Arnold Kling  

Resolving U.S. Indebtedness: Various Scenarios

Why oh Why Can't We Have a Bet... What Would Coase Say?...

"Winterspeak" wonders if I would bet on the U.S. defaulting. Here are the alternatives that I can think of, and the odds I would give to them:

1. Muddle through. No major change in policy, and no major change in economic growth, but somehow the ratio of debt to GDP remains stable. I give this a 10 percent chance, although it implies that I am miscalculating the path that we are on. I really don't see how it can happen.

2. Technology to the rescue. Some major technologies, probably either wet or dry nanotech, produce so much economic growth that the ratio of debt to GDP stays under control. I give this a 20 percent chance. Sometimes I think the chances are higher, maybe even 50 percent. It's a difficult estimate to make--today, I'm in a mood to say 20 percent.

3. Policy changes. Congress increases taxes (but does not enact a wealth tax) and/or takes steps to rein in Medicare and Social Security spending. I should point out that I have been writing about the race between Medicare spending and economic growth since 2003. I give this a 25 percent chance.

4. Inflate away the debt with moderate inflation (between 5 and 10 percent per year). I think this would be politically costly, and it might not be enough to really inflate away the debt (it depends on how quickly bond investors adjust expectations and raise the inflation premium in nominal interest rates). I gives this a 15 percent chance.

5. Wealth tax. The government takes, say, 5 percent of everyone's personal assets above $100,000. It does this on a one-time basis (or so it says). I give this a 25 percent chance.

6. Hyperinflation. This would certainly expunge the debt, but it would be political suicide. I interpret Winterspeak as taking this scenario seriously. I don't.

7. Default. The U.S. simply refuses to pay some or all of its debt. I interpret Winterspeak as saying that this would never happen. I am inclined to agree, although I would just say that it is highly unlikely.

I think that the combined chances of (6) and (7) are no more than 5 percent, with (7) even less likely than (6).

Comments and Sharing

CATEGORIES: Fiscal Policy

TRACKBACKS (1 to date)
TrackBack URL:
The author at Jim's Blog in a related article titled Prospects of hyperinflation writes:
    Incumbents that engage in hyperinflation usually gain political benefit in the short run, and the short run is all they care about. The Weimar government was not punished at the polls. ... [Tracked on August 24, 2009 6:27 AM]
COMMENTS (52 to date)
Harkins writes:

It's a sign of the times that someone would call 10% p.a. inflation moderate.

E. Barandiaran writes:

Let me give you my odds based on the experience of Argentina and several other countries that faced similar situations. IN THE NEXT TEN YEARS (I'm cheating because I read about the new projections they are going to announce on Tuesday), I expect:
1. 60% this means that the debt/GDP ratio will be over 100% at the end of 2019; I'd say between 120 and 150%;
2. 5% this means a ratio between 50 and 100%;
3. 10% this means a ratio between 75 and 100%;
4. 10% same as 3;
5. 5% a ratio around 50%;
6. 5% same as 5;
7. 5% same as 5.
As of today the debt is $7.4trillion and GDP $14.1trillion.

Kit writes:

Re: Default

I can imagine a contrived dispute with China over Taiwan: the President calls for the default on Chinese owned debt supported by other Western indebted nations and probably very popular.

jkb writes:

On the wealth tax you are forgetting the constitutional restrictions on federal property taxes - which requires that taxes so levied be raised on a state by state basis, so that the money is raised proportionately to the states share of the national population. If California has 12 per cent of the national population California has to raise 12 % of the levy - even if it has 15% of the wealth. This is why we have never had federal property taxes, only excise or income taxes. And yes, the estate tax is an excise tax.

James D. Miller writes:

Palin's 2012 Campaign Slogan:

The American people shouldn't have to pay off the Obama debt.

Richard S writes:

There's an active market in credit default swaps on U.S. Treasury debt (Euro denominated, of course). Its existence and growth means that people are willing to put real money on the possibility of #7. Arnold, if you think it's so unlikely, here's a way to put your money where your mouth is...

E. Barandiaran writes:

And Obama's slogan should be "Don't cry for me America"

winterspeak writes:

ARNOLD: You continue to assume that the US Government is somehow constrained in its ability to create dollars. It is not. We are not on a gold standard, or on anything gold-standard like (fixed fx, convertible currency, non-domestic currency obligations, etc.)

The US Govt can meet any and all dollar$ obligations it may have without any problems or constraints whatsoever.

I see your list of probably scenarios, and they are all variations of "how does the US get out of this pickle" -- the pickle being the much larger debt to GDP ratio.

Have you thought about why the debt to GDP ratio matters, given that the US can pay ANY $debt? At any time? With no problems whatsoever?

Japan has a massive debt to GDP ratio, and inflation is not a problem there.

I understand that you've been taking a "hydraulic" view of macro, I urge you to take a "balance sheet" view of Macro. If you take all private sector balance sheets and sum their "paid-in equity" entry, you will get a number that exactly equals another number.

Care to guess what that number is?

Wilmot writes:


You really think the federal reserve is going to monetize all the debt the US government has created?

Dr. T writes:

One option gets omitted from most deficit reduction lists: sale of federally owned properties and of oil and mineral rights. The federal government owns huge chunks of the western states and most of Alaska. Selling some of this land could reduce the debt by hundreds of billions of dollars. (I'm not advocating this, just mentioning it.)

PeterW writes:

Note that in the '90s it appeared that (2) was happening and indeed a small surplus materialized in the late '90s. However this was a one-shot deal, and the fundamentals of the economy reasserted themselves, first with a one-shot expense for the War on Terrorism, then on Obama's large and long-term commitment to higher entitlements. So if you invoke (2) you'll have to explain why government spending won't grow along with the economy.

Mark Bahner writes:

In 1999, Ray Kurzweil predicted that in 2019 a $1000 computer would have the capability of a human brain.

In 2009, it doesn't seem like that prediction can come to pass in only 10 years. But let's say it's more like 20 years.

