Testing Krugman’s Debt Reduction Strategy (And Finding It Fails)

Nike recently published a series of ads declaring “winning takes care of everything,” in reference to Tiger Woods’s recapture of the world #1 golfer ranking. The slogan went over with certain critics like an illegal ball drop.

Many economists insist that “economic growth takes care of everything,” and the related debate is no less contentious than the Nike ad kerfuffle. Listening to some pundits, you would think there’s one group that appreciates economic growth while everyone else wants to see the economy crumble. It seems to me, though, that growth is just like winning – there’s no such thing as an anti-winning camp, nor is there an anti-growth camp.

More fairly, much of the growth debate boils down to those who think mostly about long-run sustainable growth and those who advocate damn the torpedoes, full speed ahead growth. I’ll break off one piece of this and consider:  How much of everything does growth take care of?

A typical dismissal of the dangers of excessive debt

In particular, I’ll discuss the idea that we shouldn’t worry about America’s soaring national debt because we can always grow our way out of it. Here’s Paul Krugman’s version of this claim, excerpted from his 2012 bestseller, End This Depression Now! (page 142):

We won’t ever have to pay off the debt; all we’ll have to do is pay enough of the interest on the debt so that the debt grows significantly more slowly than the economy.

One way to do this would be to pay enough interest so that the real value of the debt – its value adjusted for inflation – stays constant; this would mean that the ratio of debt to GDP would fall steadily as the economy grows.

Krugman goes on to suggest that:

  1. Using a reasonable estimate for real interest rates (2.5%), we’ll have to pay $125 billion of interest each year on the $5 trillion in additional debt that was added after the global financial crisis.  (My update: This is now almost $8 trillion.)
  2. This amount of interest is “well under 1 percent of national income,” which means that “even shock-and-awe debt numbers aren’t nearly as big a deal as often claimed.”  (My update: By Krugman’s definition of well under, the same calculation now takes the interest payment to well over 1 percent of national income.)

No way, PK, growth won’t be enough to save the day

So Krugman basically says that $5 trillion in additional debt isn’t that “big a deal.”  I’ll take a stab at showing that it is, in fact, a big deal.

First, I’ll restate the excerpt above. By suggesting we can keep the real value of debt constant and then sit back and watch growth whittle away the debt-to-GDP ratio, Krugman is saying that our primary balance (the budget balance excluding interest costs) needs to equal our accumulated debt times the real interest rate.  With our current debt of 105% of GDP and Krugman’s assumption of a 2.5% real interest rate, the primary balance needs to equal 2.6% of GDP.

Here’s the reality check, comparing a 2.6% hurdle to history:

krugman chart 1

The chart shouldn’t need much interpretation, but consider that the 1950s surplus was achieved by presidents who detested deficits and mostly ignored recommendations of the Keynesian economists of their time (as I discussed here). Consider also the post-1950s developments that make it much harder to control our finances now versus then (see here). All we could manage in the 1950s – with deficit hawks in the White House and wind at their backs – was a primary surplus of 0.9% of GDP.

In sum, reading Krugman’s book you would think that lifting the primary balance to 2.6% is easy-peezy-lemon-squeezy. But the evidence shows that just getting back to 0.9% would be Herculean, let alone raising the balance well above historic precedents.

But… but… CBO figures tell us we can spend like there’s no tomorrow

Krugman often points to CBO projections as another reason to be unconcerned about debt. Let’s look at these projections in relation to the 2.6% hurdle, zeroing in on the year 2023. I’ll use three different scenarios:

  1. The “baseline scenario” that the CBO is mandated to produce, even though it includes many “current law” assumptions widely known to be unrealistic.
  2. The CBO’s more reasonable “alternative scenario.”
  3. The alternative scenario adjusted for the effects of recessions, which are ignored in the CBO’s projections. (The CBO assumes there’s no such thing as the business cycle, as I discussed here. I’ll explain my adjustments for this nonsensical assumption in a later post.)

Here’s the chart:

krugman chart 2

I have three conclusions.  First, if you expect to close the gaps between the line and the projections at Krugman’s real interest rate of 2.5%, then I’d like some of whatever you’re smoking. The only way to make the challenge manageable is with real interest rates significantly lower than 2.5% and even negative for long stretches of time. With low real rates, the primary balance hurdle would be less than the 2.6% shown in the chart. In other words, the anemic investment yields of the last four years will be with us for a long time, as will the distortions and risks that come with them. (This conclusion shouldn’t be surprising to those familiar with past high debt episodes – see this paper, for example.) In economist-speak, these artificially low real interest rates are a form of financial repression.

Second, even with financial repression, projections show it’s incredibly optimistic to think the real value of debt can be stabilized through a full business cycle, which is a key assumption in Krugman’s scenario. More realistically, we should feel relieved if debt grows slower than the economy even if its real value is increasing.

Third, the situation gets much worse after 2023. This isn’t shown in the chart, but just wait for the CBO’s soon-to-be-published long-term projections and the recessions-really-do-exist version of those projections that I’ll post on CYNICONOMICS.

It’s also worth remembering that the 2.6% required primary balance is based on the current debt-to-GDP ratio of 105%. If policymakers were to heed Krugman’s frequent recommendations to increase deficit spending, the debt-to-GDP ratio would continue to grow and the primary balance hurdle would grow along with it.  The hurdle would rise even further above both historic precedents and future projections. This makes it even more important to see through irresponsible claims such as those in End This Depression Now! Both history and common sense tell us that “shock-and-awe debt numbers” are a big deal; an inability to control those numbers can be disastrous for long-run growth (as I discussed here); and economic growth doesn’t “take care of everything.”

Recommended link (and technical note)

Just yesterday, James Hamilton published an excellent article that also includes a history of our primary balances. He calculates the surplus required to stabilize the net debt-to-GDP ratio at different combinations of interest rates and GDP growth. This is different than Krugman’s scenario of falling debt-to-GDP ratios for gross debt. Now, Krugman’s reported debt for the end of 2010 is actually in between the net and gross figures published currently by the IMF (the source used in his accompanying chart), but other figures cited in the same section of his book are for gross debt and one would assume he’s not comparing apples to oranges.  I discussed the difference between net and gross debt here.

Bookmark and Share
This entry was posted in Uncategorized and tagged , , , , , , . Bookmark the permalink.

2 Responses to Testing Krugman’s Debt Reduction Strategy (And Finding It Fails)

  1. mofo. says:

    How much of the GDP growth predictions depend on a growing vs shrinking population? Wasnt fear of a declining population one of Krugman’s hedges in his recommendation that Japan undertake extreme Keynesianism? Is there any reason to believe that the US will not see its population shrink and its GDP along with it?

    • ffwiley says:

      Good questions. I don’t have all the answers but I’d bet that actuaries do a pretty good job estimating population growth and these estimates are built into the government’s various projections (Social Security, Medicare, OMB, CBO, etc.). For example, in its last long-term budget projections (last June), the CBO projected population growth to slow from over 1% to about 0.5% in twenty years time. Its GDP growth projections are then ratcheted down by that difference. IMO, they’re on top of things when it comes to demographics but not so much when it comes to the effects of recessions (which they ignore entirely per my article).

Comments are closed.