Why Fiscal Policy Tends Not to Work
Almost 2 weeks ago commenter Odograph posted this link. It talks about how much bang one gets for one’s stimulus spending.
This chart shows the bang for the buck that the government can expect to get from various stimulus proposals. For each dollar spent on food stamps, for example, real gross domestic product is likely to rise by $1.73. For each dollar spent on aid to state governments, G.D.P. is likely to rise by $1.36. (This is what economists refer to the as the multiplier effect.)
There is just one problem with this. The multiplier affect that first year economics students learn about has one important caveat: that the increase in spending be exogenous. That is the money is not already coming out from another part of the economy. Think of it this way:
- Take money from Paul.
- Give it to Peter.
- Peter spends it.
- Economic nirvana.
The fourth point hinges on a hidden assumption: Paul was not going to spend the money. For example, the article linked above states the following,
Why might this be true?
Well, people don’t necessarily spend 100 percent of the money they get back in the form of tax rebates. If they’re not in dire straits — which most of the people receiving stimulus checks last time around weren’t — they might save much of it or use it to pay down debt. Neither of these actions boosts the economy much in the short term.
If I pay down debts with my tax rebate I have spent 100% of the money, thus negating the opening premise. If I save the money then the money can be loaned back out to another who will then spend it.
Now in the current situation one could say that saving the money may not result it being lended back out. Fair enough, but where is this money going to come from anyways if we go with stimulus spending? Tax increases? Maybe. An inflation tax? Maybe. Borrowing? Maybe. All of the above? Maybe. But all of these activities take money that is currently in the economy and simply moves it around. It isn’t clear that borrowing $150 billion is going to end up having a bigger impact on the economy than leaving that $150 billion for other people to borrow and spend. People borrow money to spend it, not just sit on it. If people don’t spend it they are losing money. The only reason for having the government borrow it is if nobody else is going to borrow it.
Basically when a student learns about the multiplier effect the government has a mysterious bag of money, where it came from nobody knows. That money is then tossed out of a helicopter and then spent by people.
As Raj Chetty, an economist at the University of California at Berkeley, points out, many recently unemployed don’t have much cash on hand when they lose their jobs, but they are loath to fall behind on pre-existing commitments like mortgage or other debt payments. Doing so would damage their credit rating. Many of them instead cut down on their consumption of things like food and other basic necessities. Extending their jobless benefits or giving them food stamps allows them to keep spending on these items — spending that translates directly into short-term economic growth.
This maybe true to an extent but there is nothing to say that people will not compensate by reducing spending in other catagories such as utilities, household maintenance, etc. When one loses one’s source(s)of income, reducing expenditures and conserving cash is a reasonable response. Also, food is not exactly what you’d call a huge portion of personal consumption expenditures (PCE). In both 2006 and 2007 expenditures of food were about 13.7% of PCE. PCE is about 70% of GDP so food expenditures are about 9.5% of GDP and we are talking about subsidizing only a fraction of that via food stamps. The bang might be biggest in terms of bang per dollar, but it still if there is going to be a bang it will be a pretty small bang. And again, this money has to come from somewhere else in the economy. That means that somewhere else in the economy there is likely going to be a decrease in output.
The same applies to tax rebates and tax cuts as well. The only thing about a tax cut is that a tax cut also reduces the deadweight loss associated with taxes. However, this reduction in loss of efficiency is not enough to result in a large increase in economic activity. Further a one time tax rebate may not work because of the permanent income hypothesis. This hypothesis holds that if a person gets a one time increase in their income they will save the money and enjoy a permanent increase in interest income.
Basically, the idea behind the multiplier effect that underlies most of fiscal policy is that it only works if the money was not going to be spent in the first place. If Paul, using the example above, was going to spend the money anyways, then taking it and giving it to Peter really doesn’t help at all. In fact, if Paul reduces his future work effort because he knows his income will be taken then it could very well end up that the multiplier is less than 1 meaning a decrease in economic output. Similarly if we borrow the money, again Paul doesn’t have it to spend and so the increase in output is likely to be the same prior to borrowing. It only works if Paul wasn’t going to spend the money or lend it out.
Update: Over at Marginal Revolution Alex Tabarrok has a good line from one of his colleagues, Russ Roberts,
My colleague Russ Roberts says fiscal policy is like trying to raise the water level by dipping a bucket in the deep end of a pool and dumping it in the shallow end.
Pretty much sums it for me.