A commenter on my last post suggested that Greg Mankiw would have a problem with my statement that spending multipliers exceed tax multipliers. That surprised me because as was written in this textbook (p.397):
The size of the shift in aggregate demand resulting from a tax change is also affected by the multiplier and crowding-out effects. When the government cuts taxes and stimulates consumer spending, earnings and profits rise, which further stimulates consumer spending. This is the multiplier effect. (The multiplier effect for a tax change will be less than the multiplier effect of an equivalent change in government spending.)
It then goes on to contrast the multiplier effect with the crowding out effect that comes from potentially higher costs of borrowing from government borrowing (whether taxing or spending driven). You need to consider these effects together to work out the total stimulus. Greg refers to all this here.
[DDET Read more]
Greg does think that the type of tax cuts matter.
Suppose, for example, that tax cuts are not lump-sum but instead take the form of cuts in payroll taxes (as suggested by Bils and Klenow). This tax cut would reduce the cost of labor and, if labor and capital are complements, increase the demand for capital goods. Thus, the tax cut stimulates demand not only by increasing disposable income and consumption spending (the textbook Keynesian channel) but also by incentivizing more investment spending. A similar result might obtain if the tax cut included, say, an investment tax credit.
However, I do not believe he is suggesting that tax revenue would rise.
But this is what Julie Bishop is saying. Let t be the rate of taxation, T be tax revenue and Y be output. Bishop is saying this:
This path involves two assumptions. First, that we are on the wrong side of the Laffer curve. I don’t believe this and even if we must have been there prior to 2007 so what was the Coalition doing? Second, that the increase in tax revenue will increase output. This is contrary to macroeconomic theory.
What is the evidence of the impact of tax revenue increases on output? Christina and David Romer look at this in their paper. One thing they point out is that, in the US since the 1970s, taxes do not appear to have been used for discretionary fiscal policy and changes in tax rates tend to be for long-run policy reasons. And when they identify tax changes not motivated by long-run effects, these appear to have a smaller effect on output.
In my reading of the paper, they examine the taxation revenue share of GDP and changes in it on GDP itself. They find that a 1% increase in tax revenue share has a negative impact on GDP of between 1.3% and 3%. The higher multipliers appear to come through a path whereby increased tax revenues results in low business investment.
So let me put it very clearly. Even if Bishop is right and tax rate cuts increase tax revenue, then this is going to drag the economy further into recession. How much it does so depends upon how much of an increase in tax revenue there is (you could over the top of the Laffer curve) and upon the strength of the tax multiplier. Bishop things that the latter one is large. Whadda you say what?
Here is the thing. There is so much the Opposition could legitimately be arguing now to put forward a case that more of the stimulus should be in the form of tax cuts rather than spending increases. The obvious is that the spending might be misdirected and it is a good long-term thing to get taxes down so we should take the opportunity rather than trying to fine tune the economy.
They could also argue that tax cuts should be structured in a way that stimulates. Of course, given that business investment is not being hit, the obvious place for those cuts is towards the lower end of the income spectrum. But that wouldn’t be hitting Bishop’s political base, would it?