Recent Posts

Wednesday, June 14, 2017

Let's Talk About Debt, Baby

Gerard MacDonell wrote "The debt debate is relevant now" a couple of weeks ago. In it, he argues that debt sustainability in the United States is a relevant issue now, not an academic issue a couple of decades out. He realises that economists in the Modern Monetary Theory (MMT) school will disagree, and he explains why he disagrees with the MMT view. I am in the MMT camp, and I suspect that I do not violently disagree with Gerard's view on the current state of the cycle. I would side-step his concerns about "fiscal sustainability," and instead argue a slightly-modified version of his argument: fiscal policy is relevant now (and it always is). However, political economy matters. That is, I do not think we can discuss fiscal policy in the dry technocratic terms our elites prefer to use; we need to accept that fiscal policy is inherently political. "Debt sustainability" is best labelled "political sustainability of debt." Given the drift in the Debt Ceiling debate, "(political) sustainability" is an issue that may hit in a matter of months.

The Economic Theory of Fiscal Sustainability

I will return to "political sustainability" in later sections. In this section, I will attempt to address the topic in terms that are closer to what Gerard might use. (In doing so, I may not use standard MMT terminology. What follows are my views, which I believe resemble the MMT position.)
The key plank of Gerard's argument is based around the Congressional Budget Office's (CBO) long-term estimate for the debt. (He notes appropriate disclaimers about long-term projections.)

I have serious reservations about the CBO's methodology (and forecasting track record). However, I doubt that I could come up with a superior alternative amid all my other projects right now. Therefore, I will not completely dismiss the CBO forecast: it is telling us something. The projected ratio is growing exponentially, and that would presumably be considered "unsustainable" (which is admittedly a weasel word).

I am fairly confident that the actual debt ratio will not resemble the CBO projection. Instead, if their estimates about the fiscal settings are roughly correct, what would certainly happen is that nominal GDP growth would overshoot the CBO projections. This would either be greater real GDP growth (yay!) or higher inflation (boo!).

In other words, we are back to the inflation limit on deficits, as per Functional Finance (link to primer).  What Gerard is diagnosing as a debt problem I would diagnose as a potential inflation problem.

If policy making in the United States were half-way rational, we would ask ourselves a very simple question: does it make sense to tighten fiscal policy now in response to a future inflation problem (with an unknown horizon)? If you believe in the Fiscal Theory of the Price Level, you would agree with that sentiment. For the rest of us, the answer is: no.

Unfortunately, policy making in the United States is hardly rational (which I will return to after a slight detour).

Why Not to Extrapolate Present Trends Too Far Forward

From the perspective of MMT, policymakers do not control the level of debt: they only set spending parameters and tax rates. The level of debt is not under direct control. (If you want to sound like an economist, you can say that "the debt level is endogenous, not exogenous.")

The current rise in the debt-to-GDP ratio reflected a number of structural forces in the "non-Federal Government sector."
  • There has been a negative correlation between nominal GDP growth and the debt-to-GDP ratio. (This article discusses the relationship between bond yields and the debt-to-GDP ratio; bond yields tracked nominal GDP growth over that period.)  Whether or not that correlation will always holds (or was just the result of other common factors), nominal GDP growth appears to have little room to fall going forward.
  • Foreign central backs (mainly in Asia) have been snarfing up Treasury securities as part of their trade policies. Meanwhile, the policy framework pushed the private sector to amass large pension fund assets (to "fund" the transition of the Baby Boom to retirement). Financial assets being taken out of the circular flow of the economy result in a demand deficiency, which is counter-balanced by greater fiscal deficits as a result of the automatic stabilisers of the welfare state.
In other words, the Federal Government debt-to-GDP did not rise to its current level by budgetary decisions; it was the result of "private sector" (lumping mercantilist foreigners in the "private sector") responses to non-budgetary policy decisions. Those structural factors may have run their course.

Sigh, Back to Political Economy

As I argued in Understanding Government Finance, the default risk for floating-currency sovereigns (like the United States) is political (although sufficient incompetence could do the job). Either the country ceases to exist (revolution or losing a war), or lawmakers decide to repudiate the debt. There are no "unsustainable debt ratios" to point to.

What makes a debt load politically unsustainable? That depends. During the Depression, the Canadian Federal Government was perfectly content to let the people on the prairies starve, rather than risk the sanctity of the budget balance. A few weeks later, when Canada joined into World War II, the same government was happy to overpay for first class passage to England for the previously starving persons who enlisted in the army. (I owe that historical insight to Pierre Berton.)

