A silly question for anti-austerians

Let's make up some silly numbers.

Suppose the national debt was, let's say, 1,000% (ten times) annual GDP. And suppose the budget deficit was, let's say, 50% of GDP. And suppose your economy hit the Zero Lower Bound, and suppose you thought that your own central bank's monetary policy could do no more to increase aggregate demand, but more aggregate demand was needed.

1. Would those numbers lead you to hesitate before recommending tax cuts or increased government spending to increase aggregate demand?

2. If I divided both those numbers by fifty, would you still hesitate?

If you answered "yes" to the first question, the punchline writes itself: we have already established that you are an austerian; we are just haggling over the price.

I'm not looking for some sharp dividing line, because I don't think there is one. (There is nothing magic about a 100% debt/GDP ratio, for example, because if we measured time in months instead of years, 100% debt/GDP [edit: monthly] would become 8.333% debt/GDP [edit: annually], and 8.333 doesn't look like a magic number.) But I am looking for some sort of recognition that there's some sort of convex trade-off, or increasing marginal costs of debts and deficits. And the slope and curvature of that trade-off, or the height and steepness of that marginal cost curve, may not be easy to estimate accurately.

Is the difference between "austerians", and those who accuse others of being "austerians", merely a matter of degree?

I think it is.

Comments

[louis]

Question for austerians:
If NGDP is growing at 5% annually, debt is 90% of GDP, and the deficit is 4%, would you argue for consolidation?
If suddenly the NGDP is falling at -2% annually, do you advocate more or less fervently for consolidation?
The austerians I've seen in the public sphere trotted out their arguments for fiscal consolidation after the recession hit. The anti-austerians are arguing that they have the wrong sign on their reaction function.

[Jason Smith]

Why is annual GDP important relative to a level? Is this nation-state only going to be around for a year?

I think debt thought experiments like this could only be meaningful if humans had a lifetime on the order of a nation-state. Going with your gut feelings is going to produce something that is strongly dependent on human lifetimes.

Imagine you will live for eternity. How much debt would you take on? There are only two answers to this: zero or infinite. Any finite non-zero amount is meaningless.

[Britonomist]

"1. Would those numbers lead you to hesitate before recommending tax cuts or increased government spending to increase aggregate demand?"

Those numbers would lead me out of that country or to building a bomb shelter :).

But in all seriousness, a part of me says "YES! absolutely", assuming it would have to be debt financed. On the other hand, if debts and deficits are so ludicrously high, what does it matter? Any more increase in spending is a tiny drop in the ocean of debt and wouldn't affect my overall level of anxiety one bit.

I think it's important you specify what's causing the insanely high deficit. Is it some recklessly irresponsible government that just chose a gigantic level of discretionary spending? Or has it arose from a collapse of the tax base - essentially an economic collapse?

If the former, I would absolutely advocate austerity. I believe that extremely high government spending (far beyond levels will scene in western economies today) could be theoretically contractionary, and probably one of the main causes of the economy being at the zero lower bound. Of course in the real world you practically never see government spending at these extreme levels other than in a disaster or war perhaps.

If the latter, austerity while keeping the same government and economic structure would be absolutely hopeless; no amount of surpluses and contractionary fiscal policy would bring the burden of debt down and restore the tax base in any reasonable amount of time. I'd be advocating default or monetization combined with a radical restructure of government and society, I would have considered the nation to have 'failed' and we would have to start again from the ground up.

"2. If I divided both those numbers by fifty, would you still hesitate?"

I probably would not hesitate (assuming we were in a recession).

" But I am looking for some sort of recognition that there's some sort of convex trade-off, or increasing marginal costs of debts and deficits."

I agree that there probably is a convex trade-off, but that's because I'm partly old fashioned. Not everyone will, I'm certain some post-Keynesians would disagree: they might regard debt as an irrelevant by-product, and might advocate permanently near zero rates, and even debt monetization if necessary.

I think it's important to distinguish between high deficits and high debt, those are actually separate problems: an extremely high public debt might not be so bad if you're able to suppress the interest burden (there's mechanisms to do this, even without permanently expansionary monetary policy). An extremely high deficit however usually represents a fundamental collapse of the tax base or massive government failure. Almost certainly, new money would need to be constantly created to finance the deficits, you would very soon if not immediately be running up to a hard inflation constraint; which means any post-Keynesian would agree that the deficit must be reduced to avoid high inflation. It's hard to imagine how in this scenario rates would be at the zero lower bound: if there was such enormous inflationary pressure, interest rates would want to go up, AD would be surging and Keynesians would not be advocating more spending in the first place.

I think it would have been better if you present scenarios only about debt, while assuming deficits were low in both scenarios.

[Britonomist]

Uh, not sure how I managed to confuse "seen" and "will scene", major brain fart there.

[Nick Rowe]

louis: fair point, but you are dodging the question by changing the subject and counterattacking. "The best form of defence is offence"? This is not a war between two sides. Getting the answer right, or at least thinking clearly about the question, matters.

Jason: A bond promises to pay one apple per year forever. How much is that bond worth today? Is it infinite? No. We use interest rates to compare stocks with flows. interest rates have the units 1/years.

[louis]

Nick - I don't think I am dodging the main question in your post.
You asked: "Is the difference between "austerians", and those who accuse others of being "austerians", merely a matter of degree?"
I answered that for at least some of the people we are talking about, it is a difference in sign rather than a difference in degree. Some people think that the answer to a recession and the ZLB is expanding the deficit, and some think that the answer is shrinking the deficit. Others think that the ZLB is an artifact of unevolved policy rules, but that's a different topic.
Your question as to how the costs of taking on incremental debt change as the stock of debt increases stands as an interesting one, and I made no attempt to answer it as I'm simply not informed enough to try. To me it feels a little like finance theory tossing in an omnibus "cost of financial distress" to aggregate the reasons why firms don't use more debt in the capital structure, despite the tax advantages.

[Nick Rowe]

Britonomist: I think you are right that those two numbers (debt/GDP and deficit/GDP) are not the whole story. For example, if GDP collapsed by 98% due to some purely temporary natural disaster, so that both ratios would drop fifty-fold next year, nobody should be panicking. But I think they are part of the story.

Maybe past irresponsible governments, who are now out of office and will never return, just for a clear thought-experiment. Or compare two countries, that are otherwise identical now and looking forward, except they had a different history of debt accumulation.

[foosion]

If more AD was needed, I would support tax cuts or increased govt spending (whichever is likely to be more effective).

It appears the market is not concerned about the debt or deficit, unless you're postulating a ZLB that somehow includes high interest rates on govt debt.

A temporary increase in the deficit should not matter in the long run. However, if the policy doesn't improve things in a reasonable amount of time, more thought would be needed.

Where is inflation in your scenario?

[Lord]

Probably depends on how much of that debt is owned by the central bank and how much new debt would be bought by them, but it would be more useful to focus on the bank's actions and membership of the board. I can see circumstances under which those levels of debt may be justified though they usually involve tech utopias. Cyclically averaged deficits exceeding the growth rate would also indicate a problem but this current deficit would have to be due to low gdp since war leads to excess demand.

[Market Fiscalist]

Suppose I take over the government in the situation you describe and have the following beliefs about economic policy.

- The national debt is way too high and we should aim pay it down at n% per year until it is a much lower % of GDP
- An NGDPT is good and needs to be hit every year
- Monetary policy should be used to hit the NGDPT whenever possible
- Monetary policy doesn't work an the ZLB
- At the ZLB , NGDPLT can still be hit by adjusting the size of govt deficit
- It is possible that in some years we will be at the ZLB , and the adjustments made to to the govt deficit will be so large that the national debt will increase
- These occasions will be so rare that they are perfectly consistent with a long term strategy to reduce the national debt

With this set of believes I could consistently be an anti-Austerian in the scenario you describe.

[louis]

Market Fiscalist - Additionally, anti-austerians Larry Summers and Brad Delong would add that larger deficits during ZLB periods may be consistent with lower long-term debt as well, using some back of the envelope math.

[foosion]

louis - anti-austerians tend to believe larger short-term deficits are consistent with lower long-term debt/gdp, mainly by growing the denominator. Counter-cyclical policy has been recommended even when not at the ZLB.

[Nick Rowe]

foosion: "It appears the market is not concerned about the debt or deficit, unless you're postulating a ZLB that somehow includes high interest rates on govt debt."

If people expect deflation, and/or worsening recession, that would bring market interest rates down. But if your fiscal policy works, nominal interest rates would rise.

Lord and MF: one can think up (implausible) scenarios where you would still want to increase the deficit. But, other things equal?

louis: so did Laffer, though with a different mechanism in mind. Along the Laffer curve there are two tax rates compatible with one level of tax revenue, so we can have an unstable equilibrium if the government targets tax revenue on the wrong side of the curve. I worry about stability of equilibrium in Brad deLong's and Larry Summer's model, even if you accept the parameters.

[foosion]

Nick: "If people expect deflation, and/or worsening recession, that would bring market interest rates down. But if your fiscal policy works, nominal interest rates would rise."

Don't reason from a price change or interest rate level, try to determine the underlying reason. :-)

There are countries with recessions and high nominal interest rates.

In any event, I'd have no problem increasing the short-term deficit under the circumstances.

[Ralph Musgrave]

Nick says, “But I am looking for some sort of recognition that there's some sort of…… increasing marginal costs of debts and deficits.”

I’m happy with MMT ideas on how interest on the debt increases with increasing debt, which are thus.

The stock of debt and base money can be less than the private sector wants to hold. In that case, full employment won’t be achieved even at a zero rate. Put another way, the private sector will try to save so as to acquire its desired stock, the result being paradox of thrift unemployment.

At the other extreme, if the above stock is more than the private sector wants to hold, the private sector will try to spend away the excess which will result in excess demand. In that case, interest on the debt will have to be raised so as to dissuade the private sector from spending away the excess.

As to the optimum position between those two extremes, I suggest that’s and maximum stock that’s possible without interest being driven above zero. Put another way, I see no point in the currency issuer issuing so much currency that interest rates have to be artificially raised in order to persuade people not to spend so much.

[Jason Smith]

Hi Nick,

Why would an immortal discount an interest paying bond at any finite discount rate?

The value of a bond paying one apple per year for eternity to an immortal is either effectively zero (discounted, doesn't matter the rate) or infinite (not discounted). Or there is a finite discount to every apple

If the only dimensionful quantities you have are two time scales T1 (lifetime) and T2 (discount period, time to pay off debt, interest rate, GDP, etc), then there are three possible combinations:

(1) T1/T2 = ∞ as T1 → ∞

or

(2) T2/T1 = 0 as T1 → ∞

and

(3) T2/T1 = 1 as T1 → ∞, T2 → ∞

For the apple interest, either every apple is discounted the same amount (including the current period apple), which is (3) or the apples aren't discounted (1), or they are (2).

In (3), relevant to the question in your post, it's the rate of debt accumulation that matters, not the level.

For an immortal, there is no time for the one apple per year to reach half its value. When would that be? Put two events on the timeline of your life. Now zoom out to a billion years. Those two events are simultaneous. Now zoom in to a nanosecond. One is infinitely far in the future and the other is infinitely far in the past.

[Market Fiscalist]

Assuming that monetary policy doesn't help you hit the NGDPT at the ZLB then (at the ZLB) there is a tradeoff between the "increasing marginal costs of debts and deficits" and the costs of the NGDPT miss.

I'm sure you could produce models that come out either way, but it does not seem unreasonable that it might come out that the costs of the NGDPT miss (suboptimal RGDP) outweigh the "increasing marginal costs of debts and deficits".

The key issue then become whether the assumption about monetary policy at the ZLB is valid or not. If it not valid then you can avoid both the costs of the the NGDPLT miss and the costs of the increased debt.

[Mike Sproul]

Here's a possible sharp dividing line between two groups:

1) The group that thinks "aggregate demand" is a meaningful concept, and that it can be "stimulated" when the government, or anyone else, spends money.
2) The group that thinks aggregate demand is a meaningless concept, that "stimulative spending" is just the broken window fallacy on a large scale, and that the occasional successes of stimulative spending were actually a case of the government creating a shortage of currency and causing a recession, and then relieving the shortage by printing money and spending it. ("Group" might be the wrong word here, since I seem to be the only member.)

[Roger Gomis]

My answer would depend on the multiplier and other factors ( most importantly the relation between deficit and interest rates), if the multiplier is high and interest near zero austerity would do little to reduce the debt/gdp ratio, it might even increase it. If the interest rate is 0 and assuming end of period notation:

Debt(t)/Y(t)=Debt(t-1)/Y(t)+Deficit(t)/Y(t)

transforming the first term on the RHS it becomes

d(t)=d(t-1)/(1+g)+Deficit(t)/Y(t)

where d is debt to gdp and g is the growth rate of the economy. Let g be a function of g=mu*(Deficit(t)/Y(t)) (mu=sort of multiplier), taking the derivative with respect to the Deficit(t)/Y(t) indicated by DELTA:

DELTA(d(t))=d(t-1)*(-mu)/((1+g)^2)+1

So the sign is not always obvious. However it is increasing with past debt to gdp ratio so if took this exercise seriously I would say the higher the debt to gdp ratio the higher the need for fiscal expansion. Of course with interest rates, denoted by r, it is different, taking a linear approximation (and using primary balance as the measure of deficit):

DELTA(d(t))=d(t-1)*(1-mu+DELTA(r))+1

so it really boil down to mu>mu it could still be that very little debt reduction is obtained for large drops in gdp, so it is not worth it. Of course the answer also depends on the dependence of future interest rate and future growth on current spending.

PS: Sorry if there are mistakes, I've done it in a rush.

[Majromax]

Wait...

> Suppose the national debt was, let's say, 1,000% (ten times) annual GDP. And suppose the budget deficit was, let's say, 50% of GDP. And suppose your economy hit the Zero Lower Bound, and suppose you thought that your own central bank's monetary policy could do no more to increase aggregate demand, but more aggregate demand was needed.

How does this happen?

Looking at the first two sentences, my mind immediately goes to "hyperinflation or default". Both of those would tend to increase interest rates, the first through the Fisher effect and the second through a risk premium.

But if interest rates on government debt are increased, then the economy should be nowhere in the vicinity of the ZLB.

At first blush, I can only conceive of this situation applying in bizarre circumstances. Maybe space aliens have promised to abduct anyone who holds onto pieces of green paper, for example, so there is zero demand for holding base money. But in a case like this, we'd expect the private economy to have also collapsed.

In that case, then maybe more deficit spending would be somehow appropriate?

[louis]

Nick: "I worry about stability of equilibrium in Brad deLong's and Larry Summer's model, even if you accept the parameters."
Very reasonable to do so! But you still must acknowledge that their conclusion is different from that of their Austerian cousins by more than a "matter of degree". They are not intimidated by a large stock of debt to shy away from fiscal stimulus, nor do they believe reducing the debt stock will help near-term growth for a country borrowing in its own currency.

[Ralph Musgrave]

Majromax,

You suggest that a debt of 1000% of GDP and a deficit of 50% could only occur where the private economy had collapsed. While the above numbers are obviously very extreme, they could occur at least in theory where the private economy was perfectly healthy, but there was a HUGE INCREASE in the stock of paper assets that people wanted to hold. I.e. peoples’ “savings desires” (to use an MMT phrase) had increased enormously, and the state was running a HUGE deficit so as to meet those desires.

[JF]

Assuming that the deficit is now so high because the economy has tanked and govt revenues tanked too, but there never was any counter-cyclical programs whose increasing outlays contribute to the deficit (this is an austerian nation). The 50% deficit is because of a shortfall in tax collections, even though modest spending cuts were enacted (and waste and fraud was wrung out too!).

If little wealth exists, but it is a large country with somewhat stable basic resources (water, energy, productive land for agriculture) , I'd print new currency and also sell bonds to finance cash flow deficiencies (until Weimar perspectives say you need to do something else; in the past some countries have seized the assets of powerless minority residents). But even in the non-Weimar scenario, where there are no existing counter-cyclical programs, this likely means that spiraling downward demand causes revenues to continue to fall.

In this situation we definitely should have the govt spend above historic levels in order to overcome the falling aggregate demand. What is needed is a new contributory source of revenues that steps forward, such as a foreign country's aid. That might be troubling, especially if you are proud that your nation exists.

But expecting that another country does not give them grants/aid, in order to finance the spending needed by wise economic policy you need to look at wealthy residents as a source. In a wealthy nation, taxes on the wealthy need to be increased to cover the financing gaps. It is a short-time, turn-around, maybe three years at most, when the economy will continue to grow at good rates, advanaging the wealthy economically more than the rest of the population (who does better too, not so much better on the economics side then).

You expect the wealthy residents to consider stepping forward, particularly those who are thoughtful. After all, at this time of 50% deficits, one can clearly say that the wealthy elite screwed up and did not foresee events, or worse, they used events and may have even engineered events for political-economy purposes. So in this happier scenario of a wealthy nation, these residents are then asked to contribute more to the financing of new govt spending programs designed to fill the gap in aggregate demand while also financing efforts to keep civil order (domestic and foreign), to regulate to ensure markets are orderly too, and to make other needed infrastructure investments.