That still means that sometime in the next 10-20 years, we can expect economic growth to accelerate beyond anything in history. After all, if a $1000 computer is equivalent to a human brain, that means $7 trillion effectively doubles the world population.

I still expect a 50/50 chance of U.S. per-capita economic growth averaging over 4.5 percent per year in the 2020-2030 period, and over 6 percent per year over the 2030-2040 period.

Jeff writes:


In a sense, you're hung up on the unit of measure and forgetting *what is being measured* --- a quantity independent of the unit of measurement that's used.

Of course the fiat currency allows the devalued payment of prior debt through inflation, but the fundamental problem is threefold:

First, that GDP is valued in those same units, making the ratio of debt to GDP *unitless* and a fundamental measurement of *how much of our productivity we are using to fund government programs* regardless of monetary inflation.

Second, since our government continues to spend more than it collects in taxes, and further continues to grow that overspend, we continue to have to borrow from foreign sources --- and there's no end in sight. (Particularly when you consider unfunded liabilities...)

Finally, the appetite on the part of other nations for our debt will decline, and the cost of borrowing increase, the larger we allow debt-to-GDP, unfunded liabilities domestically, and unpaid foreign debt to grow. That decrease in appetite will accelerate even further as we inflate the currency, hence diluting its relative value. And paying back a fixed-amount debt with a drastically reduced-value dollar is also likely to curb foreign appetite for our government's IOUs.

Eventually, all of the liabilities would eat too much of GDP / future GDP for our debt to remain viable. At that point --- and when all other measures are exhausted --- the United States government will have only two options: confiscation of assets and even, potentially, more obvious indenture of its citizens; or default. (Shutting down almost all programs and further spending is at this point a given...)

Only by reining in deficit spending *first* and / or some combination of the productivity enhancements Arnold and others have mentioned can we avoid this eventuality, fiat currency or not. There's only so much *domestic product* --- and only so much discounted *future domestic product* --- to fund our continued fiscal irresponsibility. At some point, the rest of the world is going to realize that they can find other "sinks" for their excess capital; and at that point, assuming we've continued on the same course, it'll be game over for our spending, for the dollar, and in all likelihood for our sovereignty.

Now, whether or not we get there is another story; there will be plenty of opportunities along the way for us to remedy the situation and *stop the overspend* and *stop the borrowing.* We'd have to be hell-bent for fiscal suicide in order for it to happen. But is it a possibility?

More than ever before.



VangelV writes:

For whatever it is worth I will comment on the various scenarios provided.

1. Let us begin with the muddle through option. I just can't see how the ratio of debt to GDP can remain stable because we do not have a stable situation at this point of time. I give this a less than 1% percent chance, and like you I really cannot see how it can happen.

2. Like you, I like the technology to the rescue scenario but cannot see it given the anti-capitalist stance in the US. If I were an investor I would clearly establish the facilties that will be used to generate income in other countries where capital is treated with more respect than in the US and where corporate taxes are lower. Why would you risk your own capital to have the Democratic or Republican party tell you how much of a return you can make? Why not set up a subsidiary off shore in a tax free jurisdiction and have the operations in an entirely different jurisdiction where other investors or friendlier governments share some of the risks? For the US, I give this option a less than 10% chance.

3. I can see Congress increases taxes but I can't see that increasing the total take for the government because people will adjust and the total take will remain at around the same level as it has been for quite some time. While steps to rein in Medicare and Social Security spending could help most politicians that try to argue for this before there is a crisis will not get elected. I give this a less than 10 percent chance.

4. Inflating the debt away makes a lot of sense. But I do not see how such an effort will be controlled to a point where you could have a range of 5 and 10 percent per year with any degree of certainty. Given the bond durations I don't see how this allows the government to save much on interest payments or to keep revenues growing by enough to remain solvent for long. I give this a less than 10 percent chance.

5. I can't see the wealth tax working because if the politicians tried to take money from Americans to pay off Asian creditors there would be a revolt at the grass root level that will make the current tea parties and protests quite mild in comparison. I give this a less than 5 percent chance, even if the laws allowed Congress the power to tax wealth.

6. Hyperinflation is a much more serious scenario than you think. If America's foreign lenders lost confidence in the US government all those dollars that are now outside of your borders will flood back in and be used to purchase assets while the currency still had purchasing power. As USD holders expect its purchasing power to decline, they will reduce their cash holdings and will try to acquire goods while they are still cheap. If enough USD holders become convinced that the price increases will keep accelerating, there will be a flight away from the currency and into real values. Mises argued that this would cause a "crackup boom." I give this a 35 percent chance of happening because Congress and the Fed would have no power to stop the process unless they were willing to destroy the economy and permit foreign creditors to buy up assets at fire sale prices from bankrupt Americans.

7. Default is a strong possibility (around 40%) because of my last point. Any options designed to prevent a massive devaluation of the currency would cause a major problem for the economy and would drive many Americans into bankruptcy. That would allow foreign creditors to step in and buy assets from Americans and that would be too unpopular politically.

It might be a lot easier to have the government declare that Americans have a few days to convert all of their old cash for new bills that look very differently and that the old bills would become worthless at some point in time. At the same time, US government can declare that its debts will be paid back in specially serialized bills of the old currency and that they will trade at a discount to the new currency, which will be backed by a certain amount of gold, silver or some other commodity.

If the new currency is backed by gold it will still be in demand by foreigners who have no hard currencies to choose from. That would allow Americans to trade with foreign producers of goods and services and to bid for food, oil, or base metals or other goods that are required for consumption or as inputs for domestic production. Given the very productive American manufacturing sector, the US will be able to remain a relevant manufacturing power that will be able to exchange real goods and services for the things that it needs from foreign producers.

Of course, the last option wipes out many of the people who made a decent living by pushing paper assets around but such an outcome was inevitable no matter which scenario was chosen. The power elite would make out fine because many have probably hedged their bets by holding precious metals, real estate, and other real assets.

The major losers, in addition to the creditors, would be American savers, people on fixed incomes, and workers who had counted on their SS and Medicare benefits to retain their purchasing power.

