Bryan Caplan  

How I Was Wrong About Government Debt

PRINT
Constitutions Matter: The Case... My Path to Open Borders...
When an individual owes three times his annual income, we think it questionable: Okay for recent home-buyers, but probably a bad idea.  But when a government owes 300% of GDP in peacetime, we think it blatantly irresponsible.  I've often been puzzled by the disparity.  Why do so many freak out about U.S. government debt as soon as it exceeds 50% of GDP?

Only today, though, did I realize that these two private versus public debt figures are not comparable.  The right comparison is government debt relative to annual tax revenue.  By this correct measure, the U.S. is indeed in irresponsible territory.  Take 2011: With government debt at 67.7% of GDP, and government revenue at 15.4% of GDP, the U.S. debt/income ratio was already about 440%.

Today Jeff Sachs wrote:
Yet the two parties just ran the most populist campaigns this side of a banana republic, and I do a grave disservice to banana republics these days to say so.
My initial reaction was to dismiss it as hyperbole.  Now I'm actually worried.


Comments and Sharing





COMMENTS (38 to date)
Alex Godofsky writes:

Why can't this just be a supply-and-demand story? There is more demand for government debt (per dollar of revenue) than household debt, so in equilibrium we should expect to see more government debt at a higher price.

Nick Bradley writes:

I question the motives of the Austerians when real interest rates are negative out to 20 years.

Sovereigns -- states that own their own currency -- can't have debt crises and monetary collapse unless they're adding zeros to the currency on a monthly basis.

egd writes:

Is the U.S. Government really in irresponsible territory? As long as there is sufficient income to service the debt, it's not a problem, right?

If real annual tax receipts to the government increase by 3% every year, then the government can afford to increase its borrowing so that service on the debt increases by 3% every year without being irresponsible.

Only when the cost of debt outstrips the ability to handle the debt does it start to become irresponsible (unsustainable).

I haven't done the math, but I suspect we're well past that point.

Bruce Bartlett writes:

You are still not correct. The correct measure of debt sustainability is interest on the debt as a share of revenues. This is the measure the rating agencies use. They say that 20% is the limit.

Michael Huemer writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring your comment. We'd be happy to publish your comment. A valid email address is nevertheless required to post comments on EconLog and EconTalk.--Econlib Ed.]

8 writes:

Sovereigns have crises of existence. Most get into debt during wars or civil wars. Losers go into default/hyperinflation.

Now we have an interesting situation where governments are increasing the debt before the war. This just puts the cart before the horse, making debt a good reason to exit a financial or monetary union. The profit from secession goes up as government debt rises and becomes a self-fulfilling prophecy once the first state/region exits.

The U.S. government is already in deep trouble, based on current trends. Cash debt costs are low, but adding in accrued interest (mostly on the depleting SS trust fund) runs at over $400 billion. Interest rates are about 2.5% on total Treasury debt because so much is packed into the short term and the Fed is buying the long-term debt. So a realistic interest cost today, one that is even arguably optimistic, would be to assume USG will see it's debt service on a budget similar to today rise $600 billion, to $800 billion about 30% of revenue.

If GDP growth picks up, if employment picks up (not unemployment, but the civilian employment rate), then you start closing that gap. But if this were a stock I'd definitely dump it based on the behavior of management.

kingstu writes:

In Corporate Banking we often look at Total Debt/(EBITDA - Maintenance CapEx) as a measure of Leverage. Probably the best measure of Leverage for Government would be Total Debt/(Tax Revenue - Mandatory Spending). I realize Mandatory Spending can be reduced by an Act of Congress but it takes an Act of Congress to do so. If we use this measurement it would likely imply Leverage of about 10x which is typically where you see Leverage just prior to filing Chapter 11.

Nick Bradley writes:

"I question the motives of the Austerians when real interest rates are negative out to 20 years."

Prices are set by supply and demand. Period.

The number of floating treasuries with duration above 10 years has fallen substantially relative to total outstanding treasuries. Virtually all new debt is financed with 5 year or less duration treasuries.

In other words long duration are in drastically short supply, so it's doesn't mean anything that long-debt is expensive. It's simply an artifact of supply and demand.

Well, you might respond, the fact is that there are still some marginal buyers of long-debt at low rates. So at least they must have low interest rate expectations.

No, no, no, no. Every single fixed income trader knows that long rates have little to nothing to do with actual market expectations of rates. This is an academic myth that has no basis in reality and has shown to empirically be wrong.

