with the data and look at changes in health rather than levels, and see if, for example, a change in health from one person leads to an increase in health effort by those in the reference group. There are plenty of sociological studies that adopt this ‘signal-response’ or ‘cascading behaviour’ approach. For example, there is a study showing people are more likely to purchase a new car if their neighbours have recently purchased one.

        Is this what we expect from a utility function? We don’t really know, because we don’t have a good story about aggregation and reactions by heterogeneous agents.

        Is there a way to test if health arms races are observable in the HILDA data?

        Could we potentially use a large charity fun-run as a natural experiment on the impact of non-runner’s investment in health activity? Aren’t people always arguing that such health externalities are a good reason for these activities?

        I will think about a research approach that could work here.


Comments are closed.

%d bloggers like this: