In his article on the 2009 stimulus package ("Think Again: Obama's New Deal," September/October 2012), Michael Grunwald argues that the stimulus increased economic growth, contrary to a common view among Americans.
First, he suggests that there is a consensus among economists that the stimulus worked. He fails, however, to mention economists like Robert Barro of Harvard University or John Cochrane of the University of Chicago, whose models show it did not. Instead, Grunwald focuses on economists who predicted that the stimulus would work before it was passed and simply used the same models over again to claim it worked after it passed, providing little evidence about what actually happened. Grunwald also cites a Washington Post blog post that found, according to Grunwald, that mine was the only one of seven "useful" studies to find that the stimulus failed. But he says nothing to challenge my empirical evidence, and the list he cites also omits economists such as Barro and Cochrane.
Second, Grunwald reviews the sequence of events: The stimulus was passed in February 2009, and quarterly data show economic growth turned positive in the second quarter of that year. Timing, of course, does not prove cause and effect, but if you want to argue on the basis of timing you need to look at the monthly data around the time the stimulus funds were first disbursed. Most monthly indicators -- retail sales, exports, new orders for capital goods -- show that the sharp declines ended in December 2008 or January 2009, indicating that the economy was stabilizing before the stimulus was passed.
Third, Grunwald suggests that those who find that the stimulus failed are motivated by political considerations, writing that "before Obama took office, just about everyone agreed" on the Keynesian stimulus idea. But though there has been a resurgence of Keynesian ideas in the past decade, for many years in the 1980s and 1990s there was a near consensus among economists that these short-run stimulus packages did not work -- a view that emerged after implementation of Keynesian ideas in the 1970s produced terrible economic results. And in the research of mine that Grunwald mentions, I found that the stimulus passed in 2008 was just as ineffective as the one passed in 2009.
While Grunwald states his case clearly and concisely, his argument that the stimulus worked has been made before. For three years now, the American people have not bought these arguments, and they have seen that year after year the recovery is slower than the administration has predicted. That's the reason the stimulus is a political as well as an economic failure.
JOHN B. TAYLOR
Professor of Economics, Stanford University
Michael Grunwald's masterfully argued case that the stimulus mattered suffers from the following inference problem: A patient is very sick. The doctor quickly looks him over and immediately prescribes X, which is very expensive. The patient gets better, but not fully, and may be impaired for years. Would the patient have gotten as well without X? Is there a medicine Y that is absolutely free that would have cured him? The doctor is being sued by the patient for dispensing medicine without a clear diagnosis. Who's right?
Here's my take. The argument for raising spending and cutting taxes during a recession is based on the Keynesian model, which says that the economy isn't working either because firms are setting their prices too high (so that there is too little demand for their goods and services) or because workers are setting their wages too high (so that firms aren't willing to hire many of them). These market failures can sometimes be corrected by having the government print money or spend money. I say "sometimes" because if firms and workers respond by setting their prices and wages even higher, neither policy will work.
The trouble is that this crude Keynesian diagnosis, which Grunwald likely doesn't realize underlies the theory behind the stimulus, doesn't fit the facts. John Maynard Keynes, were he alive and kicking, would be the first to point this out. (The Keynesian model was developed by apostles of Keynes, not Keynes himself, who was skeptical of it.)
When Lehman Brothers collapsed in 2008, it started a massive panic. It wasn't Lehman's failure per se that fueled the panic, but rather fear on the part of employer A that employers B, C, D, etc. had grown afraid and were laying off workers, and that because of those layoffs employer A would face fewer customers, forcing it to lay off employees too. The ensuing firing of 8.5 million American workers over the next 19 months did not reflect prices or wages suddenly being raised to excessively high values, as the Keynesian model would suggest. Instead, this is what economists call a coordination failure. Had Presidents George W. Bush and, later, Barack Obama properly diagnosed this coordination problem, they could have sat down with the largest 1,000 employers and then the next largest 1,000 and so on to persuade them not to panic but instead to hire 5 percent more workers. I'm not talking about tax or fiscal bribes. I'm talking about using the bully pulpit to induce patriotic acts that, lo and behold, would have left those employers who hired with more customers -- namely all the workers the other employers would be hiring.
If fear is indeed the only thing to fear, then having the president spend huge sums of our children's money -- as Bush did in 2008 and Obama did through the 2009 stimulus that Grunwald's article praises -- could well make employers even more scared. Their reaction could be, "Gee, times must really be terrible if Uncle Sam is spending all this money." In short, both presidents prescribed medicine without properly diagnosing the problem. That means the patient should win his lawsuit.
Professor of Economics
Michael Grunwald replies:
I hesitate to argue with two distinguished economists, but Laurence Kotlikoff's preferred response to the 2008 meltdown does seem, well, novel. He suggests that after Lehman Brothers collapsed, Presidents Bush or Obama could have averted an economic crisis through simple appeals to patriotism. Specifically, he says they should have "sat down with the largest 1,000 employers" -- foreign companies too? -- and urged them to expand their workforces, instead of slashing them, for the national good. Problem solved! Presumably, Lennar and Toll Brothers would have hired more homebuilders, even though no one wanted any homes built. Confidence restored! I would respectfully suggest that you can't stop a capitalist stampede by asking the herd to remain calm.
In fact, while the economy shrank at an 8.9 percent rate in the fourth quarter of 2008 and lost 800,000 jobs in January 2009, it had the best quarterly jobs improvement in 30 years after the stimulus passed in February. The Washington Post's review now encompasses 15 studies, and, with the exception of John Taylor's, every one that was statistically significant concluded that the stimulus helped. Some studies did depend on models, but quite a few were empirical.
There is intense disagreement today about government micromanagement of the business cycle, but until recently there was broad consensus that epic private-sector collapses require public-sector intervention. Every 2008 presidential candidate had a stimulus plan; Mitt Romney's was the largest. Even Taylor's critique of the stimulus suggests it failed because it didn't increase government purchases enough -- in other words, it wasn't big enough. That is hardly a smackdown of Keynesian stimulus.