If you look historically 10 year forward rates have little to no correlation with the actual rate on a 10 year horizon, across G7 nations.

Virtually all the demand in long-bonds (and it doesn't take much given the constricted supply) is driven by insurers and pensions that are legally mandate to match the duration of their liabilities. Many parties must buy long-bonds to fulfill their legal obligation regardless of whether the price is fair or not.

So combine a fixed price inelastic demand, with a shrinking supply. What do you get? Rising prices (or falling yields) at the back end of the curve.

But it has next to nothing to do with rates expectations. If you want to see that how many performance targeted bond funds are actually buying 20+ year paper? Virtually zero.

wd40 writes:

If you are correct about the dangers of US government debt, then why are interest rates on US bonds so low?

Tim writes:

I would say that this market has been largely distorted in the US since the beginning of QE.

The Fed has created an artificial demand for US government debt, and almost certainly driven Treasury prices into bubble territory.

MikeDC writes:

The correct measure of debt sustainability is interest on the debt as a share of revenues. This is the measure the rating agencies use. They say that 20% is the limit.

I prefer Caplan's measure since the interest on debt is a function of the Fed setting the market.

Saying the debt load is "sustainable" because the government is basically depressing the economy by ensuring 0% interest on government debt forever is technically correct, I guess, but it's also cutting off the nose to spite the face.

Julien Couvreur writes:

Government debt is problematic regardless of its level, as it creates an inter-generational burden. It also increases the level of political conflict, as people have to argue over how to pay for goodies that were already bought (do not try this with your friends).

Regarding the level of US debt, if you want to get really worried, consider unfunded liabilities. I've seen numbers from 80 to 220 trillions (present value). The higher estimate comes from Laurence Kotlikoff.

Mr. Econotarian writes:

Although I am no fan of our current long-term deficit situation, the situation is a little different than consumer income versus government tax revenue.

A typical consumer can't simply go get a job that pays more. However a rich kid can hit up his dad for a bigger allowance. Government more closely resembles the latter. The question is when our "rich daddy" will run out of money.

Nonetheless, I suspect that US tax revenues could increase with tax rate increases in the US to a large amount. It is unclear to me that Greece, for example, has anywhere near as much wealth to tap as the US does.

Thaomas writes:

Does Bryan really think the best way to judge fiscal policy or the size of the national debt is by analogy to a household budget?

egd writes:

wd40 writes:

If you are correct about the dangers of US government debt, then why are interest rates on US bonds so low?

Supply and demand. The US is issuing a lot of debt, but there is currently a lot of demand for the (real or perceived) stability of US bonds.

Karl writes:

The obvious answer is that the government who issues it's own currency is not analogous in any way to a family. Building a model of government debt based on the limitations of a family living in Ohio has no meaning at all.

After the experience of the last 5 years, you'd think that pointing to impending hyperinflation would require a better argument.

wd40 writes:

Julien Couvreur writes:
Government debt is problematic regardless of its level, as it creates an inter-generational burden.

This is a common argument, but beginning economic textbooks show it to be wrong. In the US almost all of government debt is either owned by the government itself or by US citizens (contrary to the news, China does not own a large proportion of our debt). So, if and when a future generation pays back borrowed money, the money will go to the same age cohort. US taxpayers will pay US bondholders of the same vintage.

Alexei Sadeski writes:

A VAT and some minor tweaks to entitlements will solve this "crisis."

Voting public will have to be dragged kicking and screaming down this route, of course, but the only question is when - not if - such policy becomes enacted.

Bryan Willman writes:

Arguing about soverign debt is arguing about accounting.

What really matters is what part of the productive capacity of the economy has government promised to various groups, and what can it actually deliver?

If the demographics get bad enough (not in the current US future but could be) then it wouldn't matter how much money was held for social security. If there are 100 retired people for every working age person, the retired people will find life to be quite difficult, regardless of good or bad promises made by government or others.

The real question is how bad will US demographics versus commitments get to be?

MichaelM writes:

This is a common argument, but beginning economic textbooks show it to be wrong. In the US almost all of government debt is either owned by the government itself or by US citizens (contrary to the news, China does not own a large proportion of our debt). So, if and when a future generation pays back borrowed money, the money will go to the same age cohort. US taxpayers will pay US bondholders of the same vintage.

As if that means the debt just never has to be paid.