Fenn writes:

Is number 5 the worst scenario to you? Under what conditions would it not be?

Would it be the worst scenario for a median income American?

Devin Finbarr writes:

Jeff, Wilmot, and Arnold,

The "debt" does not need to be monetized. It already is monetized. The dollar is a fiat currency. It's just a piece of paper backed by the USG. A T-Bill is just a piece of paper backed by the USG. A T-Bill is simply a dollar with a not-valid date.

Since T-Bill's pay interest, and dollars do not, most people who do not need dollars for an immediate transaction tend to hold on to T-Bills instead (directly or via a bank account or money market fund). There is no reason that the debt cannot be rolled over forever.

Think of it another way:

Imagine the Fed printed dollars and bought all the T-bills in the world at market prices. Would this cause inflation? No. No one's balance sheet has changed a bit (since the t-bills were bought at market prices). Since no private actors financial position has changed, they cannot increase spending and bid up the price of goods. So you get no inflation.

Note that this is not the result under a gold standard. If the U.S. was under a gold standard, its credit rating would be F-. Whenever it issued new bonds, it's credit rating would fall further, and the mark-to-market value of the bonds would fall far below face value. If the government somehow built an alchemy device, and "monetized" its debt, that result would cause enormous inflation. The market value of all its debts would zoom back up to face value, everyone's balance sheet would bulge, and people could afford to spend more and bid up the prices of goods.

But the U.S. is not under a gold standard. When USG issues new T-Bills, its credit rating does not drop, nor does the market value of existing t-bills drop. Since the U.S.G. can issue an arbitrary number of T-Bills without ruining its credit rating, the treasuries cause inflation at the time they are issued.

So deficits can indeed be a large problem. The deficit to GDP ratio is one component of the overall inflation rate. If that gets too high, bad things will happen. On the plus, side the government's budget is naturally counter-cyclical. So as inflation heats up, tax revenues rise and the deficit falls.

But the overall level of "debt" is not a problem. Once the "debt" is issued, it's water under the bridge. "Paying off the debt" would be deflationary, and as destructive as contracting the money supply sharply.

One final way to look at it:

Imagine the Federal Reserve stopped calling "the national debt" the national debt. They simply called it "T-Bills outstanding" and updated their statistics to show T-Bills as a component of the money supply, rather than as debt. All government publications were purged of the word "debt" and replaced by "T-Bill". T-Bills are interesting paying dollars, when they mature, the T-Bill disappears and you get a dollar. Or you can trade your dollar in for a another T-Bill.

If the Fed did this, absolutely nothing would change. This is already how the monetary function system functions. Thus with a simple change in nomenclature, the national debt disappears!

Jeff writes:


The question is not whether t-bills (or dollars) are money, whatever that might mean; the question is how does the US government continue to fund its overspend (and resolve its previous overspend) relative to total domestic productivity? It can only resolve its obligations relative to prior and future, already-committed spending in three ways: "pay" the absolute amount (and future amounts) with an undiluted instrument --- requiring greater confiscation of existing units of value, i.e. claiming a greater part of past and future productivity (or the same or lesser part of a future, increased productivity); pay the absolute amount of the obligation with a diluted instrument, which it does by greater issuance of t-bills as noted, causing inflation; or write it off. And in doing either of the latter, beyond some point it dramatically impairs its ability to continue to fund future spending, the obligations of much of which are already incurred.

Again, whether we're talking t-bills or dollars is irrelevant. The key quantities in question are: already incurred obligation vs. total productivity, and the "cost" of future unfunded promises relative to total productivity. As noted, both costs and productivity are denoted in the same units which makes the ratio itself *unitless.*

Consider winterspeak's "frequent flyer miles" analogy. Can e.g. American Airlines "run out" of frequent flyer miles? Of course it can; consider those miles demand-redeemable "shares" in the total (discountd; think NPV) future travel miles on offer. If American Airlines issues so many of them that, upon demand for their use, it could not transport enough passengers for enough "pay" to cover costs, for a long enough period of time, it would be financially unable to continue. It could print as many of them as it wanted; but in doing so, it would find that its takers were increasingly disinterested (as the likelihood of redeeming them decreased, or the cost in redeemed miles per unit of travel increased). And in any case, at some ratio of obliged travel-miles vs. total travel miles available, it would be financially defunct.

In part, the appetite for US treasuries as an "investment vehicle" by foreign governments is a function of two things: first, it has historically been a good "sink" for "excess foreign capital." Investing in US treasuries has been an anti-inflationary or inflationary-smoothing measure for foreign governments, particularly in transitional and developing economies where rapidly-growing productivity leads to inflation that would destabilize their economy. Second, it rests on the de facto use of the dollar as the mostly-central currency / value store.

The former is likely to decrease as the level of (and ongoing slope of gains in) productivity in these foreign nations begins to more resemble our own; it is also likely to increase as the unitless ratio of the ongoing accumulation of these investments vs. our total productivity begins to reach some (unknown) level. And finally, it should be clear that the global appetite for allowing the dollar to remain the de facto central currency or value-store will decrease as its absolute value is increasingly diluted.

You are correct that to date "the market" (i.e., creditors or purchasers of US debt instruments) has not demanded dramatically higher interest or "lowered our credit rating" as a function of ongoing dilution *to date.* But that's a limit process bounded by discounted total future productivity, and at some ratio of government "consumption" vs. collection vs. productivity that could easily change.

The conclusion is inescapable: the present levels and growth of obligation by the US government are untenable without some major event that fundamentally changes the overall macro picture: either productivity needs a "magic happens here" moment; value taken in by the government as a fraction of total productivity needs to increase; massive dilution of value-bearing US instruments needs take place, or the obligations need to be wound down and, perhaps, written off. Some combination of the above is inevitable.


Shayne Cook writes:

It frankly amazes me that no one has seriously proposed a dramatically increased gasoline/diesel/motor-fuels tax to at least partially address the debt/deficit issue, either within economics circles or political circles.

Before the "flaming" begins, I don't advocate such a measure, I'm merely stating I'm surprised it hasn't been prominently discussed or proposed.

Jeff writes:

One small correction and one clarification.

I said:

"The former is likely to decrease as the level of (and ongoing slope of gains in) productivity in these foreign nations begins to more resemble our own; it is also likely to increase as..."

That "increase" should be "decrease."

Second, as noted, monetary inflation (i.e., value-dilution of the unit of measurement of the dollar) happens at the time of issuance of e.g. a t-bill, not at redemption time. In this we're agreed; I think our major disagreement is in how long the government can *continue* to play this game vs. total future productivity simply by selling more t-bills without changing anything else.

The game is one of relative value all the way around the world. Our ability to deficit-spend for most of the last forty years is due largely to unique factors about our productivity vs. the rest of the world's, and the unique use of the dollar as a global value-store that has stemmed from that relationship. There's absolutely no guarantee that these relationships are fixed, and every reason to believe that they are not.

Once these relationships are fundamentally altered --- and there's every indication that we're at the beginning of that process now --- then things could very quickly become highly unstable.



Deryl G writes:


This is probably a pretty basic question, but I'm hoping you can help me out. You wrote "...rapidly-growing productivity leads to inflation ...". I would have thought increased productivity would lead to deflation. If there is a sufficient competition, being able to produce more with less should lower prices. What am I missing?


John S writes:

Those aren't odds, those are probabilities. Odds are something like 3:1 (3 to 1) or (1 to 3).

Joe writes:

The American Peopel should not be forced to pay off the debt for a mistaken war in Iraq and Reagans love for Military spending.

Deryl G writes:


What kind of debt is okay to force the American People to pay off?

winterspeak writes:

WILMOT: "Monetize the debt" is a gold standard notion and is not applicable in the floating fx, non-convertible regime the US has.

JEFF: I certainly agree that inflation is a possibility, but not default, and the two are not the same. Likewise, size of Government is a separate issue, it can get larger (likely) or smaller (unlikely).

Nevertheless, you still seem to believe that the Govt (currency issuer) needs to tax people (currency users) in order to have currency. When put in so plain a way, doesn't that seem backwards to you? Why should an entity that can create currency need to *take* it from people who can't in order to have have currency?

In fact it is the opposite.

I wish they would rename the "deficit" "paid-out capital" because it more accurately describes how the Nation Debt must, to the penny, equal the sum of all private sector "paid-in equity" balance sheet entries. Government deficits FUND private savings.

Government tax NOT to have money to spend -- they are currency issuers, remember -- but instead to create demand for their currency AND to neutralize the inflationary impact of their spending.

An economy produces total goods and services worth Y. The Government imposes a 30% tax on the economy. The Economy is no longer wealthy enough to purchase all of its own output, so the Government can step in and buy the last 30% without triggering inflation.

MHodak writes:

Sorry to break the news to you all, but the odds of a U.S. default is 95%.

The only thing that prevents it from being a mathematical certainty is a productivity revolution that pulls us to significantly higher than the 3 to 4 percent economic growth we have historically experienced. Fat chance if the rate of government spending as % of GDP continues to go up, and especially if the area of greatest economic promise (biotech) gets throttled by some version of ObamaCare.

Trying to tax enough income or wealth to make up the gap would wreck our economy, meaning it would not yield enough to make up for what we owe.

Spending cuts are out of the question, no matter how "resolved" we are.

Selling assets its a laugh. We don't have $100 trillion in assets to make up for the current unfunded liabilities that will drive us to bankruptcy.

For those of you who think we can print our way out of debt, you're deluded. We currently need to roll $200 billion per month to keep up with our unfunded obligations. According to government projections, that level should eventually go down to $100 billion per month, if you believe that will happen with a Congress that does not believe that deficits matter. The prices at which we would have to pay for even $100B per month would spiral out of control if we start inflating to the degree necessary to monetize even a fraction of our existing debt. We could never get ahead of that curve. Also, the political calculus would drive default sooner than inflation to that degree, which would have a similar impact to a wealth tax, except broader-based.

Unfortunately, there is no plausible scenario around a default some time in the next 20-30 years. Our only hope is that the rest of the world can get out of the way of this impending train wreck enough to help us pick up the pieces on the other side.

Jeff writes:


I would state that eventual default is the *least likely* and *final* event in a terminal sequence; however, if the *growth* of public consumption of GDP continues unabated, eventually it consumes all current and future GDP; at which point, presumably, the currency has been diluted to the point of virtual worthlessness, which is in effect a default whether or not it is declared as such. (I.e., whether the default is de facto or official doesn't make much difference; if the creditors stop sopping it up, game over.) And at that point, we can't very well continue to grow public obligation relative to productivity via any means.

Re: the issue of taxing the currency holder, note that inflation is itself a form of taxation: and arguably a very regressive and detrimental one. You're almost there in your last pseudo-equation; the problem is that by stepping in the government *automatically* dilutes the existing monetary base, which is de facto inflation. At some level of consumption by the government it simply consumes all there is to consume. And when that amount exceeds the total discounted future productivity, it collapses.


"If there is a sufficient competition, being able to produce more with less should lower prices."

Increased productivity also leads to increased domestic demand for the product itself; a fixed number of people producing a larger amount of product still leads to increased competition for allocation of that product --- consider the increase in average home size over the last 40 years, or disproportionately-escalating home prices in "high-tech" cities during the dot-com boom. With increased productivity comes an increased demand for "quality of life" improvements; and worse, "wealth" is a relative thing psychologically, so everyone vies not just for more absolutely, but more relatively.

This balances itself out over time as the society settles into some kind of chaotic equilibrium with respect to allocation; however in the short term, until that equilibrium is reached, you see price inflation. This happens when the rate of productivity increase is greatest relative to its short-term historical average value.



Jeff writes:

One other bit, for the misguided neo-Keynesians out there... ;-)

"At some level of consumption by the government it simply consumes all there is to consume. And when that amount exceeds the total discounted future productivity, it collapses."

