I've started reading Professor DeLong's "The Modern Revival of the “Treasury
View”,January 18, 2009 draft. He certainly does write well.
[T]he silliest and stupidest arguments made against
Keynes's policy proposals were made by the bureaucrats of H.M.
Treasury, with their so-called "Treasury View"1 of Britain's economic
problems: that each extra pound sterling of British government spending
had to be financed by borrowing an extra pound from Britain's savers,
which meant a pound less for Britain's firms to invest. Hence investment
plus government spending was constant. So fiscal policy could never
boost employment or production no matter what.
[I]t is as obvious a fallacy as you ever find in economics. If no
government bureaucrat can boost employment and production even in the
shortest run by deciding to borrow and spend more—as the "Treasury
View" maintains—than an immediate corollary is that no private
entrepreneur can boost employment and production by deciding to borrow
and invest more in his firm's capital stock. If the "Treasury View" is
correct, then homebuilders' and financial intermediaries' decisions to build
more homes were not the cause of high employment in the mid-2000s. If
the "Treasury View" is correct, then venture capitalists' decisions to
finance internet startups and telecom companies' decisions to invest in
fiber optics were not the cause of high employment in the late 1990s.
Similarly, the huge unemployment of the 1930s was not due to any
unwillingness of businesses to invest produced by the panic of the stock
market crash and the waves of bank runs and failures in the early 1930s.
And the high employment and output in the 1920s was not driven by
private business enthusiasm for investing in the "new era" technologies of
radio, electricity, and internal combustion after World War I.
We can immediately recognize that Fama’s argument must be wrong.
First, it proves too much: not just that government spending cannot boost
employment and output, but also that private enthusiasm like the
enthusiasm for housing construction in the mid-2000s or high-tech
investment in the late-1990s cannot boost employment and output either.
Later in the post, Prof. DeLong mentions that if Fama is willing to use a classical full employment model, his conclusion might hold, but that Fama didn't in his original post. Let's try going through the story now, though.
Case 1a. Suppose that everybody in the economy is working and there is perfect information. When the entrepreneur borrows money from the bank and hires a new worker, he must hire the worker away from an existing firm. Thus, employment does not change. Output does rise, however, because the entrepreneur wouldn't be doing this unless he had a higher-return project than the existing firms and hence can bid away the worker with a higher wage. Or, what happens is that he bids away the capital by offering to pay a higher interest rate to the bank, which calls in its loan from some other firm, which therefore cannot afford to hire the worker any more.
The example uses labor, but what the entrepreneur hires away might be machines, real estate, or iron ore instead.
This story is one I use in teaching my students about opportunity cost. For Silicon Valley to grow, Detroit must shrink. It is Schumpeter's idea of Creative Destruction from The Theory of Economic Development. There is a Circular Flow of production in the stable economy, and The Entrepreneur breaks it by diverting resources to an innovation. Brahma can't create without Shiva destroying.
Case 1b. Now let there be full employment, but imperfect information. The entrepreneur and the public generally think that the new project is better, but it's actually worse. The bank knows this, but also the entrepreneur has enough capital in his firm to repay the loan even if the project goes sour.
The bank will make the loan. At first, the price of the entrepreneur's company will rise, as will the apparent wealth of the economy. (Will the price of the existing company fall when it loses the worker? I don't know.) Later, the failure of the project will be apparent, and it will be clear that the true wealth of the economy has fallen. The bank, however, will make a profit.
GDP's course is interesting. Suppose the entrepreneur uses the loan to hire workers to build houses. Those houses have high market prices, and GDP rises that year because it is measured using the price of those houses (or, perhaps, what has previously been the price of houses of that size-- this works either way). Then, it becomes apparent that nobody wants to buy those houses. They have little value. The entrepreneur (or whoever bought the houses at first, if they're not still in his inventory) gets a lot poorer. Notice, though, that GDP does not fall because of this. GDP is a flow value, and doesn't change when the value of stocks change. Also, we don't go back and change GDP figures just because the output turns out to be less valuable than we thought. Nonetheless, we shouldn't think that in that mistaken year the economy was doing wonderfully. It's as if the houses that had been built turned out to be magical castles that turn into mist when someone tries to live there.
Something like that is what happened in the Telecom Bubble and the Housing Bubble.
If the government did the borrowing for a stimulus package instead of the entrepreneur, then it too would have to take the worker from some existing job. Employment wouldn't change. Output would fall, though, because projects in a stimulus package are by definition those that the government doesn't think pass a cost-benefit test in normal times. (I'm distinguishing here between stimulus spending and normal spending.)
We have to do these first two cases of analysis of the Treasury View to get to the more relevant cases:
Case 2. Some of the workers the entrepreneur hires come from existing jobs, and some were not employed before. (This is the real Telecom and Housing Bubbles case, I expect.)
Case 3. None of the workers the entrepreneur hires come from existing jobs. This is the case to understand when we come to analyze the Obama stimulus package. And, of course, we need to figure out if it is a possible case.
I'll need to return to thinking about Cases 2 and 3 later. I should mention, though, that I have no firm opinion on them. I do oppose the Obama stimulus, but mainly because I think the government would botch it even if it's true that a Keynesian stimulus would be helpful now. I'm a microeconomist, so that's what I pay most attention to. Also, though I'm a fan of Schumpeter, don't think that I am an "Austrian School" economist. I'm not sure what that means, actually, but I associate it with a distaste for equilibrium analysis, mathematical modelling, and price theory. I am a firm believer in all those things, and proud to be part of the MIT-Chicago Synthesis which is standard among modern economists. (I'd put both DeLong and Fama in that category too, despite their disagreements about Keynesian stimulus. Their methodology isn't all that different, just their conclusions. Though maybe I should put Fama in the old Straight-Chicago School; I'm not sure.)
Labels: Economics, Keynes, research, stimulus