What 'debt burden' means isn't that there is some contractionary drag on your brain-frying aggregate, it means that there is a burden on tax payers that must eventually be paid. 'We' do not owe it to 'ourselves', the Federal government owes it to its bond holders. The Fed will pay that debt by borrowing more and paying interest costs (you should be familiar with how this works). That interest is paid by the tax payer (hopefully). It then represents a transfer from tax payers to bond holders, who are not always the same people.

Aggregates are beautiful things. They're even fundamental, in a way.

They can also be blinders that hide reality behind a kind of new fangled mysticism.

Costard writes:

Karl:

The obvious answer is that the government who issues it's own currency is not analogous in any way to a family.

I would agree that governments and families are not analogous. But the ability to issue currency does not, in any qualitative way, change the nature of a default. Bankruptcies and hyperinflations are equally ruinous to credit.

Wd40:

In the US almost all of government debt is either owned by the government itself or by US citizens (contrary to the news, China does not own a large proportion of our debt). So, if and when a future generation pays back borrowed money, the money will go to the same age cohort.

A third of U.S. debt is held internationally (5.3T/15.9T). This is not an insignificant portion. Furthermore your claim that "money will go to the same cohort" is a possible outcome but by no means an inevitable or even probable one. Notwithstanding the rigor of beginning economics textbooks, debt cycles are complicated and even non-beginning economists struggle to make sense of them. But I can't believe your point was serious when you didn't bother to supply it with an argument -- or at least a bibliography.

Methinks writes:

wd40: If you are correct about the dangers of US government debt, then why are interest rates on US bonds so low?

There are two important things going on and one whose importance I'm unsure of. First, the rest of the world is in worse shape. Europe is in crisis and Japan's Debt/GDP is approaching 300%. We are the cleanest dirty shirt and capital still runs to the U.S. when Europe trembles. The easiest place to park money is in U.S. Treasurys.

Second, the U.S. Treasury market is, by a very wide margin, the largest, most liquid in the world. The effect of that liquidity should not be underestimated. It's very easy to move large sums in and out without affecting price and so the lower yield includes a liquidity premium.

Finally, the Fed is torturing the yield curve. I think it's pretty difficult to figure out how much of an effect it's actually having. I doubt it's much of an effect when compared to the market.

Mr. Econotarian: Nonetheless, I suspect that US tax revenues could increase with tax rate increases in the US to a large amount. It is unclear to me that Greece, for example, has anywhere near as much wealth to tap as the US does.

I've always found this argument difficult to swallow. Higher tax rates reduce the return on work and investment, reduce incentives to engage in either and encourage a shift of energy toward avoidance instead - particularly in the group targeted. Reducing taxable income is not that hard. Higher taxes rates have never resulted higher tax revenue as a percentage of GDP, so why would we expect it to be different now?

http://mercatus.org/publication/tax-rates-vs-tax-revenues

Methinks writes:

I don't know where the bright line is for sustainability of sovereign debt. The only thing I know for sure is that the answer is always relative to other opportunities available to U.S. creditors. I suspect that means the U.S. will get a pass for a much longer period of time before the inevitable crisis that clears the balance sheet and destroys the the final tattered shreds of the constitutional republic. I doubt the collapse is right around the corner, but that also means the market will allow the U.S. to get into much deeper water and sustain more damage when it does come.

I also know that European countries' government spending is not much higher than America's as a % of GDP, their tax rates are much higher for everyone, the standard of living is much lower....and they can't afford their welfare state either. So, I guess there is something to worry about.

egd writes:

Karl writes:

The obvious answer is that the government who issues it's own currency is not analogous in any way to a family. Building a model of government debt based on the limitations of a family living in Ohio has no meaning at all.

Perhaps you could explain the proper analogy, or why the analogy used is inapt?

wd40 writes:

