Tuesday, January 20, 2009

Tax Cuts vs Government Spending

There's been a lot of debate about the stimulus and its use of tax cuts as opposed to a pure government spending stimulus.

When you want to stimulate the economy, traditional economic canon says that tax cuts are less effective than government spending, especially with bad trends in unemployment or GDP. Most of the models predict that a certain percentage of a tax cut gets saved and not put into economic effect. Government spending all gets put into economic effect, so traditional economic models prefer government spending to tax cuts.

For the United States, however, I don't see how an increased savings rate is a bad thing. Americans have a negative savings rate (meaning, if someone in America makes $100, they'll typically spend it all and borrow a few dollars as well), and this credit crisis is one that has made it much more difficult to borrow. Therefore, it seems unlikely to me that a tax cut will result in a true savings boost. Given Americans' preferred past and present spending habits, you'll likely see one of two outcomes:

1) Americans will use a tax cut to spend the way they would have if credit were available but they didn't receive a tax cut. That means they'll spend all of their tax cut, plus or minus a little bit, depending on credit conditions and overall level of confidence.
2) Americans will save their tax cut, which doesn't mean putting it into a bank for most, it means paying off their debt. A major source of this crisis is bad bank balance sheets. A tax cut therefore lowers the default rate on loans, helping get Americans out of debt while reducing trouble at banks almost as effectively on a dollar for dollar basis as a government bailout of bad assets.

Additionally, there's evidence (put forth by Greg Mankiw in a recent NYT article) that a tax cut may be more effective than government spending without these factors. He claims:

1) Government spending isn't that effective in stimulating GDP.
2) Government spending REALLY isn't effective in utility-economic terms (even if it shows up in GDP) if you don't use it on smart things, which often happens with infrastructure spending.
3) Government spending sets a terrible precedent for future government spending, which is already dangerously high.
4) Via a paper by Christine Romer, tax cuts even with some savings are doubly as effective in stimulating the economy as government spending in normal conditions. I suspect that even increasing that savings rate bc of poor economic conditions may not wipe that effect out, especially given that we save so little anyway.

Given these factors, then, why are we using government spending at all?

This one's a bit simpler. The US has significantly underinvested in its infrastructure since the '70s. Many energy experts are amazed we've had as few transmission problems as we have, and expect major brownouts and blackouts in the next 10 years. Our roads are in trouble, our water infrastructure is in trouble, we don't have enough nuclear scientists to service the projected demand for nuclear power plants. If we're going to grow after this recession, we need to fix these issues.

The typical response against that (espoused by Eugene Fama) is that stimulus transfers employment to public projects from just-as-legitimate-and-employment-heavy private ones ("crowding out"). That is a legitimate concern, but in this particular case, it may be ok as long as the crowding out isn't complete.

Some of the stimulus will transfer engineers and construction workers to public projects, but as long as the government projects only partially replace private ones, not completely (which I get the sense is likely, from reading a few too many oil and infrastructure company 10-Ks), it also leads to training of more engineers and construction workers. In the medium and long run, that training will provide us with more bodies to help us fix our much-larger-than-this-stimulus infrastructure problem, and may help us get off of fossil fuels faster because of less constraints on construction.

Overall, we do need some stimulus in the form of government spending, not just for getting out of this recession, but also for making sure we have the necessary infrastructure and people to sustain growth after it. However, that government spending should have a cap at what can be applied usefully and efficiently. Any stimulus level beyond that should be in the form of highly temporary tax cuts.

No comments:

Post a Comment