The fiscally conservative case to borrow and spend -- and feel good about it

615 flag america building nyc.jpg


What if borrowing money made you so much richer over the long-term that it paid for itself? It's not crazy. Millions of families make such a decision every year when they take on debt to pay for school. Indeed, investing in yourself is a bet that often pays off. But can the same be true for an entire country?

Brad DeLong and Larry Summers say yes. In a provocative new paper, they argue that when the economy is depressed like today, government spending can be a free lunch. It can pay for itself.

It's a fairly simple story. With interest rates at zero, the normal rules do not apply. Government spending can put people back to work and prevent the long-term unemployed from becoming unemployable. This last point is critical. If people are out of work for too long, they lose skills, which makes employers less likely to hire them, which makes them lose even more skills, and so on, and so on. Even when the economy fully recovers, these workers will stay on the sidelines. It's not just these workers who suffer from being out of work. We all do. High unemployment is a symptom of a collapse in investment. If we don't make needed investments now, that will put a brake on growth down the line. Together, economists call these twin menaces hysteresis. And if it sets in, it reduces how much we can do and make in the future. Assuming that spending now can forestall hysteresis, then this spending might be self-financing. In other words, spending now might "cost" us less than not acting.

This doesn't mean that government spending is magic. Often, it's anything but. But this is a special case. DeLong and Summers identify three factors that determine whether fiscal stimulus will pay for itself: 1) how much hysteresis hurts future output, 2) the inflation-adjusted interest rate, and 3) the size of the fiscal multiplier. Let's consider these in turn.


Economists know a lot about a lot of things. Hysteresis is not one of them.

Indeed, it's not clear whether long-term unemployment and investment shortfalls really do damage potential growth over the really long-term. Maybe hysteresis "only" wounds us for the next 20 years, but not the next 40 years. Unfortunately, there's reason to fear that this is optimistic. A recent paper by Stephen Davis and Till von Wachter finds that workers who are laid off during recessions -- who presumably take longer to find a new job -- take worse hits to their lifetime earnings than do workers who are laid off during good times. Lasting unemployment has lasting consequences. That should terrify our policymakers.

The below chart from DeLong and Summers shows the unemployment rate versus the percentage of working-age people who are actually working. Any divergence between the two shows us how many people have given up on trying to find a job after being out of work for too long. The recent numbers paint a frightening picture.
While quantifying just how much this will hurt our long-term productive capacity is a matter of guesswork, DeLong and Summers show that it doesn't have to be much to justify doing something now -- provided that rock bottom interest rates super-charge fiscal stimulus.

DeLong and Summers argue that real rates -- that is, adjusted for inflation -- don't have to be that low to make more spending a good deal. They calculate that real rates of anywhere between three and seven percent make fiscal stimulus worthwhile. Inflation-adjusted rates are negative now. But low rates don't only make borrowing cheaper. They might also make government spending more effective.


Government spending usually doesn't increase growth. Or, as economists put it, "the fiscal multiplier is usually close to zero." The multiplier just refers to how much total spending a dollar of government spending generates. For instance, if the government spends $1 billion and GDP goes up by $1.5 billion, then the multiplier would be 1.5. In normal times, the multiplier is zero, because the Federal Reserve offsets any additional spending. The Fed has its inflation target, and if more government spending pushes up inflation, then the Fed neutralizes it by raising interest rates. But with short-term rates hugging zero and inflation falling below target, this calculus might change. The Fed might allow the multiplier to be greater than one. And that would certainly make more spending a very good deal.

There are two broad objections to the notion that the fiscal multiplier might be quite high right now. First, just because short-term interest rates are at zero doesn't mean the Fed is out of ammunition. The Fed can still buy long-term bonds -- aka quantitative easing -- or tell markets that it will keep short-term rates low for an extended period. These things matter. If fiscal stimulus precludes the Fed from doing more monetary stimulus, then the apparent multiplier will be misleading. Second, it's hard to find many historical examples of a high fiscal multiplier. Critics like to point out that even during World War II -- when interest rates were also negligible -- that the multiplier was no better than during normal times. So, after all of this, does this mean that government spending isn't worth it?

Not so fast. Just because the Fed can use unconventional policy doesn't mean that fiscal stimulus is a waste. Much of the Fed's current strategy involves making (quasi) promises to keep rates low for a long time -- till late 2014, to be exact. It's a very watered down version of what Paul Krugman called "credibly promising to be irresponsible". The problem, though, is credibility. Markets might not believe the Fed. Actually, they don't. And that means that spending wouldn't be canceled out nearly as much right now. As for past instances of a high multiplier, World War II actually does offer solid evidence. You just need to know when to look. While we were actively fighting in the war, the government imposed private sector rationing. So    it's hardly surprising that government spending didn't spur on private spending when the private sector was forbidden from spending. But here's an oft-forgotten fact: we started spending on the war long before we entered the war -- to help arm Great Britain. Those were our "arsenal of democracy" days. More importantly, there was no rationing from 1939-41. Over this period Robert Gordon and Robert Krenn found that the multiplier was as high as 1.8. That's really, really good.

The Cliff Notes version of all of this is that a fiscal multiplier greater than one is not a unicorn. It's more like a black swan. It exists. It's just rare. And this looks like one of those rare times. Taken together with our historically low rates, now seems like a great time to make some investments in ourselves. Putting the long-term unemployed back to work is an investment in their human capital. Refurbishing roads and bridges is an investment in the physical infrastructure we need to keep competing globally. Both make us better off in the long run, and could conceivably pay for themselves. Of course, none of the above means that the Fed can't or shouldn't try to do more. It's more of a practical appraisal about what the Fed will -- and won't -- do.


Usually comparing the government's budget to a family's budget is a bad idea. Governments can borrow for far longer and on far better terms. And, counterfeiters aside, families can't print money. But in this case it's a worthwhile comparison. A family struggling to make ends meet wouldn't be wise to save money by pulling their kids out of college if they can afford tuition. Similarly, governments running massive deficits during a depression wouldn't be wise to embrace austerity if markets will lend to them on favorable terms. In both cases, the long-term damage outweighs any short-term benefit. 

Which is to say: When people offer you free money, don't say no.

We want to hear what you think about this article. Submit a letter to the editor or write to