When stimulus is not so stimulating

Excellent podcast last month on EconTalk on the multiplier effect, the concept that underlies federal stimulus programs like President Obama's American Recovery and Reinvestment Act of 2009.

So what exactly is the multiplier effect? Put very simply, the multiplier effect says that when the government spends money (input) we ought to see a greater output in the economy. So, for example, economists working for President Obama in 2009 said that for every dollar the government spent in the stimulus program, $1.52 of economic activity would be created. This is a good thing: for every $1 we pay someone to do work, that dollar gets spent on groceries, gas, or entertainment, and that helps us get out of a recession.

If you pay attention to economics news, you're probably under the impression like I was that the multiplier effect is settled science, a hard and fast law of economics that everyone agrees on. What's more, it's a tenet of Keynsian economics, the school of thought that argues the government should spend boatloads of money during a recession.

But it turns out the multiplier effect is still a hotly debated concept in economics.

UCSD Economist Valerie Ramey was the guest on the podcast, and she summarized for listeners the state of the art in government spending studies.

What happens when you estimate the multiplier effect incorrectly

What she found was a dramatic (for a subject as dry as this anyway) difference of opinion. Some economists have said the multiplier effect of previous recessionary spending as high as 3.7, which would be a really big bang for our buck. Others have calculated it as low as .8, which is to say that for every $1 spent by the government, $0.20 is actually taken out of the economy, a concept called "crowding out." Other economists say the multiplier effect fluctuates; maybe it's greater during a recession and lower during an expansion.

Ramey says there's actually very little certainty about the multiplier effect at all. She says the general consensus is that it could be as low as .8 or as high as 1.6. So basically, if the government spends money during a recession, these economists want us to know it might work or might not work.

What great advice! We pay these people for this?

Well, like a lot of things in the social sciences (and economics is a social, not physical, science), there's just no way to test the counterfactual, to test what would have happened had we not done the stimulus program. And therein lies the uncertainty.

There's a lot more nuance in the podcast about this topic (including the effect of far-term taxes to pay for that stimulus spending), so you should give it a listen. But I would argue that even if the multiplier effect is a bit hazy, stimulus spending is still a good thing on balance. The negative multiplier data didn't seem that rock solid to me, but, more than that, American's don't hold positive cost/benefit ratios as a supreme value over all other values. Americans also value fairness, equality, justice, and opportunity too. There are intangible benefits that occur when the government spends money, benefits that can't be measured directly, benefits that many Americans actually like. It's not just schools, bridges, and roads, stimulus spending can include job training, small business loans, tax credits, and a whole bunch of other things that real people value and that benefit not just the economy, but society as well.

So even with these new studies, I don't think stimulus spending or Keynesian thought is going anywhere.