Unless the wealthy were the cause of the economy tanking, and it was purposeful rent-sought results that went far-awry, you;d think the wealthy clearly would step forward to make such added contributions in order to stabilize their society and maintain it.

More likely the wealthy, who have means, will just move to another country, and let their first one (or second or third country of residence) go to pot. They don't call it political-economy for nothing. And taxation policy-people pay lots of attention to tax sources that are highly mobile.

[Nick Rowe]

Ralph: suppose desired stock of government debt (bonds+money) is 1,000% of GDP. Then next year it falls to 100% of GDP. Optimal policy? Increase taxes by 900% of GDP?

[K]

If rates are at zero and there is still no inflation then apparently you are still far from fiscal dominance. Maybe the market wants 2000% debt to gdp before the natural rate of interest rises to a reasonable level. Sounds to me like the market is offering you a huge free lunch. You should probably eat it.

[rsj]

If the National Debt is 1000% of GDP and both inflation and interest rates are low, then the private sector wants to hold this much debt. We can talk of structural issues -- e.g. why does the private sector want so much debt? Maybe there is enormous risk aversion, wealth inequality (only one person in the economy holds all the debt, and everyone else wants more savings) or other factors. Those other factors in principle could be addressed as part of a real structural reform program, for example to provide more social insurance or redistribute wealth. The structural reform program is not, however, running surpluses just because we don't like the preferences expressed by households.

As long as the private sector wants to hold that much debt, then who are you to say we should be cutting it? The way the private sector expresses a desire to reduce their stock of savings is to start selling those savings to each other and using the proceeds to buy consumption goods, driving up prices. Then the CB could react by raising rates. But if the CB *wants* more inflation, then it doesn't react by raising rates up until inflation is what we want it to be. As this hasn't happened, clearly the private sector doesn't want more debt.

FYI, in the U.S. households continue to deleverage and they continue to increase their savings. Personal Income went up and personal consumption expenditures fell, savings went up. Households want more savings. Interest rates on long term bonds are negative in real terms, yet there is no inflation. Core inflation is falling. Households have expressed their preferences, and policy should respond to these preferences rather than to numerology.

[rsj]

^^^In the above, I meant "clearly the private sector wants more debt".

Point being that there *are* trade-offs. But those tradeoffs are made by households *within* the economy. Your own personal trade-off is expressed by your desire to sell a portion of your wealth to someone else and use the proceeds to consume. The net effect of all of these trade-offs is a nominal interest rate and an inflation rate.

If we had a complete model of the economy, we could predict what those trade-offs would be and we would know what the "right' level of Debt/GDP is. Then we would be in a position to say that some counterfactual level is too high or too low, based on some convexity arguments. If we had a complete model of the economy, then we could focus on quantities only and ignore prices.

But as we do not have such a complete understanding of the economy, we follow along with everyone else and use prices as signals of scarcity or abundance, rather than ignoring prices and concluding that a given level of debt is "too high" in and of itself.

[Ralph Musgrave]

Nick, you ask “suppose desired stock of government debt (bonds+money) is 1,000% of GDP. Then next year it falls to 100% of GDP. Optimal policy? Increase taxes by 900% of GDP?” My quick answer is: “either that or a combination of much more tax and much higher interest rates.”

Moreover, I see no alternative: if the private sector has far more than its desired stock of paper assets, it will try to spend them away with inflation being the result. But of course it would be utterly bizarre if the average household’s desired stock of paper assets changed by more than very roughly 10% of its annual income in any one year (which is 1/100th of the above 1,000%). So in the real world the changes in tax doesn’t need to be anywhere near the above 900% figure.

[Nick Rowe]

Ralph: "My quick answer is: “either that or a combination of much more tax and much higher interest rates.”

Increasing taxes by 900% of GDP is going to be bit of a problem. I would go with the higher interest rates, myself.

[rsj]

You can tax wealth as well as income. Estate taxes and property taxes are an example, as are transfer taxes.

[Nick Rowe]

A 1% tax on bonds is like cutting the interest rate by 1%, if the supply is perfectly inelastic. It's interest rate policy by another name.

[rsj]

A 1% tax on bonds is like cutting the interest rate by 1%, if the supply is perfectly inelastic. It's interest rate policy by another name.

