Most fiscal stimulus arguments suffer from three basic fallacies.

First, if money is not going to be printed, it has to come from somewhere. If the government borrows a dollar from you, that is a dollar that you do not spend, or that you do not lend to a company to spend on new investment. Every dollar of increased government spending must correspond to one less dollar of private spending. Jobs created by stimulus spending are offset by jobs lost from the decline in private spending. We can build roads instead of factories, but fiscal stimulus can’t help us to build more of both1 . This is just accounting, and does not need a complex argument about “crowding out.”

Thanks to Bruce Kesler. Second, investment is “spending” every bit as much as consumption. Fiscal stimulus advocates want money spent on consumption, not saved. They evaluate past stimulus programs by whether people who got stimulus money spent it on consumption goods rather save it. But the economy overall does not care if you buy a car, or if you lend money to a company that buys a forklift.

Third, people must ignore the fact that the government will raise future taxes to pay back the debt. If you know your taxes will go up in the future, the right thing to do with a stimulus check is to buy government bonds so you can pay those higher taxes. Now the net effect of fiscal stimulus is exactly zero, except to raise future tax distortions. The classic arguments for fiscal stimulus presume that the government can systematically fool people.

I agree with the assessment in this article, and would like to point out a thought I have in regard to Mr. Cochrane's first point. He states that "Every dollar of increased government spending must correspond to one less dollar of private spending." I think the problem is much worse. He assumes a 100% efficiency in government spending, i.e. each dollar taken from the private sector will wind up being spent in the public sector. Any entity will have administrative costs, government even more so. The money used up by administrative expenses can not be spent publicly. The net effect is that a dollar taken from the private sector will result in somewhat less than a dollar spent publicly. We have a situation where, in order to create three jobs in the public arena, four will have to be killed in the private sector.

John Cochrane has the basics nailed down, of course, but there is one point he overlooks. If government raises its absorption of resources by spending more, then it must tax people more, just as Cochrane says. What he does not mention is that people therefore feel less wealthy and typically respond by working more, which raises national output/income. Thus the spending really does stimulate output. HOWEVER, in so doing it reduces welfare. People already had the opportunity to work more and have higher incomes before the increase in spending. That they chose not to do so tells you that they valued their remaining leisure time more than they valued the extra income they could have gotten by giving up some of that leisure and working more.

When the government comes along and taxes people more, the resulting loss in income does induce those people to give up some leisure and produce more, but that is a change in behavior they would have not have chosen in the absence of coercion. Forcing people to work more by first impoverishing them lowers their welfare. It is just the same as if disease had wiped out some of their crops or a fire had burned down their house. Yes, they work more and thus raise national income, but they are not happy about it. So indeed the extra government spending does what its advocates say it will do, which is to induce more output.

Unfortunately, for *precisely* that reason it lowers welfare. What a great idea!

You may be referencing specific experimental or empirical research here when you say what people will "typically" do, but it doesn't ring true to me. I don't accept that higher taxes act as an incentive to more labor contribution. Just the opposite. Surely you couldn't raise taxes to, say, 99% and expect people to work harder to make up for it. Atlas does shrug at some point. And I'm prepared to think that the shrug function, at the margin, is linear -- that is, there is not necessarily a tipping point below which people will work harder in response to higher taxes, and then suddenly work less hard in response to even higher taxes.

We are talking about different effects. We must distinguish between the average tax rate and the marginal tax rate and the very different effects they have. An increase in taxes has two effects - a substitution effect and a wealth effect. The substitution effect arises through the marginal tax rate (the amount by which tax paid rises if you earn an extra dollar of income). A higher marginal rate reduces the marginal return on economic activity and thus reduces the activity itself. That is what the supply side people have emphasized for decades, quite correctly. There also is a wealth effect, though, which is what I was talking about.

Let's consider a purely hypothetical tax for the purpose of illustration. Suppose the government imposes a pure lump-sum tax, say $50,000 that everyone pays no matter how much income he has. A person pays the tax, and then the next hour of his work is not taxed at all.

The marginal tax rate is zero, there are *no* distortions of relative prices, and the supply side type effects do not arise. Nonetheless, the tax affects behavior because of its negative wealth effect. The tax reduces after-tax income and thus makes people poorer. In response to that, people work more. What we have here is a tax with a positive average tax rate (tax paid divided by income earned) and a zero marginal tax rate. There is a wealth effect but no substitution effect. In such a case, the implication is unambiguous: people will work more to replace in part the income lost to taxes.

The tax you are talking about - a 99% constant tax rate - is a completely different type of tax. It has a marginal rate of 99% on all income. That will have exactly the incentive effects you describe, though it also has a wealth effect. With a proportional tax, if people work more, they also pay more tax. The incentive effects are quite different from the lump-sum tax of the preceding paragraph.

In the real world, there are no pure lump sum taxes. All real taxes have positive marginal tax rates. However, all taxes also have a wealth effect. The two effects work in opposite directions on labor supply.

The substitution effect reduces labor supply, and the wealth effect increases it. Which of the two dominates depends on the detailed structure of the tax (such as whether it is a purely proportional flat tax or instead is a progressive income tax) and the individual taxpayer's personal characteristics.

In my original comment, I was talking only about the wealth effect: to the extent that the tax reduces a person's wealth, it induces that person to work more. *If* there is going to be any stimulus in higher spending, it will be there *only* because of the reduced disposable income. Any such effect reduces welfare, as I explained in my original comment. If, in contrast, the substitution effect dominates, then the negative wealth effect is still there reducing welfare, but it is overwhelmed by the substitution effect and we don't even get a stimulus.

The particular mix of substitution and wealth effects in a change in the tax law depends on the structure of the change. For example, we could increase taxes by reducing exemptions ("broadening the tax base") or by raising the tax rate. The former has mostly a wealth effect, and the latter has a strong substitution effect. The stimulus effects differ correspondingly. Of course, we don't know how the government plans to raise the extra taxes it needs to pay for all this new expenditure, and I bet the government doesn't know, either. In the past, it sometimes has emphasized broadening the base, and at other times it has emphasized raising the tax rate. It is clear that, given the magnitude of the expenditure being proposed, the government is going to have use all means at its disposal to raise taxes. Right now, though, the government is avoiding any discussion of how it is going to pay for what it is proposing, so we can only guess what the net mix of substitution and wealth effects is going be.

You suggest that the "shrug" function is linear. It probably isn't because preference functions are non-linear. Indeed, there is some well-known work in economics estimating that the welfare loss associated with a proportional tax increases with the square of the tax rate.

Similarly, I would expect the disincentive effects of taxes to increase at an increasing rate as the tax rate is increased, i.e., the supply side disincentives should get more and more serious the higher the tax rate goes.

Posted by Donald L. Luskin at 11:01 PM | link  

Monday, January 26, 2009

Posted by Donald L. Luskin at 7:15 PM | link  

Posted by Donald L. Luskin at 7:02 PM | link  

Sunday, January 25, 2009

Posted by Donald L. Luskin at 2:39 PM | link