This article was originally published in The Atlantic as Megan McArdle's guest column

Many thanks to Megan for inviting me to guest blog. 

Most of my research is on how intelligence--IQ--matters more for nations that it does for individuals. I've also done work on how monetary policy does or doesn't influence the economy, and on how we really didn't need to give unlimited bailouts to the big banks. 

I'll talk about a few of these topics in coming days, but let's start off by rehashing the battle over whether the 2009 stimulus bill--ARRA, the American Reinvestment and Recovery Act--really worked. I'm not talking about the tax cuts--I'm talking about the spending: green jobs, new government buildings, health clinic staffers. With my coauthor Dan Rothschild (now at AEI) I wrote two papers last year on the stimulus for the Mercatus Center at George Mason that got me thinking quite a lot about this. 

The Keynesian theory of stimulus is elegant: When a recession needlessly throws people out of work, the government can hire them, and then those people take their paychecks and buy stuff made in the private sector. So it's a win-win: The government and the private economy both expand. No cruel tradeoff: the pie is bigger than before, there's a "multiplier effect." I imagine Keynes genuinely loved the thought of saving his beloved market economy. 

So, how often does it work out that way? Well, Valerie Ramey of UC San Diego (FD: one of my dissertation advisers) has looked at the major studies and written some of her own, and in a new article she concludes that in the U.S., "on balance government spending does not appear to stimulate private activity." Yes, it boosts government hiring--and lowers the unemployment rate--so on average you're getting a free lunch there. But Faberge shampoo-style stories you read about in freshman economics texts where "he spends money at her store, and she spends money at another store, and so on, and so on" just doesn't seem to show up in the real world. 

You might think, "Textbooks don't really claim that government spending sets off waves of private-sector spending---that's just a caricature." But I've got a textbook right in front of me--Schiller/Hill/Wall, in its 13^th edition so it must have moved some product--and they talk about a multiplier of 4: One dollar of government spending sets off three dollars of private spending. If that's how the real world works, let's double the Department of Defense.

Case/Fair/Oster, a popular text coauthored by deservedly famous economists, is more careful--but even once they throw in all of their caveats, they conclude "in reality the size of the multiplier is about 2." Every dollar of government spending sets off an extra dollar of private-sector spending.

The Congressional Budget Office is a bit more conservative than both of those textbooks: They offer high-low ranges for the multiplier, so I'll cavalierly report a midpoint: 1.5. They also note that after post-ARRA "discussions on this topic with its panel of economic advisers" (to be a fly on the wall at that meeting) they dropped their lower-bound multiplier estimate down to only 0.5: So they think there's a reasonable chance that a dollar of government spending shrinks the private sector by fifty cents.

How could that possibly be? How could extra government spending shrink the private sector?  Well, in the simplest Keynesian model, it can't. You have to add some bells and whistles--let's call that "reality"--to get a multiplier less than one. 

Here's one path to a tiny multiplier: Let's suppose the government is trying to find the right person for the job.  Now, what if that person already has one....

I'll talk about that, and about much else, later this week.  Glad to be aboard.