Following up on Brian’s post: Let’s imagine a city with two families, the Riches and the Poors. The Riches earn $100,000 a year and the Poors earn $10,000. Each household consumes two 12-ounce sugar-sweetened beverages a day, and the city imposes a 1.5-cent-per-ounce tax on sugar-sweetened beverages. 365 x 2 x 12 x $0.015 = $131.40. Assume no behavioral adjustments (i.e., the tax doesn’t reduce sugar consumption by anyone). Also, assume full pass-through of the tax (i.e., each household bears the entire $131.40 out-of-pocket cost.)

Is the sugar-sweetened beverage tax regressive? On first glance, yes: the Riches pay $131.40/$100,000 = 0.13% of income, and the Poors pay $131.40/$10,000 = 1.314% of income. A tax that takes a higher percentage of income from the poor than the rich is the very definition of regressive.

But now assume that the revenues raised from the tax ($262.80/year) go entirely to fund universal pre-K. And since the Riches already send their children to private pre-K, let’s assume that all $262.80 is going to pre-K for the children of the Poors. Now is the soda-tax-plus-pre-K package “regressive”? Assume that the benefits of pre-K to the children of the Poors are equal to the amount spent ($262.80) (although research by James Heckman and others suggests that the benefits of pre-K are actually multiples of the amount spent). So the soda-tax-plus-pre-K package leaves the Riches $131.40 worse off and the Poors $131.40 better off (minus $131.40 for the cost of the soda tax plus $262.80 for the benefits of pre-K). Is a policy that makes the rich worse off and the poor better off in after-tax terms “regressive”? Seems hard to argue that it is.

Read more at Whatever Source Derived