[kissickas]

> Total sin tax burdens are poorly explained by demographics (including income).

While two most-taxed clusters may be low-income, the researchers found a poor correlation between income and sin tax burden overall.

More detail from the paper:

> The second takeaway is that saying “sin taxes are regressive” or “sin taxes are progressive” largely misses the point. There is much more variation among households within income groups than across them in purchases of sin goods (Figure 3). Even among the lowest-income groups, the majority of households pay negligible amounts of sin taxes, and there are heavy smokers and heavy drinkers at all levels of education and income.