If one believes that financial markets work well as I suspect that Bryan and many readers of this blog believe, then one is put into an especially difficult position when one argues that the market for Treasury bonds underestimates the true risks involved. Hence my earlier rhetorical question to Bryan: “If you are correct about the dangers of US government debt, then why are interest rates on US bonds so low?” Egd’s response “Supply and demand. The US is issuing a lot of debt, but there is currently a lot of demand for the (real or perceived) stability of US bonds,” essentially says that the market discounts the dangers posed by Bryan. Egd’s response does not provide an answer as to how Bryan could be right when professionals who are actively in the market are betting to the contrary. Tim argues that the Fed’s quantitative easing is pushing prices into the bubble territory. Because outsiders are still buying US bonds at these high prices, a bubble implies that the market is mispricing the risks. A believer in efficient markets might ask Tim why he knows more than the professionals in the market. Methinks has an apocalyptic view of the “inevitable crisis that clears the balance sheet and destroys the final tattered shreds of the constitutional republic” and suggests several ways to get around the conundrum that I have proposed. (1) “We are the cleanest dirty shirt and capital still runs to the U.S. when Europe trembles.” This just says why US bonds are more expensive than European bonds, not why they are not all paying a much higher interest rate. (2) “The lower yield [on US bonds] includes a liquidity premium. “ Again this just says why US bonds are more expensive than European bonds, not why they are not all paying a much higher interest rate. And (3) to paraphrase his remarks, the US crisis is inevitable but the date is uncertain. What evidence is there that financial markets more generally are incorporating this inevitability?

Julien Couvreur writes:

wd40, you oppose my claim that there is inter-generational burden due to debt.

Feel free to point me to a textbook. Here is a simplified analysis which highlights the problem I referred to.

Year 0: Group A loans money to government voluntarily. Group B receives benefits paid with said money.

Year 20: Group A receives money back from loan as expected. Some people from Group B are still alive and taxable, they contribute to paying back the debt. But Group B2, constituted of new tax payers, also contributes to paying back the debt.
Group B2 bears a burden which was not voluntary, for which they didn't directly receive a corresponding benefit (Group B received that money), and for which they couldn't possibly have even voted.

The problem lies with Group B2, which is the new generation of tax payers.
It does not matter what nationalities compose Group A.
It also does not matter whether Group A and Group B may overlap, as Group A and Group B2 most certainly don't.

Even if you abstract money out of those transactions, and look at how resources are distributed, government debt necessarily involves an involuntary sacrifice on part of group B2.

Keith K. writes:

@wd40

Actually Nick Rowe,Don Boudreaux, and Bob Murphy have shown that this IS a problem with government debt financing.

http://www.fee.org/the_freeman/detail/debt-cant-burden-future-generations#axzz2Gwci8JRW

Methinks writes:

wd40,

That was quite a tantrum. You seem to be very confused about things.

Risk is relative. Prices are relative. If you do not grasp this concept, any hope of understanding is out of your reach.

I do not know what you mean by "so low". Low relative to what? To 10 years ago? to Friday? To what you think they should be? What are "they all" and why should "they all" be paying a higher rate? In other words: what are you trying to ask?

Here's what I'm trying to tell you: If expectations for, say, Spain begin to turn negative, I may dump my Spanish equity holdings and buy Treasurys until I can figure out my next move. Treasurys are so liquid it's like moving into cash. Ceteris paribus, I prefer Treasurys for that reason alone. If enough people do that, the Spanish stock market will tank and the U.S. Treasury market will rally. Even if I purchased a long duration bond, I'm not making some long range prediction about the United States. I just have to be fairly certain that the United States Treasurys market will do better over my investment horizon than alternative assets available for me to purchase. That "investment horizon" could be overnight or even a few hours. I don't really care what will happen to the United States 30 years from now.

Markets are not inefficient because they don't correctly predict events that may occur in several decades.

wd40 writes:

Julien Couvreur I was responding to your statement that going into debt creates an intergenerational burden. That is, the younger newer generation is burdened by paying off the debt of the older previous generation. But if most of the US debt is owned by the US (and most, but by no means all of the debt is owned by US government agencies and US citizens), then most of the costs to the younger generation of taxpayers is at the same time a benefit to the younger generation of bondholders. Of course, taxpayers and bondholders need not be the same set of people, but as long as most of them are US citizens, most of the aggregate effect on the younger generation is awash. In contrast, your example talks about the burden on the future generation of US taxpayers qua tax payers. I have no disagreement with the statement that government borrowing today increases the burden on future generations of taxpayers (ignoring the benefits they might receive as bondholders). And just to forestall other possible side issues by other commentators, I am not arguing that government investments in unproductive projects (whether borrowed or paid for by present taxes) is not a problem.


Methinks writes:

Oh, not the "we owe it to ourselves" meme again. Your purchase of government bonds is a claim on my income and you don't see the problem with that? Well I do and you'll have to fight me for it.

