Tax Cuts, Tax Revenues, and Growth: Reconciling Theory with Evidence

Yesterday I made a post about how federal tax revenues have not increased as a result of the Bush tax cuts from 2001. Actually much of the post is based upon an analysis of official data that David Cay Johnston did.  In the comments though, William Sullivan asks:

I realize that economic growth, as measured across the last decade, is anemic at best. I think it is abundantly clear that our employment situation is awful. What I don’t understand is how these issues can be attributed to lower taxes. I recognize that ECON 202 is not the end of economic learning but Johnston’s statement seems to be at odds with what we’ve learned.

I thought his question was excellent. It’s the type of question people and principles students should ask in response to an article.  Since I think my response will be useful for future students I decided to answer Mr. Sullivan with a new post (I can find it easier in the future then).

Initially, a quick read of both my post and Johnston’s article does leave the impression that we’re saying “tax cuts don’t stimulate the economy”. Yet we (I) teach in principles of macro classes that “tax cuts are one way of implementing fiscal policy to stimulate the economy”.  So how do we reconcile the theory with the evidence?

Here’s how I do it.  We need to be more precise in both what’s being measured and in what claim or theoretical assertion is being tested.  First let’s look at the linguistics of the three variables of interest here. The controversy or confusion is about the relationship between these 3 variables are:

  • tax rates – the % of income that by law needs to be paid as income (in more sophisticated analyses we should distinquish between marginal rates, the rate paid on additional income vs. average or effective rates, which is tax $ paid divided by income. For now, tax rates are sufficient.
  • tax revenues – the total dollars of taxes collected by the government in a particular year.
  • GDP – the total size of the economy.  Since GDP is roughly equal to gross domestic income, we can basically say this is what the economy earned each year. For our purposes here, we’ll call this “the economy”.

The first issue is lazy semantics that politicians and sometimes economists often use.  They often will use the phrase “tax cuts”. But what does “tax cuts” mean?  We always have to careful about what’s meant when somebody says “tax cuts”.  It could mean “reduction in tax rates” or it could mean “reduced tax revenues”.  We really have to look at context to see what’s meant as I’ll illustrate a little later.

Now the next question is what are the different propositions or assertions that are made about the relationships between these variables.  The first relationship is the purely accounting, definitional relationship and shouldn’t be controversial at all. Namely, in any given year, the average tax rates times GDP = tax revenues.  This is really the idea that taxes are basically income taxes. We earn income as a nation (GDP) and we pay a percentage (tax rates) of it to the government (tax revenues).

Now let’s look at the claims that have been asserted.  Roughly put, there are two separate questions or assertions to be tested here:

  1. Lower tax rates will stimulate the economy (GDP) so much that tax revenues will rise (or at least not fall).  In other words, if tax rates are cut, both GDP and tax revenues will increase.
  2. Lower tax rates will stimulate the economy (GDP) but tax revenues will drop.  In other words, if tax rates are cut, GDP will grow but not grow enough to keep tax revenues from falling.

The difference is important.  If #1 is true, then we don’t have to worry about increasing the federal deficit when deciding to cut taxes. This is what Bush and Republicans asserted to be true in 2001 and most recently in 2010 and 2011. Republicans in Congress have been arguing recently that tax cuts won’t make the deficit worse.  They only way that can be true is if #1 holds true. The essence of the data that David Cay Johnston presented and that I quoted was in response to this assertion.  The data clearly shows that tax rates were cut and tax revenues fell.  They not only fell, they fell by a lot. It’s important to note that tax revenues fell even in the “growth” years before the Great Recession hit in 2008.  One implication of this is that assertion #1 is clearly false. Tax rate cuts do not stimulate the economy enough to keep federal tax revenues from falling.  Therefore, anytime tax cuts are proposed we need to realize they will add to the deficit.  Tax rate cuts make the deficit worse.

Now let’s consider assertion #2 above.  This is really a simplified version of what we teach people in principles classes:  tax cuts can stimulate the economy. This is simple Keynesian theory.  I think, though, that economists have been too lax in how we state this idea  (myself included – I ask for forgiveness!). We don’t clarify what’s meant by “tax cuts” and we don’t clarify the mechanism.  Strictly speaking, the idea is that tax rate cuts will reduce the taxes collected by government (revenues) and if they result in increased spending (aggregate demand) then GDP will increase. The trick here is that tax rate cuts will allow households/corporations to have more after tax income (also called disposable income). Then, if they spend that additional after-tax income, then GDP will rise.  In this sense, when Keynesian economists say “tax cuts” they’re usually meaning lower tax revenues, which most probably happened by tax rate changes.

So, in looking at the data, should we reject or accept the idea of assertion #2.  In this case, I think it’s important to look at a broader historical record than just the Bush tax cuts of 2001. What we see is that tax cuts (as in lower tax revenues) tends to have a stimulating effect on GDP, but the strength of the effect varies widely.  In other words, reducing tax collections (lower tax revenues) will sometimes grow the economy strongly and at other times it will barely have any effect.  If virtually never has a negative effect, unless it was combined with spending cuts.  What makes the difference?  A lot depends on how the taxes are cut and for whom they are cut.  This is because the critical linkage is whether they tax cut will increase spending.  To the extent a tax cut is just saved by the household or used to pay down debt, it doesn’t help the economy grow. This is the difference between the Bush tax cuts and say an earlier, more successful tax cut like the Kennedy-Johnson tax cuts in 1964. The Bush tax cuts were overwhelmingly targeted at top-bracket, high-income individuals and corporations.  Instead of spending the tax savings, these people largely saved the money and converted it into financial assets. Spending in the economy did not get much of a boost.

So in conclusion, I think we can say assertion #2, what we teach in class, is viable, but we really need to be clearer:

tax cuts will stimulate the economy and grow GDP if we implement the tax cuts in a way that they get spend and not saved, but they will worsen the deficit.


 

What we have here is really three different variables that we need to keep track of