Let's not assume that the supply of bonds is perfectly inelastic. That is insane. It doesn't cost any real resources to either create or destroy a bond. A bond is a notional electronic entry, and we have lots of electricity and storage. How about we assume that the supply of bonds is perfectly elastic instead?

[Nick Edmonds]

"It's interest rate policy by another name."

Interestingly, though, it's the opposite interest rate policy to what you're suggesting (lower rates as opposed to higher rates).

In this rather extreme example, people are substantially trying to spend out of wealth rather than income, so you would have to tax wealth (or expenditure) rather than income.

[Max]

K, the problem is that what the market is offering can change on a dime.

BTW, in Krugman's liquidity trap paper from the '90s, he argued that Japan should *not* try to use fiscal policy because of high debt/GDP (it's a lot higher now!). I don't know if he has ever explicitly renounced that view.

[rsj]

Max,

The market offers what the central bank offers. Interest rates are set by policy. You can argue that the policy reacts to inflationary pressures, etc, but that means you are assuming that there will be inflationary pressures, in which case the private sector does not want to hold this much debt, and the inflationary pressures can be dealt with via an increase in taxes and a reduction in debt.

But in either case, households are the ones making convex trade offs and we need to listen to them, and use prices as a signal of scarcity rather than deciding that households are wrong simply because one number is bigger than what we like it to be. People seem to be OK with taylor rules and have learned to not look at ratios of monetary aggregates to GDP when deciding to raise rates -- If we strip out the patronizing morality of "debt = bad" then we'd get the same thing here.

[Max]

rsj, "the inflationary pressures can be dealt with via an increase in taxes and a reduction in debt."

Yes, but that may be expensive if the required adjustment (to quell a debt panic) is large.

[Nick Rowe]

Nick E: yep. But we need to distinguish between the rate of interest looking forward and the rate of interest looking back. Suppose people want to hold a large stock of government bonds from now until time T, and a small stock of bonds thereafter. A one-time proportional tax on government bonds at time T, if anticipated, means a low rate of interest from now till time T, and a higher rate of interest thereafter. Which is just what you need to eliminate the excess demand or supply of government bonds.

rsj: I was referring to government bonds. If we treat the stock of government bonds as an exogenous policy instrument, that the government can manipulate, that is akin to treating the supply as perfectly inelastic.

[Dan Kervick]

1. Would those numbers lead you to hesitate before recommending tax cuts or increased government spending to increase aggregate demand?

Yes.

If you answered "yes" to the first question, the punchline writes itself: we have already established that you are an austerian; we are just haggling over the price.

Not so fast. The hypothetical circumstances you describe are so extreme that it is hard to know what kinds of economic policies would make the most sense without knowing more background.

- Is the 50% deficit the primary deficit, or just the deficit? If the latter, then presumably the large size of the deficit is attributable largely to debt service?

- What kind of spending is the government doing? Is it running a large public sector with government consumption, investment and production etc. constituting a large share of GDP? Or is it just servicing debt and sending out transfer payments, so that the government contribution to GDP is close to zero?

- Who owns the debt?

Anyway, without further information, and in such extreme economic circumstances, I would recommend:

1. Direct cancellation of a substantial portion of the national debt. Given which segments of the population generally hold government bonds, debt-to-GDP levels of that size (at least assuming a tax structure such as the one we have now) represent a massive upward redistribution scheme and should simply be turned off. Also, the outgoing debt payments are probably all being recycled to purchase more bonds to finance the gargantuan deficit (unless the govt is just printing the money.) We're in the realm of Ponzi as a national way of life. So start by putting a bill through the legislature to wipe out the debt.

2. Direct supply side interventions in the economy: increased production quotas, increased hiring quotas, etc. Since every supplier in the economy is concurrently a customer who generally can't increase supply without purchasing more labor and factor inputs, these actions would yield concurrent supply and demand side results. This is a drastic measure, but the hypothetical indicates a rent-and-redistribution economy that has gotten lost in a thicket of financial transactions and interlocking ponzis and is neglecting real investment and (non-financial) work. So strong action is needed to command the economy back toward a state of healthier priorities.

3. Restructure government spending priorities. If the government is running large deficits but economic activity is still stagnant, they are clearly not spending the money very well. They probably need to put a much larger percentage of the outlay into consumption and investment, and less into transfers. They may even need to increase their share of GDP, which they can do once they cancel the debt. Hire people, build stuff, conduct research, develop technologies.

