Consumption versus Income Taxes

3/20/2013 03:39:00 PM
Dylan Matthews has a post on republican proposals in several states to replace income taxes with sales taxes instead. Matthews notes that governors in Louisiana, Nebraska, and Kansas have all proposed eliminating income taxes and hiking sales taxes. He seems to have missed that Governor Kasich of Ohio has announced a similar plan, though in the past few days it has looked like the state legislature may not approve it.

I will be coauthoring some research on this topic in the near future, though we just started so no results to report here yet. A couple highlights from Matthews:

"Nicholas Johnson, director of state policy at the Center for Budget and Policy Priorities, explained the first time Jindal floated this idea, state sales taxes aren’t particularly good consumption taxes" [emphasis added]
This is an important point. On the one hand, state sales taxes actually exempt most forms of consumption, including most food and all services. Many states, such as Ohio, actually have constitutional provisions that prevent things like food from being taxed, prohibiting them from even having authority to institute a true consumption tax. On the other hand, sales taxes have a tendency to tax a lot of things that are actually investment rather than consumption--this is the case whenever the sales tax applies to a business-to-business transaction. Most economists would agree that distorting investment incentives is worse than distorting labor incentives, so the potential upside of a sales tax is pretty murky, at best.

"Economists generally prefer consumption taxes to income taxes"
Not really. Some old-school economists do prefer consumption taxes. But modern taxation theory suggests that it is rarely optimal to distort consumption incentives, and that taxing income is generally the most efficient way to tax highly heterogeneous households, because income is highly correlated with workers' type, while consumption is not.  I will have more on this in later posts.

"The academic literature is divided on the question."
Yeah, that's a more accurate description of where economists stand.

In a later post, I will discuss the theoretical reasons why some old-school economists still prefer consumption to income taxes. I will also try to write a post with a simplified version of modern tax theory, explaining why consumption taxation is unable to maximize social welfare. But the basic story is this: the error in the old-school tax literature was the assumption that revenues are only used to finance government consumption and investment expenditures, with no transfer payments. The Ramsey models don't even have a way of incorporating transfer payments into the framework. When you recognize the reality that a large share of revenue goes to transfer payments, then you get the result that it is rarely optimal to distort consumption incentives, and that income taxes are the optimal tax instrument for redistribution. Now, its true that transfer payments are more of an issue for the federal government than the states, but it remains the case that state consumption taxes are counter-productive, since consumption taxes ultimately redistribute from poor to rich--abolishing state sales taxes would reduce the amount of federal transfer payments, allowing us to cut income taxes across the board.