Brad DeLong pleads perplexity:

I simply do not understand their arguments that government spending
cannot boost the economy. As far as I can tell, they are simply burying
their heads in the sand.

At the start of 1996, the US unemployment rate was 5.6%. Then
America’s businesses and investors discovered the Internet. Over the
next four years, annual US spending on information technology equipment
and software roared upward, from $281 billion to $446 billion, the US
unemployment rate dropped from 5.6% to 4%, and the economy grew at a
4.3% real annual rate…

Back at the start of 2004, America’s banks discovered that they
could borrow money cheaply from Asia… or so they thought. Over
the next two years, annual US spending on residential construction
roared upward, from $624 billion to $798 billion, the US unemployment
rate dropped from 5.7% to 4.6%, and the economy grew at a 3.1% real
annual rate.

…[Y]ou cannot argue that these groups did not
increase their spending, and that their increased spending did not pull
large numbers of Americans – roughly two million in each case – into
productive and valued employment.

The government’s money is as good as anybody else’s. If businesses’
enthusiasm for spending on high-tech gadgetry and new homeowners’
enthusiasm for spending on three-bedroom houses can boost employment
and production, then what argument can Harvey, Fama, Barro, Steil, and
company make that government spending will not? I simply do not see it.

If all we’re trying to do is raise nominal GDP, then I agree that “the government’s money is as good as anybody else’s.”  Unfortunately, nobody’s money is very good.  To see why, we need to ask a question inspired by Bastiat: If X had not spent money on Y, what would have happened instead?  We can pose this question to both the private sector and the government. 

Suppose X is the private sector, and Y is the Internet.  Brad seems to think that if the private sector hadn’t poured money into the Internet, that spending would simply never have existed.  This doesn’t make sense to me.  If the Internet boom never happened, I say that the private sector would have spent on something else, or lent it out to someone else to spend on something else.

OK, suppose X is the government, and Y is the bail-out.  Again, Brad seems to think that if the government doesn’t spend Y, no one will.  This doesn’t make sense to me either, and for the same reasons: The money has to come from somewhere. 

No doubt Brad is eager to meet Bastiat’s challenge head-on and say: “Wrong!  If X had not spent money on Y, total spending would have been less.”  But there are only two escape hatches that allow this response.* 

The first is the interest-elasticity of money demand – when interest rates go up, people will respond by reducing their cash holdings.  To me, this seems only slightly less fanciful than inter-temporal elasticity of labor supply.  I’ve never reduced my cash holdings when interest rates went up, and I know the model!

The other escape hatch is endogenous monetary policy – when the government borrows more, the Fed has to print more money to keep interest rates from rising.  To me, this seems entirely real.  But once you admit that fiscal policy increases nominal GDP by inducing more expansionary monetary policy, why not just cut out the middle man of fiscal policy and make monetary policy more expansionary?

Admittedly, “Harvey, Fama, Barro, Steil, and
company” may not share my view that the LM view is roughly vertical.  All I can do is help Brad understand my own reasoning.  Maybe I’m wrong; but am I really so hard to understand?

*Or maybe three, if you count a “beggar-thy-neighbor” devaluation to siphon demand from other countries.