Of course government utilization of domestic product is not in itself inherently unproductive; government *spends* the units of value it collects, contributing to productivity... this is a recursive process, but it's obviously also a limit process. At each iteration, some small bit of efficiency is eliminated, i.e., some small amount of total productivity is irrevocably used up.

In theory this amount is small enough that many iterations (i.e., much more "consumption" by government) would be tolerable. In practice, the private productivity declines dramatically as the fraction of total productivity attempted by the government increases; at some point far short of the theoretical limit, collapse would seem to be inevitable. (We've got ample historical existence proofs, if only we would pay attention.)

$0.02, and IMHO...


Shayne Cook writes:

Deryl G asked above: "What kind of debt is okay to force the American People to pay off?"

What a marvelous question!

Thank you, Deryl G!

Devin Finbarr writes:


Productivity has nothing to do with any of this. The dollar is not dividend producing equity. It is simply a collectible, a medium of exchange. People hold a particular medium of exchange because they think it will be a good store of value. An item is a good store of value if a) its dilution rate is lower than all other stores of value and b) other people wish to hold it as a store of value. b) depends on laws, military power, and a).

Thus what matters is the dilution rate of dollars compared to alternative stores of value (such as gold or real estate). Even if productivity stagnates, if the dilution rate of dollars is less than that of gold, people would still hold dollars. Actually the dilution rate of dollars can be considerably higher, since dollars have better tax treatment and have a lot of other legal privileges.

Conversely, even if productivity grew at 10% a year, the government would be foolish to dilute at 10% a year. While the value of a dollar relative to a basket of goods might be stable, its value relative to gold would fall. So people would exchange dollars for gold, possibly triggering an outright currency run.

Thus the number that matter is the dilution rate of the currency. In fiat world, the currency consists of all government backed notes - FDIC insured accounts, treasury, greenbacks, etc. The total outstanding quantity (minus double counting) is something like ~$15 trillion. So what matters is the deficit as a percentage of the outstanding money stock. A ~3% dilution rate could be stable indefinitely. Even a ~5% might be stable for a long time, although the economy will be more prone to bubbles and the Fed might have to tighten now and then to break inflationary spirals. Once you pass 7% or 8% you start to really fry your currency and the risk of a currency run increases substantially.

Jeff writes:


"Productivity has nothing to do with any of this. The dollar is not dividend producing equity. It is simply a collectible, a medium of exchange."

Of course it has to do with productivity, though I agree a dollar is not an equity. It is a hypothecated share (as in potentially exchangeable for some portion of, not as in actual equity in) of some unspecified quantity of future assets and / or productivity. You only hold one in order to exchange it with somebody at some future time for some other kind of value; and that value can clearly come from a variety of sources (i.e., ongoing labor contribution increases total available future value --- we call that part of it "productivity" and, as it is variable whereas fixed assets such as land are generally not, it is an essential part.)

You are correct that part of the overall value picture on which the dollar rests as a medium of exchange includes the value of fixed assets; but the total future value even of those fixed assets rests on e.g. productivity increases. I.e., an acre of farmland has more intrinsic value now than it did in 1800, because with better technology and ongoing labor we increase future anticipated productive output of the asset. The amount of acreage available has not really changed all that much, but how productive an acre can be certainly has, as has the supply / demand picture for same... (Supply / demand vs. consumption also play into changing value for fixed assets, cf. total oil, metals, etc.)

I completely agree with your general point, if not the details of the conclusion. You can't just look at the money supply alone, as it fails to capture the impact of labor, technology, or even necessarily those fixed assets already valued and represented within that supply. (I'm not actually offering up a grand unified theory, I'm merely attempting to emphasize the importance of expectations about the future in determining the value of a fiat currency at any given point in time.)


winterspeak writes:

JEFF: No, default is not the final event in a terminal sequence. It need *never* happen in a fiat, non-convertible, floating fx, domestic obligation regime.

It is impossible to consume "future" GDP. Are we sending stuff back in time to the 1940s?

An economy, in a period, gets to consume whatever it produces, in that period. You can store stuff for the future, but you cannot move stuff from the present into the past.

If Govt is 100% of economic activity, it can simply direct output and does not need a currency at all. A completely planned economy does not need price signals. And it will have access to whatever it can produce (which I don't think will be very much).

"Deryl G asked above: "What kind of debt is okay to force the American People to pay off?"

American people will never "pay off" the national debt, as the national debt is what gives them the currency they need to save in the first place.

The private sector needs to get money to save from somewhere. The place it gets it from is the Govt. When the Govt gives the private sector money, it's called Government spending. When the Government takes money from the private sector, it's called taxation. The different between the two is the deficit (or surplus). If the Government is net funding private sector savings, it must run a DEFICIT to do so. If the Government is DRAINING the private sector of net savings, it must run a SURPLUS to do so.

Keith writes:

My opinion is that the odds he places on the scenarios shows a delusional understanding of current economic reality. My thought, massive deflation first, hyperinflation several years down the road. Dow 1000, here we come.

Jeff writes:


Don't be ridiculous, nobody's suggesting economic time travel. It's possible to consume future productivity by spending it today, by borrowing; this is what we do every time we e.g. make a purchase on a credit card. Further, the total of future productivity has a present value; this is the familiar "net present value" concept. Clearly the further out you go, the lower the present value gets; so this is a finite quantity of *value* that can be borrowed against.

Two other bits: I fear that you exaggerate the role of the government in overall creation of the spendable currency. Through our fractional reserve banking, the private sector generates a large part of the required money itself.

As for the invulnerability of fiat currency, I would suggest that you may be splitting definitional hairs. Can a nation with a fiat currency become *insolvent?* Of course it can, and there are many historical examples, some quite current and relevant (cf. Iceland. Also cf. Merkel (disagrees with you) and this Spiegel article, for starters:,1518,604523,00.html

To wit:

"There's a rumor going around that states cannot go bankrupt," German Chancellor Angela Merkel said recently at a private bank event in Frankfurt. "This rumor is not true."


winterspeak writes:

KEITH: There's a good amount of delusion going on, for sure.