Bottom line: once an economy has reached such an advanced stage of decadence, sickness and Ponzi racketeering, strong government action is needed to set it back on the right course. We're beyond the realm of problems that can be solved via mere "stimulus".

[JF]

rsj above notes this: "As long as the private sector wants to hold that much debt, then who are you to say we should be cutting it? " The sequitur for him is to sell more debt to finance govt actions.

If the quoted sentence is true, it seems clear that the buying-private-sector is awash with funds and it has no idea what to do with them - on the margin for this group of people the money they have has little utility. Of course they'd rather be given a bond instead of a tax bill as they can brag about the increasing wealth number they have attained (as you continue to add in the new govt bonds), plus they know this gives them more of the future productivity of that country, which they see to be likely or they would not see these as being "safe" assets. That is a political and redistribution upward result noted by others on the board.

I would have thought that most of the comments here would have been about raising taxes on the wealthy (who are awash with money and want "safe assets"), at least in a country in which many live. (Can't tax existing wealth if it isn't owned much in a country; think about the US in its formative period, for instance, or any emerging economy).

Can people on this board discuss the raising of taxes on the wealthy in these scenarios?? Just to remind people the US is first with $80+ Trillion in net wealth, Japan second, Eurozone and the UK are up there too.

rsj is right of course for the understandable scenario to which most comment (the exaggerated one offered by Nick Rowe can't be imagined). If this hypothetical country can stay together civilly, you will clearly also need in a transitional sense to borrow too, and raise taxes on the wealthy, so the govt as an entity can replace basic demand that continues to fall.

[Peter K.]

"And suppose your economy hit the Zero Lower Bound, and suppose you thought that your own central bank's monetary policy could do no more to increase aggregate demand, but more aggregate demand was needed."

As Bernanke pointed out in his talk on deflation, the central bank can always do something even at the Zero Lower Bound.

http://www.federalreserve.gov/boarddocs/Speeches/2002/20021121/default.htm

"Like gold, U.S. dollars have value only to the extent that they are strictly limited in supply. But the U.S. government has a technology, called a printing press (or, today, its electronic equivalent), that allows it to produce as many U.S. dollars as it wishes at essentially no cost. By increasing the number of U.S. dollars in circulation, or even by credibly threatening to do so, the U.S. government can also reduce the value of a dollar in terms of goods and services, which is equivalent to raising the prices in dollars of those goods and services. We conclude that, under a paper-money system, a determined government can always generate higher spending and hence positive inflation."

I would have the central bank cooperate with the Treasury and do a money-financed tax cut or "helicopter drop." This wouldn't raise the debt anymore. I would also do money-financed government spending and investment. This wouldn't raise the debt and deficit anymore.

They would be done at regular intervals until aggregate demand was increased, the output gap closed and we started to see 4-5 percent inflation. Then I would balance the budget and bring down the debt via growth which would mean greater tax revenue and less spending on safety net programs like unemployment insurance and aid to state governments.

[baconbacon]

Going to make an ass of myself here, but isn't the Debt to GDP example backwards? Debt is fixed- not dependent on time, and GDP is measured withing a time period. So 100% debt to GDP over 1 year if measured monthly would be 1,200% debt to GDP (monthly GDP is 1/12 of annual GDP, total debt remains X trillion dollars).

[dan]

I think maybe not.
The curve the anti-austerians are on is the ratio of GDP/available resources.

ie. if resources are wasting and not being utilized that is the cost. And the way to reduce the cost of wasted resources is to start putting them to work...

So in your example there is not enough information to answer the question.
We need to know what available resources are, potential gdp etc relative to actual GDP. So I don't think the debt level and deficit level are irrelevant, but they aren't determinant. The first challenge is to get GDP humming at full potential, and then deal with the right deficit level and fiscal policy.

I think you also have to stipulate if this is a closed economy or of borrowing is happening externally and in different currencies.

[Nick Rowe]

bacon: Ooops! It was me that made an ass of myself. I shouldn't even try to do math. But if debt is 100% of monthly GDP, that's 8.33% of annual GDP. Which must be what I meant to say!

[Min]

Your conditions make me think that GDP is too low. If we increase GDP we will probably decrease those "scary" ratios. Speaking from a mathematical standpoint, it has always bugged me how the nearly all of the public discourse about those ratios ignores GDP. At first I thought that it was a cognitive problem of foreground and background. But now I think that propaganda has something to do with it, as well.

