Larry Summers stirred a lot of interest with a talk he gave last week at an International Monetary Fund conference. The former Treasury secretary asked an unsettling question: What if the U.S. needs financial bubbles to maintain full employment?

Not for the first time, you might think, life imitates the Onion. The blog Zero Hedge notes that the satirical website got here first. “Recession-Plagued Nation Demands New Bubble to Invest In,” it reported in 2008. It quoted the chief financial officer of a “bubble-based investment firm”: What the U.S. needed was “a concrete way to create more imaginary wealth in the very immediate future. We are in a crisis, and that crisis demands an unviable short-term solution.”

In saying this might actually be true, Summers was indulging his taste for provocation and carrying lines of thought to extremes. This tendency has attracted attention before -- as when Summers seemed to suggest that pollution should be exported to developing countries or that women’s brains aren’t wired to do science. It’s an appealing trait in an intellectual, uncommon in eminent officials.

Summers’s thinking on the here and now of economic policy isn’t the least bit radical, in fact. His ideas on the longer term were more arresting; so was hearing them from the man who was President Barack Obama’s first choice for chairman of the Federal Reserve. Would-be Fed chairmen aren’t supposed to ask unsettling questions about bubbles.

The Argument

Summers’s starting point was to note that, four years after the beginnings of recovery from the crash, the economy still has idle capacity and high unemployment. With interest rates already at zero, the Fed has to use unconventional monetary policy -- quantitative easing -- to support demand. This hasn’t been enough, and the result is painfully slow growth.

No quarrel so far. The problems of the “zero lower bound” on interest rates are well-known. The standard remedy, which Summers and most other mainstream economists support, is short-term fiscal easing. Unfortunately, fiscal policy is being tightened; households, meanwhile, are still trying to cut debt and restore their net worth after a once-in-a-century crash in housing prices. In other words, there’s no great mystery about the current shortfall of demand. Under these circumstances, a slow recovery is exactly what you’d expect.

But Summers then asks, what if this isn’t temporary? Maybe there’s a chronic shortfall of demand, beyond what’s due to the crash. Maybe we face an age of secular stagnation, in which the zero lower bound is normal. If so, short-term fiscal easing can’t be the answer. Permanent fiscal stimulus may be necessary to maintain demand. And if that can’t be done because politics makes it impossible, what’s left? Bubbles. Summers doesn’t advocate them, mind you, he just poses the question.

To deal with secular stagnation, could we need reckless lending and borrowing, facilitated by negligent regulation? You’ve heard of the paradox of thrift -- the idea that saving is an individual virtue but (under certain circumstances) a collective menace. Now we have the paradox of prudence. Those ninja loans might have been a dumb thing from the point of view of individual lenders and borrowers, but the economy as whole needed them to keep output growing and people employed. Perhaps the Onion had it right.

Before we award “America’s finest news source” a Nobel Prize, let’s reflect. The slow recovery since 2009, as I say, doesn’t support the idea of secular stagnation because you can so easily account for it in other ways. For evidence of a deeper-seated problem, Summers turns to the years before the crash.

Slow Recovery

During the recovery from the recession of 2001-2002, the U.S. had a house-price bubble (the proximate cause of the subsequent economic collapse). But during that expansion, Summers argues, the economy returned to full employment without encountering inflation in wages or consumer prices. Without the bubble, the economy would have recovered too slowly. The implication is that bubbles are sometimes a good thing.

It’s a puzzling argument. Suppose, as Summers asks us to imagine, that the U.S. does face a chronic shortfall of demand, that fiscal stimulus isn’t available and that you therefore need bubbles to get the economy back to full employment. What’s the point, if periodic crashes such as the one the economy has just endured are part of the outcome? Bubbles eventually burst: That’s what they do. Wouldn’t slow growth and less-than-full employment be better -- if that’s the alternative -- than crash-bubble-crash?

Moreover, it’s plausible that bubbles do more to destroy jobs and output when they burst than they do to support jobs and output as they inflate. The process isn’t likely to be symmetrical. So this isn’t just a matter of choosing, for any given level of employment over time, more volatility or less. Crash-bubble-crash probably means higher average unemployment than secular stagnation.

Bubbles have other dire consequences, too. They draw resources into financial services and pump up pay and profits in that sector, and the subsequent crashes don’t fully reverse those changes. So crash-bubble-crash feeds an ongoing misallocation of resources and worsens inequality. One could go on. Altogether, the bubble cure -- if we’re being asked to take it seriously as a cure -- is worse than the stagnation disease.

That’s supposing the economy even has the disease. Summers’s proof that it does is the claim that the economy didn’t show signs of overheating as the bubble inflated after 2003.

That claim is questionable. There was no great inflationary surge, it’s true, but underlying inflation was creeping up and the headline rate was running at 3 percent or more by 2005. (Include house prices in the measure of inflation, as some argue you should, and the numbers would have looked much worse.) Unemployment fell to less than 5 percent -- the rate thought to be consistent at the time with stable inflation. Above all, the U.S. was running a huge and growing external deficit, which carried the risk of currency devaluation and inflation in later years.

So there were signs of overheating. Action to gently deflate the housing bubble before it burst might have left the economy at or close to full employment with consumer-price inflation nearer to 2 percent than 4 percent; house prices wouldn’t have had so far to fall; and the subsequent crash might have looked more like an ordinary recession than the cataclysm that transpired.

Secular stagnation isn’t impossible. It’s a disturbing possibility, worth thinking about. But there’s no strong evidence it’s upon us, and even if it were, bubbles wouldn’t be the answer.

(Clive Crook is a Bloomberg View columnist.)

To contact the writer of this article: Clive Crook at ccrook5@bloomberg.net.

To contact the editor responsible for this article: James Gibney at jgibney5@bloomberg.net.