Jim Glass writes:

There is one and only one solution. Remember that S&P projected US Treasury bonds would be "junk" by 2027, and since then Obama's accelerated that debt growth by a good half-dozen years, so there is not a whole lot of time left to get there...

1. Muddle through. No major change in policy.

0% chance by arithmetic.

2. Technology to the rescue. Some major technologies, probably either wet or dry nanotech, produce so much economic growth that the ratio of debt to GDP...

Dreams are nice. But when per capita GDP has been increasing 2% a year for near 200 years at a remarkably steady rate, why would it explode upward now?. Why not 50 years ago? Or 50 years from now, when it will be far too late?

Note that even a surge in the growth rate large enough to be deemed "huge, unprecedented" still won't work, because Social Security benefits are tied to wages which rise with the growth rate, and demand for health benefits rises faster than income, which rises with the growth rate.

We need a specific "miracle technology breakthrough" -- one that slashes the dollar cost of health care. Since it hasn't happened in 200 years, I give less than 0.5% chance.

3. Policy changes. Congress increases taxes ... and/or takes steps to rein in Medicare and Social Security spending.

Nothing else is possible by process of elimination. There's an easy way (Congress responsibly does this in advance to head off calamity) and a hard way (Congress irresponsibly delays until calamity arrives and then does it to avoid worse calamity) and I'd certainly more expect the latter. But this is going to happen, because in the end there are no other options.

And to paraphrase Sherlock, when you eliminate everything that won't work...

4. Inflate away the debt with moderate inflation (between 5 and 10 percent per year).

0% chance. Impossible. Won't work because >$50 trillion of the >$60 trillion of US liabilities are for Social Security, which is inflation-adjusted, and health care, the cost of which rises faster than inflation, and unfunded federal/military retiree pensions, which are inflation-adjusted.

And it won't work even for Treasury debt, because "inflating away" debt only works if you actually pay it off with inflated dollars.

Lenders aren't total dummies. If you roll over debt after you inflate, you get clobbered rocketing interest rates and wind up back where you started, cash-flow wise. And if you keep borrowing more ... fuhgetaboutit.

To pay off debt (even with inflated dollars) and stop borrowing requires a massive change in real fiscal policy -- you still have to push revenue above spending.

5. Wealth tax. The government takes, say, 5 percent of everyone's personal assets above $100,000

0% chance. A one-time fix won't work. Future expenditures rise above revenues forever at an accelerating rate. Literally exponentially as interest compounds. No one-time fix will do, be anywhere close to large enough.

6. Hyperinflation.

Won't work for the same reason as "inflation". That's the "beauty" of owing all your big debts in inflation-adjusted and real terms, regardless of nominal dollar cost.

7. Default. The U.S. simply refuses to pay some or all of its debt.

Even if it happens it doesn't give any money to pay Medicare/ Social Security/ Medicaid/ unfunded federal-military pensions. So it is no solution.

Remember, the fiscal killer is not the $10 trillion or whatever is projected of Treasury debt, but the >$50 trillion current value liability for Medicare, Social Security, Medicaid, & unfunded federal and military pensions. Even walking away from $10 trillion doesn't give the money to pay $50 trillion.

The ONLY solution for paying this >$50 trillion is increasing taxes to pay promised benefits or cutting the amount of them paid (either by reducing benefits by formula or just not paying, "defaulting" that way) or -- in reality -- both.

To see the rough size of tax increases/benefits cuts needed to keep the government solvent, and for perspective on them compared to past budget events (like World War II) here are some numbers.

Mark Bahner writes:

"Productivity has nothing to do with any of this."

Productivity is very important to this question. If productivity grows "double exponentially"--as Ray Kurzweil and 200+ years of economic history suggest it will--then money itself may become irrelevant even towards the middle or end of this century.

Here are my predictions for U.S. per-capita GDP through the end of the century, assuming that the U.S. GDP grows at a rate equal to the world average per-capita GDP growth rate. Note that *all* values are in year 2000 dollars. The decadal growth rates are for the decade ending in the year stated:

Year, Decadal Growth Rate (%), Per Capita GDP (2000$)

2000, N/A, $35,000;
2010, 2.0, $42,000;
2020, 3.5, $59,000;
2030, 4.5, $92,000;
2040, 6.0, $165,000;
2050, 8.0, $356,000;
2060, 11.0, >$1 MILLION;
2070,...who cares? (Money has become irrelevant.)

Mark Bahner writes:

"But when per capita GDP has been increasing 2% a year for near 200 years at a remarkably steady rate, why would it explode upward now?"

Two answers:

1) Per capita GDP has not been "increasing by 2% a year for near 200 years at a remarkably steady rate." At least not world per capita GDP. World per capita GDP has been increasing "double exponentially" for the last 200 years.

Here are world per-capita GDP growth rates for the 50 years ending in the year stated:

1800: 0.18%
1850: 0.87%
1900: 1.65%
1950: 1.76%
2000: 2.83%

2) As to why "it would explode upward now": in one word, "computers."

That's the short answer. The long answer is:
I'd say less than 0.1 percent of the people on this planet, including less than 1 percent of economists, really know what creates wealth. What creates wealth? Julian Simon knew: (free) human minds create wealth. And Ray Kurzweil knows computers.

Combining Ray Kurzweil's estimates of the computing power of $1000 with the number of personal computers sold each year provides an estimate of the number of "human brain equivalents" added to the human population (by personal computers) each year.

The numbers are:

1990: 2000: 50
2010: 50,000
2020: 50,000,000
2030: 100 billion
2040: 1 quadrillion

Needless to say, adding 100 billion or 1 quadrillion to the number of human brains each year should result in pretty spectacular economic growth.

Mark Bahner writes:

The value for 1990 should have been less than 1 human brain equivalent. (And 50 human brain equivalents in 2000, and so on...)

