Taxes play a number of roles. First, they are a way to raise revenue to finance public goods. Maintaining law and order or invading another country is not cheap and it is easier to finance such projects by compulsory levies than by voluntary contributions. Second, taxes can be used discourage undesirable behavior. Many countries, for example, levy ‘sin’ taxes on alcohol and tobacco to discourage their consumption and thus to reduce their harmful effects. Finally, taxes redistribute income from those who are (relatively) rich to those less fortunate. The taxes that we all pay reflect these various objectives: virtually all of us pay taxes but those of us who are relatively well-off and those who engage in ‘sinful’ activities such smoking, drinking or driving a car tend to pay more, sometimes a lot more, than others.
The current tax system, however, is far from perfect. Consider redistribution. Surely, it is only fair to tax the rich and subsidize the poor. Life is a lottery in many respects. If we care about equality, we should therefore redistribute so that those who happen to have drawn the short straw are compensated for their bad luck. This argument, however, only holds if distribution of earnings is sufficiently random. Earnings are not random, however: they reflect one’s wage (which may be, to some extent, determined by luck) as well as effort (broadly defined to include both the amount of hours that one works as well as their investment into education and other human capital). If we tax effort, we will discourage people from working and acquiring human capital. Therefore, a far and egalitarian tax system should tax the income-earning ability but not effort.
The problem is that it is not always easy to tell what part of one’s income is due to luck or effort. We do know, however, that some people tend to earn more than others. It is well know that men tend to earn more than women and whites tend to paid better (and find jobs more easily) than members of ethnic minorities. There is more: a well-known paper by Daniel Hammermesh and Jeff Biddle (American Economic Review, 1994) found that attractive people get higher wages. Another study, by David Jonhston (Economics Letters, 2010) concludes that women with blond hair get paid more than women with hair of other colors. And, finally, tall men and women also earn more; this is in addition to enjoying other benefits such as having higher education and being on average more confident, happier and less likely to suffer various psychological problems (see the discussion in my own paper with Michèle Belot in Economics and Human Biology, 2010).
Therefore, a fair and egalitarian tax system should take into account characteristics that are related to tax-payers’ ability to earn high wages. Beauty is highly subjective and hair can be dyed. Height, in contrast, does not change during a person’s adult life and can be measured easily. It would make, therefore, an ideal criterion for setting tax rates. This insight is not mine: a similar idea was originally formulated by George Akerlof (American Economic Review, 1978) and a recent article by Gregory Mankiw and Matthew Weinzierl (American Economic Journal: Economic Policy, 2010) fleshes this proposal out.
Not surprisingly, this idea is controversial: one person that I tried to convince about its merits even likened it with the Nazi predisposition to decide one’s fate based on their physical characteristics. Yet the labor market puts a premium on certain physical characteristics, even on those which (unlike education or hair color) workers can do little about. Taxing according to height would only correct for this inherent unfairness of earnings distribution. Under the present system, two persons with the same income usually pay the same tax. Under height-dependent tax, the taller one would be taxed more. This is fair because, holding everything else constant, the taller person has an easier time to earn a given level of income while the shorter person has to work harder for it.
Now how about the role taxes play with respect to discouraging undesirable behavior? One particularly pressing issue these days is the increasing prevalence of obesity. This has grave consequences: obese people are significantly more likely to suffer from heart disease and diabetes. This, in turn, is costly for the public purse (plus it lowers the wellbeing of obese people and their families). The increase in obesity rates reflects unhealthy life style choices: mainly caloric intake that is too high and too little exercise. We already have so-called sin taxes on alcohol and tobacco that seek to remedy this. In the UK, the VAT rate is higher (20%) for sugary drinks and salty and fatty snacks than for other food (0%). A recent Finnish paper (“Health and distributional effects of differentiated food taxation,” by Kaisa Kotakorpi, Jukka Pirttilä, Tommi Härkänen, Pirjo Pietinen, Heli Reinivuo and Ilpo Suoniemi) argues that taxing sugar at a relatively modest rate of 1€ per kilogram would result in a 6% reduction in the prevalence of type-2 diabetes as well as lower incidence of heart disease in Finland. They also argue that these higher sin taxes should be complemented by lower taxes on healthy food such as fruit, veg and fish (which is already the case in the UK).
Sin taxes, however, are a blunt instrument. Taxing sugar would also affect the prices of other foodstuffs (bread, for example). More importantly, this change would hit equally those who are obese and those who are not: two persons can have very similar sugar intakes and yet have very different BMIs (body mass index, a common measure of obesity, computed by dividing one’s weight in kilograms by the square of their height in meters), depending, for example, on how much exercise they do. Taxing them based on their BMI, therefore, should be preferable to taxing sugar, fat and/or alcohol content of foodstuffs and beverages.
Income tax should therefore depend on how far one is from the median height (1.75m for men and 1.62m for women in the UK) and also on how far their BMI is from the ideal range of 20-25. One problem, however, is that a person’s current BMI reflects their life-long choices. Therefore, linking income tax with the level of BMI would unfairly penalize them for their actions before this reform. To avoid this, the tax could be levied according to one’s height and the annual change of BMI. Variations in one’s BMI within the 20-25 range should be ignored, while those with BMI above 25 should be taxed more if their BMI increases. This obesity penalty should be further increased for those with BMI over 30 (the threshold between being overweight and obese).
If this idea catches on, being audited by the tax authorities may soon mean that you would need to strip down to your underwear to have your height and weight checked.
Jan Fidrmuc (height 1.83m, BMI 29.6)