Does the U.S. Economy Need Bubbles to Live?

Larry Summers thinks the economy has fallen into "secular stagnation." Only more regulation, more inflation, and more spending can save us from our boom-and-bust fate.

Larry Summers can't spot any catchup growth that doesn't come from a bubble (Reuters)

Larry Summers thinks the economy is broken, and has been for awhile.

It's not just the Great Recession and the not-so-great recovery. It's the 8 years before that too. As Summers points out, even low interest rates, deficit-financed tax cuts, and nonexistent lending standards weren't enough to overheat the economy then. Sure, housing prices boomed, but nothing else did. Inflation was low, and unemployment wasn't that low. In other words, demand wasn't excessive, but it wasn't sustainable either.

Summers worries that this isn't only a story about our economic past, but about our future, too. That we won't be able to create enough jobs without bubbles, now and forever. It's a new old idea called "secular stagnation." Economist Alvin Hansen first proposed it in the late 1930s when it looked like slow population growth might slow investment, and create a permaslump. But, thankfully, the baby boom, and 16 years of pent-up consumer demand proved him wrong.

Goodnight, Sweet Sustainable Growth?

This time, though, might be different. See, there's something called the "natural rate of interest," which is the inflation-adjusted one that gets the economy to its Goldilocks state of low inflation and low unemployment. If the Fed sets rates above it, inflation should fall and unemployment should rise (though, in practice, wages and prices don't tend to fall below zero).

As Summers points out, this natural rate has been negative for most of the last decade. And that's why we've alternated between bust and boom. The Fed's 2 percent inflation target means that even if it sets rates at zero, it might not be able to generate negative enough real rates to hit the natural one. That's the bust. But these kind of persistently low rates might eventually encourage risky behavior that, well, bubbles over. That's the boom. And our economic trap.

There are three ways out: more regulation, more inflation, and more spending. The first is what economists call "macroprudential policy"—which is just another way of saying that we should make it harder to borrow during a boom. Summers, though, thinks it's a "chimera" that this alone can give us the "growth benefits of easy credit ... without cost." He might be right. A recent paper by Kenneth Kuttner and Ilhyock Shim looked at 57 countries, and found that, aside from limits on debt-service-to-income ratios, these kinds of policies don't limit household credit growth all that much.

If we're lucky, all we need is a little more inflation. And maybe a little less of China manipulating its currency. A higher inflation target, say 4 percent, would let the Fed engineer more negative real rates if necessary, and recover faster from a slump.  It should also push up nominal rates, and make it less likely that we stuck to begin with. That'd be even more true if China stopped sucking out demand by manipulating down its currency. That last bit would be nice, but maybe isn't necessary. Consider that Australia hasn't had a recession in over 20 years—yes, really—despite running chronic trade deficits. How? Well, it's mining boom is a large part, but even more so is its smart central bank policy that targets 2 to 3 percent inflation averaged over the business cycle. In other words, it makes up for any inflation undershooting with overshooting (and vice versa).

A Free (Fiscal) Lunch?

But we can make the Fed's job easier with some smart spending. The case is overwhelming right now. As I've pointed out before, the Fed's forward guidance is really them telling Congress how much fiscal stimulus they'll allow. See, normally there's not a strong argument for stimulus spending, because the Fed might just negate it by raising rates. But this time the Fed has told us it won't raise rates before unemployment falls to 6.5 percent (and probably not until "well past" that time). That means the "multiplier"—how much total spending a dollar of government spending creates—should be relatively high.

The case is even more overwhelming when you consider the long-term costs of a slump. It's what economists call hysteresis. For one, the long-term unemployed will eventually become unemployable in the eyes of companies that discriminate against them. For another, putting off investments today means the economy won't be able to grow as much tomorrow. So a slump begets a slump. Or at least a slower-growing economy. But if spending today prevents GDP—and tax revenue—from disappearing tomorrow, it could maybe, just maybe, finance itself. That is, we could create future tax revenue that would cover today's interest payments. That's what Larry Summers and Brad DeLong argued two years ago, and, as you'll see, that's a pretty low hurdle right now.

Whether or not stimulus is self-financing depends on three factors: the 10-year real rate, the multiplier, and hysteresis. Of course, it's easier at lower rates, but it can still work at higher ones if the multiplier and hysteresis are high enough—and both are above zero. The chart below shows you the maximum real rate that would make it work for any combination of multipliers and hysteresis effects. Each color corresponds to a different multiplier, and each group corresponds to different hysteresis levels. So, for example, the group at the far right shows you the highest rates that make it work for each multiplier assuming hysteresis knocks off 0.2 percentage points of productivity a year. In the most extreme case, with a multiplier of 2 and hysteresis of 0.2, stimulus would finance itself even if real interest rates were 41 percent; they're currently 0.7 percent.

One final note on reading this chart. Anything above the black line shows when stimulus should pay its own interest. That line shows the average real rate on 10-year bonds between 1990 and 2005—so when the economy was "normal."

Now, plenty of economists have found evidence of multipliers between 1 and 2, but hysteresis is a trickier question. We don't know as much about it. A Bank of England study found that it was 0.2 percentage points for them during the Great Recession, though that could be an extreme case. Still, as long as there's any multiplier and any hysteresis, stimulus should finance some of itself right now. Not that this is an important test for whether we should do it or not. It's just a test of how mind-numbingly obvious it is. Even if it wouldn't finance itself, there's still every reason to do more. As Evan Soltas points out, there are plenty of projects we're going to need to do eventually, so we might as well do them now when interest rates are so low, and save money. Oh, and put people back to work too.

A Secular Boom

Our early-21st-century economy was built on sand. The sand of rising home prices, rising household debt, and ever-rising incomes for the top 1 percent. Then it all fell apart. Home prices fell, households started paying down debt, and the only thing that didn't change was the top 1 percent still got most of the growth—95 percent of it, to be exact. If we want to build an economy on stronger foundations, we'll need less debt and more investment. More rules to rein in leverage, and more inflation to eat away at debt. And more public investment if the private sector won't.

Or we could just continue pointless austerity, and cross our fingers.