Devin Finbarr writes:


Productivity matters very much from a standard of living perspective. If you're trying to figure out how much your dollar will buy you in 2020, productivity growth rates matter.

But productivity does not matter from the point of view of the sustainability of deficits (which is the issue we are discussing). If you have 15% productivity rate growth, and 150% dilution rate, you'll still get a currency run and dollars will become worthless. If you have a 2% dilution rate, and a negative 3% productivity growth rate, you will not get a currency run and the dilution rate is perfectly sustainable. Your population might rise up and have your head for other reasons, but not because of the deficit.

winterspeak writes:

Devin: You are quite correct that productivity matters from a standard of living perspective, but does not matter from a POV of sustainability of deficits.

The Federal deficit is what FUNDS private savings. We, the private sector, get the money we need to net save, from the Government SPENDING MORE than it taxes.

I'm not going to say any more, because this is a pretty simple and basic point that underlines the fact that the US does not have a gold standard, or anything gold standard like, which Arnold et al don't quite seem to embrace. Latest line: "The singularity is a scenario under which government debt is not a problem." Reality is also a scenario under which government debt is not a problem -- just ask Japan. Not that I wouldn't mind a singularity -- maybe suddenly everyone would understand accouting!

Mark Bahner writes:

"But productivity does not matter from the point of view of the sustainability of deficits (which is the issue we are discussing)."

Productivity does matter from the point of view of sustainability of deficits. If productivity is doubling every decade, annual deficits of $1 trillion are trivial. If productivity is growing by 1 percent per year, annual deficits of $1 trillion are...well, definitely not trivial.

winterspeak writes:

MARK: You are not correct.

The US can self finance any deficit it runs without any operational problem. What you can buy with that money is a different question.

Productivity tells you what the economy can create (real).

Deficit is simply the amount of paid in capital the Government has issued to fund the private sector. If it's too high, you'll get inflation. If it's too low, you'll get unemployment and deflation.

Imagine the US economy with 300M seniors on medicare and only 1 person working. Is the problem *really* whether the seniors will have enough money to pay the 1 worker? Clearly the problem is if that 1 worker is productive enough to produce everything the 300M seniors want to buy.

We are no longer on a gold standard.

EricR writes:

Hi Dr. Kling,

I've been very interested in scenario #2 for a while now (technology coming to the rescue). I browsed (can't say I read it entirely - as non-scientist it's really beyond me) Ray Kurzweil's "The Singularity is Near" and became very interested in the technology that could be coming down the road, and soon. Kurzweil is big on three revolutions:

1. Biotechnology - life extension
2. Nanontechnology - tiny machines
3. Artificial Intelligence

If any of these three really develops in the way K predicts, our whole notion of what an economy is will change. Imagine, as K predicts by around 2040, if we had the ability to manipulate matter at the atomic level? If resources were not an issue, what would happen to our economies?

Also, I think not enough press has been given to the solar power revolution that is coming. Given the technology curve and its rate of progress (similar to Moore's Law), K and other leading scientists see solar competitive with the electronic grid within 4-5 years. Who knows the impact this will have, especially if electric cars really take off. This could massively reduce our dependence on foreign oil and make us a net energy exporter (jobs!) instead of an importer.

Anyway, if you are interested in the advances being made, sign up for the newsletter at the link above. It's just a collection of daily stories in all kinds of tech areas.

Mark Bahner writes:

"MARK: You are not correct.

The US can self finance any deficit it runs without any operational problem. What you can buy with that money is a different question."

Whether "the US can self finance any deficit it runs without any operational problem" is irrelevant. What one can buy with that money is the important question! That's why productivity matters.

No one doubts that the U.S. can print a quadrillion dollars. The Weimar Republic (and Zimbabwe) have shown that's possible. But the more dollars printed, the less they are worth.

"Imagine the US economy with 300M seniors on medicare and only 1 person working. Is the problem *really* whether the seniors will have enough money to pay the 1 worker? Clearly the problem is if that 1 worker is productive enough to produce everything the 300M seniors want to buy."

Yes, I agree. But you've essentially switched to my it's not surprising I agree. You're arguing that productivity really IS what's important.

And by the way of which...imagine the U.S. economy with 300 million people and no one working. Is that possible? The answer is that it is possible, if robots can do everything humans can do.

winterspeak writes:

MARK: "Whether "the US can self finance any deficit it runs without any operational problem" is irrelevant. What one can buy with that money is the important question! That's why productivity matters."

Not meeting your nominal obligations is the definition of "default". That was the topic of this thread, as well as all the previous posts. This is a real risk on a gold standard, and it is a zero risk on the fiat currency that the US has currently.

All kinds of ridiculous policies are being proscribed because people, including Arnold and Jeff Hummel (who, may the saints preserve us, teaches a University level macro class) believe that default is possible, indeed, inevitable.

I think productivity is important too, but I also recognize that all the misconceptions about default are actively *harming* productivity.

MARTY: It makes no difference to the sustainability of the dollar if it loses its status as reserve currency. The US Treasury does not need to actually issue any debt, and I wish it would stop, at least for a while, just to put an end to THAT myth.

BofE is actually considering this.

Jim Glass writes:

Not meeting your nominal obligations is the definition of "default".

Saying that a nation can choose to avoid technical default through hyperinflation and the resulting destruction of its national credit standing and financial system is not particularly "productive" as far as policy analysis goes.

Note also that government policy is driven by politics, and the inflation option is not in fact always available even with a fiat currency. Russia had a fiat currency when in 1997 it chose default over inflation because its leaders feared the wrath of its external creditors less than that of its internal citizenry.

Also as I noted before, "printing money" in an inflationary manner in no way solves the debt problem of a nation that continues to run deficits, because it must roll over existing debt and issue new debt, on both of which lenders charge a sufficiently increased interest rate to cover the inflation.

All kinds of ridiculous policies are being proscribed because people, including Arnold and Jeff Hummel (who, may the saints preserve us, teaches a University level macro class) believe that default is possible, indeed, inevitable.