According to the Freedom Caucus (a faction within the Republican Party), practically any level of Federal debt is unacceptable. Therefore, we may be able to gear up for yet another fight over the debt ceiling over the coming months. The Freedom Caucus believes that by not raising the debt ceiling, the Federal Government can avoid default; rather it would be operating under a hard balanced budget constraint. (The legal opinions I have seen disagree with the Freedom Caucus' view. From a practical standpoint, I see no way the Treasury can prioritise payments the way the Freedom Caucus wishes without exposing the President to the charge that he is ignoring Congress' legislation.)

In any event, looking to economic theory to find a solution to irreconcilable political differences is not going to work. I doubt that there is a magical level of debt which will satisfy both the Freedom Caucus as well as the Democrats. In other words, to the extent that "fiscal capacity" is a political concept, we are unable to come up with a quantitative definition for it.

(c) Brian Romanchuk 2017


  1. It's one of the arguments for scrapping this Gold Standard anachronism entirely.

    The US is a very weird nation. After all it still uses Fahrenheit. In nations that have a working parliament and that have little issue with the central government owning and/or controlling the central bank, then there is a question that needs to be answered

    Exactly what is the benefit of a marketable, transferable welfare payment than couldn't be better dealt with by an appropriate individual non-transferable National Savings product?

    Whether it is a pension annuity or Granny bonds for saving ISTM that it is better for government just to run an overdraft at the central bank that it owns/controls (thereby setting the rate of interest at zero) and handle saving welfare via National Savings.

    There is no function issue with debt sustainability, but surely it should be subject to political consideration like any other welfare payment.

  2. The CBO projection you provide shows the debt to GDP ratio in the area of 150% around the year 2050. While that would be historically high for the US, I am not sure that it would be even politically unsustainable. While it might happen that NGDP would rise beforehand to lower the ratio from the denominator side, given the inflation-phobic policy makers here, I wouldn't count on it. And look at Japan, where the ratio is well above 200%- that was achieved without significant NGDP growth or inflation. And their ratio has been sustainable both politically and economically I would think.

    Mr. MacDonell's problem is that he will not accept the implications of a purely fiat monetary system where the debt to income ratio of the issuer has no meaning other than the possibility that the spending involved might be inflationary in some circumstances. So he has gone off fishing somewhere rather than worrying about it, which actually sounds like a very rational thing to do. I hope he is having a good time. I am more than a little envious.

  3. Nice piece Brian

    Beyond the fact that CBO is always wrong and doesn't understand finance of a sovereign govt (they repeat the same "risks of rising govt debt" in all their other publications that I critiqued here, they make 2 fundamental errors (and thus so does Gerard).

    1. they assume in 2037-2047 that interest rates will be higher than GDP growth. That won't happen. CBO has finally accepted at least for the next 20 years what I've been saying for 10+ years, which is that interest rates above growth are not sustainable without a significant current account surplus. But after 20 years, they return to the same i>g predictions that they have been wrong about since 2000.

    2. CBO is required to assume SS and Medicare are fully funded. But this is inconsistent with current law, which requires the programs to only spend according to earmarked revenues and trustfunds. As such, under current law, there is no scenario in which CBO will be correct, because under current law the govt will have to cut both programs. So, it makes no sense at all to cut these programs or any others, because then you are building in cuts that are larger than the cuts already in current law. See this

    1. STF,

      Even though the CBO may be required to assume that Social Security and Medicare payments will be fully funded in the future and that is inconsistent with current law, it is not really a bad assumption considering the popularity of those programs and the outrage among those whose benefits would be cut. The old people always vote. And I will hopefully be one of them.

      Anyways, your New Economic Perspectives article blows giant holes in the CBO analysis of the 'risks' involved with rising government debt. It is far better than the article from The Nation that you link to.

      Question about your point #1- why are interest rates on government debt above growth (NGDP growth?) not sustainable for a currency issuer? Are you using 'sustainable' in the same context as Brian uses it here? If the answer is very long, could you point me to where to find it?

    2. I've wondered that too, because you can easily model a situation where interest and bonds happen without any GDP at all and it is all perfectly sustainable. The interest growth rate trends lower.

      I put together a simple agent model that shows just this.

    3. Thanks Neil. I had read your article previously but had not realized it dealt with my question to STF. Seems sustainable to me.

    4. Hi Jerry

      The problem with CBO modeling it that way is that it misleads the public into believing the programs are going bankrupt and the govt must cut spending or raise taxes now.

