There are two ways to gauge these magnitudes. The first is to contrast the impact of excise taxation to other predictors of happiness. For example, the results in columns and suggest that being a predicted smoker with no excise tax raises unhappiness by 7.5 percentage points in the U.S. and by 9.6 percentage points in Canada. Thus, our findings suggest that a 50cent real excise tax on cigarettes would leave predicted smokers as happy as those not predicted to be a smoker in the U.S., and that a $2.00 real excise tax would have that effect in Canada.
Alternatively, we find that, conditional on all other Xs, being in the top income quartile reduces unhappiness by about 7.5 percentage points (relative to the bottom income quartile) in the U.S., and by about 6 percentage points in Canada (incorporating the impact of being in the top quartile of both family and personal income). So a 50-cent (in the U.S.) or $1.33 (in Canada) excise tax would have the same effect. In other words, such an excise tax level would be equivalent to moving a predicted smoker from the bottom to the top income quartile.
Both of these exercises imply very large impacts of excise taxes on happiness. But one difficulty with these types of comparisons is that the effect of the X variables themselves on happiness may not be well identified. While the impact of cigarette taxes on the happiness of predicted smokers is, we argue, a well identified relationship, the same cannot necessarily be said of the impact of factors such as income on happiness; those who are richer may be fundamentally less happy for other reasons, for example, understating the impact of income on happiness. A better comparison may be to consider what these results imply for the implications for happiness of reducing smoking. Estimates of the impact of excise taxes on tobacco expenditures are generally in the range of -0.5, although Gruber and Koszegi obtain a higher elasticity of roughly -0.6 using more recent data. Gruber, Sen and Stabile estimate an elasticity for Canada of -0.45. read more
These estimates suggest that each 10 cent increase in price leads to a 6% decline in smoking in the U.S. (given the base average real price of 97 cents over our sample period) and a 3% decline in smoking in Canada (given the base average real price of $1.67). Our happiness regressions suggest that this tax rise is also associated with an decrease in 1.5 percentage points of happiness amongst all those who are predicted smokers in the U.S., or roughly 10% of baseline unhappiness among smokers, and 0.46 percentage points among those who are predicted smokers in Canada, or roughly 7.6% of baseline unhappiness among smokers. Extrapolating, then, these findings suggest that reducing smoking by 60% would fully remove unhappiness among smokers in the U.S., and that reducing smoking by 40% would fully remove unhappiness among smokers in Canada.
These implied effects are once again quite large.