This is what I’m talking about in my last post:
We can continue to debate the size and nature of the stimulus, of course, but roughly $800bn seems right and the mix of spending and tax cuts currently proposed also makes sense.
Yeah, roughly $800,000,000,000, give or take a few zeros. Y’know, back of the envelope. Seems about right. Good enough for government work. Maybe throw in an extra $100,000,000,000 in case I forgot to carry the one.
Can someone point me to the relevant math that supports this conclusion? Here’s Robert Reich. I looked for a link to his analysis, but it must be broken.
But if $100 billion in spending raises GDP by $150 billion, and the marginal tax rate is 1/3, $50 billion of the spending comes back in additional revenue. So bang for the buck — increase in GDP per dollar of added debt — is 3, not 1.5. Since the main concern about stimulus is that it will add to government debt, it’s this bang for the buck measure, rather than the multiplier, that’s relevant. And 3 sounds a lot better than 1.5.
Mirabile Dictu! Government spending begets government revenue!! The obvious conclusion? Spend more and more and more!!! As I read his equations in the linked post, there’s no limit, as long as the spending isn’t being financed by higher taxes. Leaving aside (although we shouldn’t) behavioral effects, financing is the issue not really addressed, isn’t it? The spending has to be financed somehow, either through future taxes, borrowing or the printing press. It could be that this bang for the buck is well worth the expense, but how do we know? Does any of this tell us anything about the long-term consequences? I’m prepared to believe that there are benefits to smoothing out the pain, taking our (smaller) lumps later for the benefit of staving off disaster now. But I’m also prepared to believe that the cost may be worse than the benefit. How would we know? Shouldn’t we know before we engage in this kind of unprecedented experiment in government building?
The chart from Romer and Bernstein reproduced in one of Krugman’s posts above is instructive.
Krugman doesn’t endorse this, by the way–he wants even more spending and asserts that more spending would correspondingly reduce the pain, and more quickly. OK, let’s say this is correct, for the sake of argument. The cost of doing nothing, according to the chart, is the cost, whatever that is, of the space between the two lines. By 2014 we’re in the same place with or without a “recovery plan.” But is that right? As I noted, 2014 arrives, post stimulus plan, with either massive debt, higher taxes or inflation. Seems to me any of these would have a negative effect on employment. Wouldn’t a more realistic picture have the two lines crossing at some point (probably before 2014) with a higher unemployment rate with the stimulus than without it for some period of time in the future? Or maybe the cost wouldn’t show up in employment (I’m not really much of a Keynesian). Regardless, I just don’t see the cost side of this calculation. Meanwhile, I guess the claim is that stimulus now will create growth. Two things: First, building bridges and subsidizing failing industries will not produce growth (employment, perhaps; productive emplyment, not likely), and second, the relevant claim should not be that stimulus will produce growth, but that stimulus will produce more growth than non-stimulus. Has anyone demonstrated that this is true, let alone that it is true if one includes the cost of paying for the stimulus?
The prospect of pointy heads and politicians planning in excruciating detail the contours of our economy for the foreseeable future using back of the envelope calculations and unsupported assumptions without any realistic assessment of the costs and the costs of error has me . . . a tad bit concerned.
UPDATE: I was wrong, and I apologize. It’s not all about building bridges, and that was an unfair claim. Rather, it’s also about weatherizing modest-income homes. I stand corrected. (HT: Tim Kane).
UPDATE II: I notice that Greg Mankiw also wonders where the accounting is.