martedì 8 giugno 2010

Are fat taxes a good idea? Three reasons why they are not.

The idea of a "fat tax" is again on the policy menu in the U.S., this time taking the form of a tax on sugary drinks.
Economist Greg Mankiw wrote a nice piece on the NY Times about the subject (here:, arguing that the rationale behind such forms of taxation, if any, lies in state paternalism. If people are deemed to be unable to properly care about their "future" self, either because of weak will or myopic evaluation horizon, then the government could enforce healthier behaviors through substitution effects induced by indirect taxes. The same line of reasoning, continues Mankiw, is then applicable to smoking, alcohol drinking, and many more "sinful" goods.

Generally speaking I feel inclined against sin taxes, for a number of reasons.

First, for the obvious lack of trust in politicians: are they really capable to obtain detailed data about demand elasticity and to evaluate welfare gains better than individuals? Direct observation of everyday's life and theoretical findings of Public Choice literature should be enough to answer "No way, sir!".

Second thought goes to people, who are not divided into just fat soda drinkers and healthy soda-dodgers. To clarify the matter let me describe a simple world where three kinds of citizens exist: the Healthy Guy eats and drinks in each period (month, week... it doesn't matter) exactly the right amount of fats, so he never accumulates (or looses) weight beyond the healthiest amount; then there's the Guilty Guy, who sometimes indulges in over-eating and accumulates fats, but when his weight reaches a threshold which is deemed to be dangerous for the "future self" of such individual, he engages in a diet which brings weight back to healty standards; and lastly, the Fat Guy who eats too much and doesn't manage to loose enough weight to be under the safe line and reasonably reduce heatlh dangers.

Now, a fat tax is supposed to help the Fat Guy to be less fat, by nudging him toward fat-free food and beverage, but what about the other two? Maybe the Healthy Guy consumes a soda from time to time, compensating the additional fats with other, fat-free food during his weekly diet. A fat tax would worsen his purchased basket (in terms of Paretian efficiency, he is certainly worse of) without making his health any better. Consider also the Guilty Guy: since his exposure to high fat accumulation only lasts for brief intervals of time, his additional health risks are probably low. A fat tax alters his consumption basket for a minimal benefit in turn.
We should also consider age groups. If Mankiw's argument about future self care is right, then what about elder people? Fat accumulation does not increase risks of hearthstroke or other health risks immediately, but it takes some time to build up dangerous levels of cholesterol in blood veins. A fat tax moves elders away from their preferred consumption basket, but the obtained benefit for the health of their future self might be really too low to be of any interest to anyone.

Third thought is about the effectiveness of fat taxation. Is fat food demand elastic enough? Some authors offer compelling evidence that it might be not (see for example Chouinard et al., "The Effects of a Fat Tax on Dairy Products", link). Inelastic demand would mean fat taxation would obtain additional tax revenues without altering people's behavior. If this is the case, then we are going to introduce a new indirect tax for revenues needs, and the paternalistic rationale is just the political tool to attract supporters.

It seems to me these three reasons are more than enough to refute the basis upon which fat taxes, and generally sin taxes, are built and supported. Only externality arguments make some sense, for example in case of smoking in public indoor areas (but, are "sin taxes" the best tool to deal with this kind of externality?). Realistically, what kind of negative externality can we reasonably attach to fat people? Honestly, such weak rationales should never be the foundation of any reasonable policy action.

Nessun commento: