MMT and the Sustainability of Sovereign Deficits and Debt

L. Randall Wray | January 25, 2013

[This is the fourth part of a series (1, 2, 3) on sovereign deficits and debt.  The series was started in response to Ed Dolan’s original post detailing agreements and possible disagreements with the MMT approach.]

To recap very quickly, we agreed that sovereign government cannot become “insolvent” and forced to involuntarily default on commitments in its own currency. We moved on to “math sustainability” and agreed that so long as the interest rate paid on sovereign debt is below the GDP growth rate, then government does not necessarily face explosive growth of deficits and debts. And we agreed that the overnight interest rate is a policy variable, so that the central bank could keep it below the growth rate if desired. And we agreed that Treasury could use a “debt management” strategy to ensure that its average rate paid would be “low”—near to the Fed’s target rate, and if the Fed was pursuing a low rate strategy then on likely growth rates usually used in these types of models then the Treasury’s rate paid could be kept below the growth rate.

(Of course in recession the growth rate can go below zero but the interest rate would remain at zero or above; however this argument about sustainability is about the long term, not about cyclical problems.)

Now that always leads to the question: but if the Fed did pursue such a low interest rate policy, we’d get inflation that would force the Fed to raise its target rate above the growth rate to fight the inflation. Here was my one sentence response from last week:

“Here’s the preview: if deficits increase inflation rates, then “g” (GDP growth rate) rises so that even if the Fed raises “r”, we can keep g>r.”

Ed Dolan responded in the comments section this way:

“I think the part that will be more interesting to me is coming in Part 4. It will need to explain two things:
1. Even if it is possible always to keep g>r (both nominal) as inflation accelerates, would we really want to do so? Is there always some steady rate of inflation that guarantees g>r, or does it take continuously accelerating inflation? Are there any conditions under which accelerating inflation itself can undermine real output growth?
2. Suppose inflation is initially triggered not by monetary policy, but by an exogenous shock to real output (say, a natural or man-made catastrophe), or by attempts by the government to increase spending even after the economy has reached full employment (as in the “mission to Pluto” example of your MMT text). How can we be sure that the monetary policy operations needed to hold interest rates at an arbitrarily low nominal level will not induce further inflation?”

So let us begin to answer these questions. Today I’ll tackle question #1, or at least part of that question.

Let me begin with a blog I wrote a few months ago (quoted in Scott Fullwiler’s series over at NEP that I’ve drawn on several times in my own series–so here I am quoting Scott’s quote of me, which seems a bit strange!):

“As Randy Wray explained a few months ago, the negative income effect of low interest rates has been significant in offsetting any fiscal stimulus:

But there’s a darker side. The low deposit rates and the high fees are wiping out savers. I’m not telling you anything you don’t know. You cannot even get half a percentage point on your savings at banks. Sure, your mortgage rate has also fallen, but the net effect has drained consumer’s income. Here’s a quote from a Credit Suisse report:

The side-effect of the Fed’s near-zero interest medicine – the collapse in personal interest income over the last few years. The decline in interest income actually dwarfs estimates of debt service savings. Exhibit 2 compares the evolution of household debt service costs and personal interest income. Both aggregates peaked around $1.4 trillion at roughly the same time – the middle of 2008. According to our analysis of Federal Reserve figures, total debt service – which includes mortgage and consumer servicing costs – is down $206bn from the peak. The contraction in interest income amounts to roughly $407bn from its peak, more than double the windfall from lower debt service.

Let’s put that in perspective. Remember the Obama fiscal stimulus? About $400 billion a year for two years—let’s say almost 3% of GDP. There’s been a big debate about whether it “worked”. Only the truly crazy believe it did not save us from an even worse recession than what we actually went through.

Well, QE is removing an amount of aggregate demand from the economy equal to half of the Obama stimulus. And that is not just for two years—it goes on and on and on, year after year after year, as long as the Fed pursues ZIRP.

So QE is supposed to stimulate the economy by taking 1.5% of GDP away every year?

Just as we learned in the case of Japan—which experimented with ZIRP over the past two decades—extremely low rates take more demand out of the economy than they put in. So the Fed has mistaken the brake for the gas pedal: QE slams on the brakes but the Fed thinks it is sending more gas to the economy. The only thing we can be thankful for is that the Fed is driving a rickshaw, not a Buick. The damage it can do is not lethal.

Don’t get me wrong, I’m not against ZIRP—I’d have ZIRP all the time—but we need to understand that it does not stimulate the economy.

And we just found out last week that QE completely removed again almost $90 billion of income from the private sector in 2012.”

This is something that is almost always left out of analysis. When interest is paid, it goes somewhere. There’s a nearly impregnable belief that if the Fed raises rates that must slow the economy because higher rates discourage borrowing and spending. However, we know that most spending by consumers and firms is not interest sensitive. So what if higher rates have little impact on borrowing and spending? More interest is paid, and more is received. What is the net impact? Well, economists think it probably depends on who borrows and who spends, and their relative propensities to spend. If we take from those who spend 100% of their income (heck, for a decade American consumers happily spent $1.06 for every dollar of income—no matter what the interest rate was!) and give to those who spend only 50% then raising rates might depress the economy.

But what if the biggest debtor is the government? That is, what about Japan? For the sake of argument let us say that Japanese households are net savers and have managed to pay off all their debt, while government has debt outstanding equal to 200% of GDP. Then raising rates could be highly stimulative as consumers holding that debt get more interest income and consume much of it, while government does not cut its spending even as rates rise. And it turns out that you don’t have to go all the way to Japanese levels to get such an effect.

My UMKC colleague Linwood Tauheed and I modeled this effect some years ago, using a systems dynamic model he developed. We plugged-in a range of estimates for the relevant parameters (interest elasticity of spending, propensities to consume, and public and private debt ratios) obtained from mainstream empirical work. We found that once the government’s debt ratio rises above 60% of GDP then for plausible parameter values you can actually stimulate growth by raising rates. To be sure, it is a model and like any model the results all depend on assumptions. But we played it as safe as we could by taking mainstream estimates. It certainly appears that as you increase the government’s debt ratio relative to the private debt ratio then there are cases in which that will stimulate the economy. This is why I said that if the Fed raises rates, growth might rise to keep g>r. In addition to the debt ratios we need to know who holds the debt and what their propensity to consume the nation’s output is. (Note that even if the debt is held by foreigners payment of interest to them could induce them to buy our exports. We did not explicitly model that. Note also that much of the government debt is held by financial institutions and figuring out what the effect on spending might be can get very sticky or hairy or maybe sticky and hairy.)

This is certainly not meant to be an “I Win, Q.E.D. (quod erat demonstrandum)” argument. But let us think of the scenario usually imagined.

Government is hell-bent on running up deficits. That causes growth to rise, toward full employment. Inflation starts to pick up. Now this actually makes it easier to maintain growth rates above interest rates: g>r. But the Fed responds (with horror!) by jacking up rates, trying to get r>g. That increases private sector’s interest income, causing spending to rise and growth to continue to increase.

Ed wonders if I can guarantee that g remains above r? No. But I envision two scenarios.

a)      Policy-makers impose constraints to fight inflation. I realize that our hyperinflationary Austrians presume that policy-makers always and everywhere love to induce inflation. I do not observe that in the real world. Even though economists cannot find detrimental economic effects from “moderate” inflation (up to annual rates way into the double digits—even as high as 40% per year), the population hates inflation. And therefor so do politicians.

b)      But let us say I’m wrong about that. In the next election in reaction to the austerity-loving President Obama, the population chooses the party that promises to run up deficits to stoke the fires of inflation. (I’m not sure which party waiting in the wings that might be, but bear with me.) Once elected, the politicians oblige.

Here’s the deal. Tax revenues depend on economic performance. Faster growth generates even faster growth of tax revenues. Back in early 2007, my Levy Institute colleague Dimitri Papadimitriou and I showed that by the mid 2000s federal tax revenues were growing at a pace well over double GDP growth. We warned:

Revenues are now increasing at a rate of almost 15 percent on a year-over-year basis, far outstripping growth of government spending (growing half as fast), nominal GDP growth (less than 7 percent), and real GDP growth (just over 3 percent). The current situation, with tax revenues growing five times faster than real GDP, is historically unusual, reminiscent of the period before the Reagan-era recession. However, the late 1970s and early 1980s was a period of high inflation, which drove tax revenue growth through “bracket creep.” The mounting real tax burden was somewhat moderated by the rapid growth of nominal income. Comparing tax revenue growth with the rate of growth of nominal GDP, one finds that the current period is even more unusual: there is no other extended period since the 1970s in which taxes have risen twice as fast as nominal GDP. Thus, the real tax burden is rising at a faster pace today precisely because nominal incomes are not growing very quickly. Finally, many of the previous periods that saw tax revenues increasing at more than 10 percent per year were followed closely by recession: 1972–74 (average = 12 percent); 1977–81 (average = 15 percent); 1999–2000 (average = 10 percent). The exceptions (1984–85, 1996–98) were during earlier stages of economic expansions, with average growth rates of tax revenues a bit lower than 10 percent. Source:

As an aside, we predicted that these fiscal headwinds, combined with rising oil prices and overburdened consumers would lead to a recession and financial crisis. Can anyone remind me what happened in 2007? And the Queen thought no one saw it coming.

The point here is that if we get the high growth, high inflation scenario, the deficit will fall. That is what it always does. President Clinton even got a budget surplus (that killed the economy). That reduces the necessity of keeping the interest rate paid below the growth rate to maintain math sustainability.

While I do not believe in “equilibrium seeking free markets” I do recognize that the economy has something like a self-regulating thermostat. It works something like the thermostat on my boiler that sends hot water through hundred year old cast iron radiators that heat my home. Set at an energy conserving level of 58 degrees, the actual house temperature drops to somewhere around 56, the heat kicks in and raises it to 60 then shuts off, but the hot water in the radiators continues to increase the house temperature to 64 or so (hotter near the source, colder near the windows) before it begins the inevitable fall back toward 56.

Deficit hysterians want to have it every which way all at once. Uncontrolled deficit spending causes debt ratio to rise without limit but without inducing economic growth. Even as inflation rises, for some reason nominal GDP growth doesn’t rise. Bond Vigilantes and/or the Fed push nominal rates above nominal growth rates (or you can do it in real terms—doesn’t matter because as we learned in third grade you can divide both sides by the inflation rate) and the debt ratio grows toward infinity.

Note the steps in that chain of “logic”:

a)      Bigger deficits don’t increase GDP growth

b)      Higher inflation doesn’t increase GDP growth, so cannot keep g>r, and cannot increase tax revenue

c)       Higher interest rates and thus interest income doesn’t increase spending or taxes


So forever we’ve got deficits (largely to pay interest—that never gets spent), growing debt ratios, high interest rates (again, that create income that won’t get spent), high inflation, and low tax revenues.

It’s quite a story. Requires a “radical suspension of disbelief” as the late Hyman Minsky would say.

Next time, let’s look at this “self-regulating” system some more. Preview: watch the Supersize Me movie and recall Herb Stein’s quip that unsustainable processes eventually stop.


Leave a Reply

You can use these XHTML tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>