Our primary finding is that most state and local tax systems take a much greater share of income from middle- and low-income families than from the wealthy. That is, most state tax systems are regressive.
What they show is that while the share of income that goes to pay income taxes rises with income, the share of income that goes to pay property taxes is roughly constant with income--until you get to the top 1 percent of income, which pays a lower share of income as property taxes. The share of income that pays sales and income taxes declines steadily with income.
The authors interpret the patterns of sales and property tax burdens as showing that the tax systems are regressive. However, an alternative interpretations is that property holdings and consumption are better measures of permanent income.
It would be interesting to redo the study using consumption rather than income as the denominator. My guess is that the ratio of sales and excise taxes to consumption is lower for individuals with low levels of consumption than for individuals with high levels of consumption. If so, then when consumption is the indicator of economic well-being, it might turn out that sales taxes and property taxes are progressive. By the same token, income taxes might be less progressive than they appear in the study, because some of the high incomes taxes are paid by people with low permanent incomes, and vice-versa.
For Discussion. Using the approach in the study, would an income tax with a large exemption for savings likely be regressive?