Washington is one of the few states that does not have an income tax. Instead, it has a hefty sales tax. Every few years politicians dust off the old state income tax playbook and try to sell it to voters as an alternative to the sales tax.
The last time this happened was 2010, when we voted on I-1098, which was a voter initiative to lower real property taxes, eliminate the B&O tax (a tax that businesses pay), and impose a state income tax. Voters resoundingly rejected it.
Proponents of I-1098 argued (among other things) that the sales tax is “regressive,” meaning that it hurts poor people more than rich people. But when measuring who is poor and who is rich across an entire population – in order to argue whether a tax is fair – it gets complicated quickly.
The United States Bureau of Labor Statistics’ 2010 Consumer Expenditure Survey helps us out. The CES tells us on average how much money people made by dividing them into five separate groups from the poorest twenty percent to the richest twenty percent by average income, and then tells what they bought. It is available online at http://www.bls.gov/cex/2010/Standard/quintile.pdf so you can do the math yourself.
According to the CES, the poorest twenty percent (measured by income) earned an average $9,906 in 2010. The top twenty percent earned an average of $157,369. When we talk about who is rich, and who is poor, this is who everyone is usually talking about.
Things get interesting when we measure who is rich and who is poor by what is consumed in a year rather than what is earned. Remember the poorest quintile, who averaged $9,906 in earnings? In that same year, their expenditures averaged $20,953. This group consumed twice as much as they earned! (I’ll explain this anomaly later.) The richest twenty percent, who earned averaged $157,369 in earnings, spent $92,870.
So how can the poorest twenty percent, the group that earned $9,906, spend twice as much as it earned? Borrowing may account for a small portion of that figure, but more likely the culprit is the ultimate consumption smoothers: retirees. Many of these people have low earnings but live from their savings. They show up as low income but may not be “poor” in the traditional sense. Simply classifying them as poor because of low earnings is misguided.
Income in any given year is not a good measure of who is rich and who is poor. This is because income for a lot of people is dynamic over time. Measuring a group’s average expenditures in a given year may be a better measure of wealth because people tend to smooth their consumption over time, even as their income varies. (By the way, this revelation, seemingly obvious, helped win Milton Friedman a Nobel prize in economics.)
Labeling the sales tax a “regressive” tax is an oversimplification. Whether it is regressive, flat or progressive depends on the specific goods that are subject to the tax, and who tends to buy them. For example, a sales tax on luxury yachts is not regressive because poor people do not tend to buy luxury yachts. But what about food? Cigarettes? Or gas?
Let’s examine a hypothetical sales tax – say ten percent – on “gas and motor oil” and see what it does to these groups. The poorest twenty percent spent an average $1,009 on gas and motor oil in 2010. The wealthiest twenty percent spent $3,240. Doing the math, a ten percent sales tax on gas and motor oil equals one percent of the poorest quintile’s average income, but only two-tenths of one percent of the richest peoples’ incomes. This hypothetical tax takes a larger percentage of the poorest quintile’s average income than the richest quintile's. This is a regressive tax, at least when we categorize who is poor and who is rich by income.
The results are even more staggering if the tax were imposed on tobacco products. In that case, the poorest quintile spent a lot more money on tobacco than did the rich. In other words, a smoking tax is really regressive.
The result may not be so staggering on food or rent. Groceries (with a few exceptions) are not subject to a sales tax in Washington State. Neither is rent.
The same ten-percent gas tax, when calculated as a percentage of total expenditures, rather than income, proves the tax is nearly flat. It equals fourth-tenths of a percent of the poorest quintile’s total consumption and three and one-half tenths of the richest quintile’s total consumption. In other words, in terms of expenditures, our hypothetical tax hits both groups nearly the same.
So, is the sales tax regressive? It depends on what goods are taxed and who buys them. Attaching the “regressive” label to a sales tax requires a measurement of who is rich, and who is poor, and what types of goods those groups buy and how much. Measuring that is not as simple as many would think.