Will Higher Taxes on Gasoline Lead to Higher Government Spending?

One of the big objections to the Pigou Club idea of raising taxes on polluting activities is that governments will not use the money to off-set other taxes. Rather, the added revenue will allow the government to waste even more money on pork-barrel projects. This is related to the idea of the Starve the Beast theory of government spending, where tax cuts today will cause spending cuts tomorrow. If we raise additional tax money today, won't government just spend it all tomorrow?

Examining the Evidence of the Starve the Beast Phenomenon

I decided to examine the literature and see what evidence there was for the starve the beast phenomenon. I found three recent articles on the subject:

• "Starve the Beast" Does Not Work by William A. Niskanen of the Cato Institute (a PDF summary of the article is available here)
  
  
• Jeffrey Frankel comments on: The Euro, Stabilization Policy, and the Stability and Growth Pact by Adam S. Posen (PDF available here)
  
  
• The "No New Taxes" Pledge by William G. Gale and Brennan Kelly of the Brookings Institution (PDF available here).
  

All three articles find the same thing - that there is absolutely no statistical evidence for the "starve the beast theory". What recent evidence we have actually supports the opposing theory, that is increases in taxes lead to lower government spending, and vice versa.

What Do the Statistics Say on the Starve the Beast Theory?

Gale and Kelly give an excellent breakdown of the experience in the United States between 1981 and 2004. In their words:

...Recent history displays exactly the opposite pattern as predicted by "starve the beast" theorists; even after controlling for the business cycle, changes in spending and changes in taxes are negatively correlated over three major periods since 1981:


• Between 2000 and 2004, standardized federal revenues fell by 3.5 percent of GDP, but standardized total federal spending rose by 1.4 percent of GDP and standardized federal non-interest spending rose by 2.3 percent of GDP (only half of which was due to increased defense and homeland security...)
  
  
• Between 1992 and 2000, standardized federal revenue rose by 1.8 percent of GDP, but standardized total federal spending fell, by 2.4 percent of GDP and standardized non-interest spending fell by 1.5 percent of GDP.
  
  
• Between 1981 and 1992, standardized federal revenues fell by 2.0 percent of GDP, while standardized federal outlays rose by 0.6 percent of GDP.
  
  
All of patterns above are inconsistent with the "starve the beast" view of revenues.

If that were not enough evidence, Gale and Kelly describe the behavior of particular lawmakers. They found that lawmakers who signed in favor of "large and permanent tax cuts" were more likely to also vote in favor of large spending bills, such as the Medicare prescription drug bill, which they describe as "the largest entitlement increase in decades".

The statistical evidence, for the United States at least, looks particularly clear - if there is a relationship between government spending and tax levels, it works in the opposite direction of what the "starve the beast" theory would indicate.

Why the Starve the Beast Theory Fails

The starve the beast theory is intuitively attractive - I must admit that for a long time I believed it. But even a cursory examination of the spending patterns of the U.S. administration would cast doubts on it, and a more rigorous examination does a pretty thorough job of discrediting it. So why does such an intuitively appealing theory fail so miserably in the real world?

Niskanen gives the following interpretation of the data:

"What is going on? The most direct interpretation of this relation is that it represents a demand curve--that the demand for federal spending by current voters declines with the amount of this spending that is financed by current taxes. Future voters will bear the burden of any resulting deficit but are not effectively represented by those making the current fiscal choices. One implication of this relation is that a tax increase may be the most effective policy to reduce the relative level of federal spending. On this issue I would be pleased to be proven wrong."

The Frankel argument (as edited by Posen) is presented slightly differently:

"What is the mechanism through which the Starve the Beast approach is in theory supposed to restrain spending? The mechanism is that if you create huge deficits, citizens will worry so much about the national debt that they will come complaining to their Congressman: "I'm worried about raising taxes on my grandchildren." The Congressman will then be less likely to vote higher spending. Maybe people worry about the national debt, about taxes on their grandchildren. But surely they don't worry about such uncertain prospective future taxes... more than they worry about certain taxes today... Unpopular taxes today must put more pressure on Congressmen. Thus as a political economy argument, Starve the Beast just doesn't make sense, if the alternative is the regime of the 1990s."

My belief is that lowering taxes and raising spending are complements and not substitutes, so any theory must that into account. The following Gale and Kelly argument makes a lot of sense to me:

"The data above suggest that policymakers go through periods of fiscal restraint and fiscal largesse and the restraint or largess occurs simultaneously on both the tax and spending sides. That is, periods of fiscal largesse tend to generate declines in taxes and increases in spending (as shares of GDP). Periods of fiscal discipline tend to provide declines in spending and increases in taxes."

Implications on the Pigou Club's Proposals

It should come as no surprise that I believe this evidence supports the aims of Pigou Club. Pigovian taxes can be used to off-set income taxes, but a portion can also be used as a form of 'belt-tightening' to reduce the massive U.S. federal deficit and to reign in excessive pork-barrel spending. Higher Pigovian taxes today can herald in a new period of fiscal restraint and remove the need for higher income taxes in the future.