Also, the 10/1 debt/GDP ratio makes me suspect that too much gov't expenditure is going to creditors who are not keeping the money in circulation. To call raising their taxes, if that is indicated, "austerity" seems misguided.

[Nick Rowe]

Dan: some of your questions are a bit like asking: "It depends whether you are offering me $1 million Canadian or US for the night!"

"Bottom line: once an economy has reached such an advanced stage of decadence, sickness and Ponzi racketeering, strong government action is needed to set it back on the right course. We're beyond the realm of problems that can be solved via mere "stimulus"."

That sounds very similar to what an "austerian" would say. They would just have a criterion for what counts as an "advanced stage of decadence" that was more strict than yours.

Peter K: "As Bernanke pointed out in his talk on deflation, the central bank can always do something even at the Zero Lower Bound."

And I would strongly agree. In which case, we don't need to consider the trade-off between cutting the deficit and increasing aggregate demand.

[Dan Kervick]

@Nick

That sounds very similar to what an "austerian" would say.

Well, maybe. But I don't think austerians would argue in favor of cancelling the debt. It might be that the net result of my recommendations would be less overall government spending, but only because the government might need to stop spending so much money on debt service so it can spend more on other stuff.

[K]

Nick,

"the problem is that what the market is offering can change on a dime"

Arbitrage in economics is like a false statement in logic. If you assume it, you can derive literally anything.

If the odds are so high, then why doesn't anybody (everybody) just short government bonds?

[Dan Kervick]

The more I think about this question, the more it strikes me as too abstract unspecified for any answer. We aren't told anything about what kind of economy we are looking at and why it is under-performing. Without such information it is impossible to guess what the right kinds of policies would be.

[Niveditas98 .]

Didn't we agree a few weeks ago that if the interest rate is zero, there is no cost to additional government debt?

[Nick Rowe]

Nive: yes. But will it stay at 0% in future? Not if the fiscal policy works, and lifts the economy off the ZLB.

[Frank Restly]

Nick

"But I am looking for some sort of recognition that there's some sort of convex trade-off, or increasing marginal costs of debts and deficits."

Total interest paid in any annual period = Total Debt * ( 1 + Term Premium * Short Term Interest Rate ) ^ Average Duration / Average Duration

Let the term premium be some multiple (alpha) of the natural log of the average duration:

Term Premium (TP) = alpha * ln (Average Duration)

For any combination of alpha and short term interest rate, there is a sweet spot non-zero average duration that results in the lowest amount of annual interest being paid.

For instance let the short term interest rate = 1%, and alpha = 0.5.
The least amount of annual interest will be paid when the average duration of government debt is about 43 years.

A couple of other calculations:
ST INT% = 5%, Alpha = 0.5, Sweet spot duration = 12 years
ST INT% = 1%, Alpha = 0.75, Sweet spot duration = 31 years
ST INT% = 5%, Alpha = 0.75, Sweet sport duration = 9 years

In using the term "average duration", I am thinking that the government issues all the same duration of bond with 1 / average duration bonds coming due every year (accrued interest is paid from tax revenue, principle is rolled over). If government is using coupon securities, then interest payments on all debt come due every year and there is no benefit to extending duration.

[Frank Restly]

Ugh,

Just realized my mistake.

For any combination of alpha and short term interest rate, there is a sweet spot non-zero average duration that results in the lowest amount of annual interest AND principle being repaid. Interest only would be:
Total Debt * ( ( 1 + Term Premium * Short Term Interest Rate ) ^ Average Duration) - 1 ) / Average Duration

Is rollover a choice or is it coerced? If rollover is completely a choice, then the term premium should be stable with a stable average duration. If it is coerced, then the term premium can become unstable. The numbers above apply to the minimum annual interest and principle repaid every year as a percentage of total debt.

For instance, with a short term interest rate of 1% and an alpha of 0.5, the government would pay out about 2.85% of total debt in interest and principle repayments ever year. To get a sustainable debt to GDP ratio of 1000% with government collecting 15% of GDP in tax payments, it would need to reduce its interest and principle repayments to about 1.5% of total debt (15% / 1000%) - here all tax collections would go towards debt payments. Assuming a constant alpha of 0.5, the short term interest rate would need to be at about 0.6%.