"So let’s talk about stimulus math, as I see it."

Paul Krugman thinks large about the stimulus in the NY Times:

"So what kinds of numbers are we talking about? GDP next year will be about $15 trillion, so 1% of GDP is $150 billion. The natural rate of unemployment is, say, 5% — maybe lower. Given Okun’s law, every excess point of unemployment above 5 means a 2% output gap.

Right now, we’re at 6.5% unemployment and a 3% output gap – but those numbers are heading higher fast. Goldman predicts 8.5% unemployment, meaning a 7% output gap. That sounds reasonable to me.

So we need a fiscal stimulus big enough to close a 7% output gap. Remember, if the stimulus is too big, it does much less harm than if it’s too small. What’s the multiplier? Better, we hope, than on the early-2008 package. But you’d be hard pressed to argue for an overall multiplier as high as 2.

When I put all this together, I conclude that the stimulus package should be at least 4% of GDP, or $600 billion."

Here's my comment:

If we’re going to end up spending this amount we might as well do it now.

However, are you including things like UI in your figures? What’s the mix?

I mean, surely on what and where you spend the money is as relevant as the total number? Or are you simply assuming that the money goes out and that’s it? Is this organic or mechanical?

— Posted by Don the libertarian Democrat

Here's Alan Blinder's number:

"Next up, after reforming the bailout plan, is the Economic Recovery Act of 2009. Given the likely severity of the economic slide, a large dose of fiscal stimulus — amounting to perhaps 2 percent of G.D.P., or roughly $280 billion — is needed either in the lame-duck Congressional session this month or soon after Inauguration Day. "