Good. Fast. Expensive. Pick Two.

It is a commonplace among conservatives that liberals just want to spend government money, as much as possible on almost anything they can find.  This is, of course, not true.  Liberals want to spend money on projects that they think will be more valuable than the equivalent usage in the private sector.

This is presenting some problems now that the actual aim should, by the theory of Keynesian stimulus, to spend money as fast as possible on almost anything you can find.

It is very obvious, now that we have the stimulus plans, that the Democrats are using stimulus as an excuse to spend money on things they want to spend money on.  Their demand for things like alternative energy programs is inelastic; it's just that it happens, right now, to be convenient to bill them as stimulus.

The problem is, that contra the Republicans, Democrats do care that money spent on these important projects is spent well.  And spending a lot of money well takes time.  It's an inversion of the old engineering aphorism:  good.  fast.  expensive.  You can only have two of the three.

Because the bill is massive, because it's focused on pre-existing Democratic spending priorities, it is going to be slow.  Matt Yglesias defends it, saying:

With the whole thing done, it seems that two thirds of the funds will flow within 18 months of enacting the plan. Of course it's true that 100 percent would be better. And even truer that if we had passed a stimulus plan back last September rather than experiencing months of delay thanks to conservative intransigence this problem wouldn't be so severe.

But in the context of stimulus, eighteen months is a long damn time.  Eighteen months is, in fact, about how long it takes a stimulus to work through the system.  If for no other reason, that ought to be a little worrisome for progressives because that means the stimulus won't have even 2/3 of its full effect until after midterms.  It is simply not "even truer" that conservative intransigence is causing worse delays than the focus on spending the money on massive new projects.  It's not even as true. It's not true at all.  No matter how you assess the relative benefits of spending to tax cuts, tax cuts could be 95% out the door in April.  So could many other kinds of rapid government spending--preventing fare cuts on transit systems, sure, but also repainting all the faded yellow lines on highways, or repairing park benches, redecorating government offices, etc.

Why does speed matter so much?  Because the primary argument for fiscal stimulus right now is not that we need to alleviate the pain of a temporary economic contraction--that's what things like beefier unemployment insurance, food stamps, and housing assistance are for (the first and the last are very good ideas, by the way.)  The argument is that we're in danger of a liquidity trap--that we could end up at a permanently lower level of output, as described by Keynes and popularized by Paul Krugman in the story of the Capitol Hill Baby Sitting Collective.  (Though it's worth noting that the ultimate solution was to double the money supply...)

The basic idea is that if everyone is afraid to spend money because they might be laid off, and sits at home in the dark, all the people who made money selling the things they used to buy will get laid off--and so will they, because they're part of "everyone".  The government basically shocks us into a higher level of output by spending the money we're afraid to.

Though you wouldn't think it from the really quite shocking incivility emanating from the pro-stimulus side, the empirical evidence that this works in a large industrial economy like ours is basically nonexistant.  The problem is, we have very, very few examples to test on:  America during the Great Depression, and Japan in the 1990s.  And neither America nor Japan managed to stimulate their way out of their troubles.  You can argue--and many do--that this is because we, and they, didn't stimulate enough.  That may be true.  But unless you can forward test your theory, it's a just so story . . . as we just painfully found out about the "It was all the Fed's fault" narrative of the 1930s banking collapse.  There is no excuse for calling people who question your highly theoretical model fools and charlatans.

What we've got, since Japan really never did emerge from its lost decade, is basically one fact: America entered World War II in a depression, and emerged from World War II without one.  Hopefully, the relevant variable was the massive, massive amount of spending, rather than any of the other explanations one can plausibly build about the effect of Total War on depressions--like the slaughter of some of your excess labor force, or the substitution of more immediate fears of being killed for panic about the financial future.

But the amount of government borrowing during World War II was truly gargantuan--roughly half of GDP by 1943.  All the relief dribbled out over the course of the Great Depression at best kept the Depression from being worse--unemployment was still in the double digits in the late 1930s.  Moreover, much of the spending FDR did do was paid for by tax hikes, which cancel out the stimulative effect of the spending.  It's hard to tell how much to credit even the improvement we did see to government borrowing, because of course the acute portion of a financial crisis does come to an end eventually even without government action, and by the time FDR entered office the thing had already been running for a good three years.

So if we're going to do stimulus, judging from our not-very-good best example, what we want to do is pack a massive wallop as quickly as possible, to shock those "animal spirits" back into a more normal economic rhythm.  I am skeptical that this will work even if tried, for reasons I have outlined elsewhere.  But if we are going to try it, we should be focusing less on the Democratic wish list and more on figuring out where money can be most quickly and effectively spent.