Arnold Kling explains very clearly how government debt shreds the fabric of society:

http://www.american.com/archive/2012/november/lenders-and-spenders-confronting-the-political-reality-of-debt

Julien Couvreur writes:

wd40,
Group A and B are the older generation whereas Group B2 is by definition the newer generation of tax payers.
The concrete application is that people in their 20s (group B2) are getting taxed for spending by their elders (group B) who are retired or about to, in order to repay other retired people (group A). That is the inter-generational burden.

Even if by some heroic assumption you claimed Group A has transformed into Group A2 (newer generation of bondholders, who inherited bonds from their parents in Group A, who are mostly dead instead of retired), the problem remains with Group B2 (new generation of tax payers) where it does not overlap with Group A2.

Aggregating the problem coarsely (only considering A+B versus A2+B2) and making undue assumptions (A transformed in A2, or A2 are the same people as B2) as you imply indeed hides the problem. But the task of economists is to look through such obfuscation.

Julien Couvreur writes:

wd40,
In thinking more about the nature of your reply, I am starting to think that you are not objecting to the analysis I presented, but rather to the name attached to it ("inter-generational debt burden").

The reason this label is applicable is that people who made decisions were in Group A (voluntarily loan) and Group B (as voters on Year 0), whereas the burden is born by Group B2 (which are not voters on Year 0, but are tax payers on Year 20).

I hope this also explains why your assumption that Group A has transformed into a new generation of bondholders (Group A2) would not invalidate this label.

wd40 writes:

In principle, calculating the price of a risky asset is quite easy. For example, if you think that the probability of a complete default of a one year "bond" is 10% within the next year and you can invest in a Swiss bank account with no risk at 0% interest, then you will demand about 11% interest on the bond (if the interest and principal is paid at the end of the year). No one is expecting a complete default on US bonds, but the very low present rates suggest that the likelihood of default is very low, Or to be more precise, the ten year treasury note yields compared to the yields on other less risky assets with lower (possibly negative) yields is not consistent with a significantly higher probability of default on treasury notes.

Jim writes:

"Only today, though, did I realize. . ."

Really? You just realized this? So all this time you've been operating under the belief that the federal government owns the entire economy?

Methinks writes:

Very good, wd40. The market has assessed the risk of a U.S. Default to be much lower than the risk of a Greek or Italian or Spanish default. Our RELATIVE default risk is lower then theirs. And the dollar is the reserve currency. And the Treasury market is the most liquid market on earth. That all helps keep the interest rate low.

Perhaps you do not understand me when I talk about what I think is the inevitable crisis. I'm not saying that it is imminent. But, do the math. The U.S. is running a budget deficit of over $1 Trillion per year. If the tax rate for income above $250,000/year is 100% and we foolishly pretend that will have no incentive effect, that fantasy will raise enough money to run government for a few months. Entitlements are untouchable - not if you want to be re-elected. Do you suppose this is sustainable? At some point (a point I won't even attempt to predict), the U.S. will go through the same crisis that has plagued every other thoroughly rotted welfare state.

Mark Bahner writes:
A typical consumer can't simply go get a job that pays more. However a rich kid can hit up his dad for a bigger allowance. Government more closely resembles the latter.

But in this case, the kid's debt is over 100% of dad's annual income.

wd40 writes:

1 Mo 3 Mo 6 Mo 1 Yr 2 Yr 3 Yr 5 Yr 7 Yr 10 Yr 20 Yr 30 Yr
0.06 0.08 0.12 0.15 0.27 0.40 0.81 1.31 1.92 2.70 3.12

To end on a note of near agreement, I personally believe that that the present path of expenditures relative to taxes is unsustainable. And I see no evidence (besides the financial market evidence) that that the political parties will bring these two parts of the ledger closer to each other in the near future. But at the same time I do not see any evidence that the financial market is viewing things this way (and that is why I brought up the argument in the first place). My reading of the above treasury-bill data is that the likelihood of either a default or rapid inflation to wipe out most of the debt 3 to 5 years from now is unlikely. The 5-year interest rate is barely above the 1-month or 1-year interest rate. And much the same holds for longer periods of time (it is clear from the back and forth in the comments that others might come up with alternative explanations for the small differential between long and short rates).

Craig Howard writes:

I question the motives of the Austerians when real interest rates are negative out to 20 years.

I question the intelligence of anyone who thinks "real interest rates" are real.

Comments for this entry have been closed
Return to top