      What would be better would be to forecast with current law, showing that the govt debt is sustainable within the traditional (not-MMT) definitions. It could then show what that means for the programs under current law--there will be cuts but the programs will continue at about 75% to 80% of current levels. This would be a far more informative approach and be better for public dialogue, rather than the fear mongering that results from the current approach.

    5. Thanks STF. I agree with all of that. Certainly it would be better and more consistent (honest?) of CBO to base their forecasts with current law. The fear mongering would find a way to continue though. It is politically motivated not reality based.

      Another way that might make sense even if they continue their current methods (of disregarding the statutes involved with these programs)-- Say SS and Medicare amount to 20% of GDP today and was projected to increase to 25% of a larger projected GDP in fifteen years. They could highlight that and it would allow people to better assess the costs in real terms. This would be consistent with MMT. Maybe they do that already but it certainly doesn't make it to the news.

    6. Perry Mehrling, on nominal page 367 (article is 4 pages long so page 3 = nominal 367), argues that social security is capitalized (funded), beyond the level of the so-called Trust Funds, by the undertakings of the state to develop human capital and other assets in a social mutual fund that is the government:

      The State as Financial Intermediary:

      In this paper on SSRN I argue that the Fed/Treasury battery provides superior financial security systems (a type of cash flow insurance) to the bank and nonbank sectors of the economy:

      Financial Instrument Generation in the US Financial System:

      In this paper I argue that nothing would be gained by attempting to "fund" social security because the federal government provides financial security system insurance to all the "funded" investments whenever necessary to perform a systemic bailout:

      Treasury Finance in the US Financial System:

      The problem with investments is that no actuarial model can accurately predict cash flows on a long run basis to due financial system complexity -- so the federal government must allocate a financial loss and cash flow insurance via some political mechanism. Most members of Congress are probably not even aware that this is the game being played regarding student loan policy and other policy debates over the so-called public debt (insured national savings), taxes, and federal credit programs.

  4. The US, along with most countries now, I think, has a monetary system where money is based upon debt. Debt and money are two sides of the same coin, so to speak. But who tells the populace that fact? The unsustainability of debt means the unsustainability of money. Who thinks that money is unsustainable? Even bitcoin is still around.

    Merle Haggard wrote, "I wish a buck was still silver." It was true that you could once redeem a dollar bill, a silver certificate, for silver. But all that meant was that you would get a bunch of quarters and other change in return. The same thing would happen today.

    There are arguments, OC, about different types of money but any question of the sustainability of government debt without questioning seignorage is lame.

  5. Another pet peeve of mine is the debt/GDP ratio. It has its uses, but it has a bad psychological effect, namely that people talk about the debt/GDP ratio as though they were talking about the debt alone. A rising debt/GDP ratio might mean that the debt is staying the same but the GDP is going down. (That might not surprise me, if we let secular stagnation become the new normal. But if so, reducing the money supply is not the answer.)

    Also, if providing for Social Security and Medicare is a problem, it is a problem for the society as a whole, not just for those programs. In an era of rising productivity and robotization, who thinks that we will be able to provide for our elderly in the next decades? If we don't do it, it is because we won't.

    But yeah, the national debt is a political sore point. Why not issue consols, which do not count as debt?

    1. Edit: I left out a "not". Who thinks that we will not be able to provide for our elderly?

    2. Consols would probably be counted as debt; if the lack of maturity date was an issue, they would just change the definition of "debt" in the statistics.

      Sure, the debt-to-GDP ratio is a useless statistic, but it is too useful politically. There's plenty of Democrats going after Republicans about potential tax cuts, and the Republicans are always happy to point to the debt levels when welfare programmes are discussed. Asking them to grow up is probably asking too much.

    3. According to Carlos Mucha (or my understanding of what he said--I may have misinterpreted), a perpetuity would not count as debt under current US law. I'm not enough of an expert to know if that's correct or not.

    4. I saw that (I had no idea who was the source). Consols might be useful as a debt ceiling work around, but the CBO would probably be forced to count the consols in their forecasts eventually (in my opinion).

  6. Even governments not up to speed with economics as they really are [MMT explains] do know that money is always there when needed. It just takes a war for them to act!

    1. Yup. Unfortunately most economists cant figure that out tho.

  7. I can't help but wonder what the economist in Venezuela are saying about the sustainability of debt. I really don't know the answer.

    Judging from the trend in Venezuela, the problem we should be discussing is not the sustainability of debt. Instead, we should be discussing where debt generated-at-ever-faster-rates takes an economy.

    Judging from what is happening in Venezuela, the results are good for some people but very bad for others. There seems to be a division between have's and have-nots that is increasing.

    Too bad that we have so little theory to explain what is happening there.


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.