I'd hope that ridiculous policies would be proscribed, but apart from that...

Some perspective: The explicit and "implicit" (for unfunded Medicare, Social Security, Medicaid, federal/military pensions) obligations of the US gov't at the end of 2008 were $64 trillion discounted to present value (some estimates are much higher), for 2009 it will be $69 trillion or so. US debt is financed with income taxes -- because cash interest must be paid on it. The Treasury, SS Trustees, et al, estimate 6% as the average long-term future rate. This interest rate on $69 trillion, divided by the 80 million payers of income tax in the US, gives an average tax bill to finance the interest larger than US average household income (about $50,000).

One can consider this a back-of-the-envelope calculation of a ballpark number for a very big ball park and still see immediately that it is totally unsustainable.

Of course, taxpayers aren't paying cash interest on all this yet, only on $7.4 trillion of it today, but another $9 trillion will be added in the next 10 years, and then the figures rocket straight up as boomers retire and "implicit" debt becomes "explicit" to pay for it. Which is why both Moody's and S&P have projected the credit rating of the US starting to fall by 2017, with S&P projecting it will hit "junk" by 2027 (on pre-Obama, pre-recession policy) as per my earlier link.

Now, alas, $62 trillion of the $69 trillion of obligations are not in "nominal" dollars but are defined in real terms -- wage level adjusted, inflation adjusted, real services. This means that, even with a fiat currency, not on the gold standard, "printing money" is no option at all to meet these obligations.

The obligation is to transfer to the beneficiaries of these programs increasing percentage point shares of real GDP (not nominal dollars) from taxpayers, the only persons from whom government can make a transfer of GDP.

That means legislated much-higher taxes in real, uninflated dollars/ legislated much-lower benefits in the same dollars, or combination thereof (in amounts approximated in the link I provided earlier) -- or indeed, absent such legislation should it prove politically impossible, "default" on these obligations and a lot of unhappiness in the streets. It has happened before.

The disposition of this totally unsustainable $62 trillion present value of implicit debt is what persons who worry about the budget worry about. Absent that, nobody thinks the US would default on the current $7 trillion of Treasury debt, or even on $16 trillion of it.

winterspeak writes:

JIM: Have you considered, even for a moment, that it is possible for the Government to print money in a non-deflationary manner?

Hasn't Japan, which has printed money like mad for 20 years, but is still in deflation, not made you think a little bit harder about the link between Gov deficits and private savings?

If not Japan, then how about the US now?

The term "national credit" is an oxymoron -- still more gold standard thinking. Fiat currency CANNOT default.

Russia's 1997 rouble was in a fixed-fx rate, making ultimate default inevitable. The US$ is on a floating-fx scheme, and is non-convertible, making default IMPOSSIBLE.

You keep talking about the US needing to borrow money. It simply does not. The US is not on a gold standard. The US could stop issuing bonds tomorrow, and it would have no problem deficit spending in perpetuity, so long as it was OK having zero interest rates (which it does now, and Japan has had for 20 deflationary years).

You also say that the US debt is financed with income taxes. It is not. The US Govt is a currency issuer -- why on earth does it need your dollars? It can make it's own.

Tax is an anti-inflationary measure, not a funding mechanism, at the Federal level.

I am pleased to see there are so many people focused on saving the US financial system. I wish a few more of then actually understood how it worked.

Mark Bahner writes:

"JIM: Have you considered, even for a moment, that it is possible for the Government to print money in a non-deflationary manner?"

I think we all can agree that the government can print money in a "non-deflationary" manner. It's the "non-inflationary" manner that seems problematic.

Let's say that technology does not save us by promoting per-capita growth rates that I've predicted for the world (and presumably the U.S. could meet the world average).

In fact, suppose all of debts Jim has quoted do come to pass:

Now, alas, $62 trillion of the $69 trillion of obligations are not in "nominal" dollars but are defined in real terms -- wage level adjusted, inflation adjusted, real services.

How does the federal government print money to cover those debts without causing massive inflation (particular since the debts are wage- and inflation-adjusted)?

Mark Bahner writes:

I should have put this after my snark on "non-inflationary" vs "non-deflationary."


winterspeak writes:

Mark: Hah! I must have deflation on the brain.

Suppose per-capita growth is stagnant.

Now suppose that Government prints $1T and gives it to the private sector. The private sector then saves it.

How will this contribute to inflation?

Inflation is too many goods *chasing* too few products, emphasis on the *chasing*. If it's saved, it has no inflationary impact.

Right now, the private sector is increasing its net saving. The *only* place this money can come from is the Federal deficit.

In fact, the national debt equals the sum of all net private sector savings to the penny.

Mark Bahner writes:
Inflation is too many goods *chasing* too few products, emphasis on the *chasing*. If it's saved, it has no inflationary impact.

Let's leave aside the question of whether money printed and given to people for Social Security would be saved (in general, it wouldn't). There's the more fundamental problem that, in a fractional reserve system, money deposited for savings in a bank actually creates new money. Potentially, the amount created is up to 1 divided by the required reserve ratio times the initial amount deposited for savings. Add http and www for this website:

So you seem to be claiming that money printed by the federal government will simply be saved by private individuals, and will cause no inflation. But it will cause inflation, even if the money given to individuals is saved, because the U.S. banking system is a fractional reserve system. Banks will use the money deposited for savings to create more money, which will be spent. More money chasing the same amount of goods equals inflation.

George in SA writes:

When a business is in this kind of trouble, what does it do to stay alive? It sells assets to raise money and pare its debt.

It would seem that if the population came to understand that drastic measures were needed, selling assets would be one way to reduce the debt load. What kind of assets? Natural resources, like oil, coal and gas, as well as privatizing assets, maybe even selling some national parks to foreigners. Texas sold most of the panhandle in the late 19th century, and it worked out very successfully for that state.

The key is that citizens would have to understand that we could not continue in our extreme environmental protectionist ways and avoid default.

Comments for this entry have been closed
Return to top