Matt Yglesias, spilt milk, and the debt burden

I visit Matt Yglesias' house (HT JeffreyY). I drink one litre of milk from his fridge. I write Matt an IOU for one litre of milk.

1. If Matt subsequently tears up that IOU, then I am richer and he is poorer. Taking the two of us together, in aggregate we are neither richer nor poorer if Matt tears up the IOU. Tearing up the IOU doesn't bring the milk back.

2. Therefore there is no aggregate cost to me and Matt of me drinking more milk from Matt's fridge and writing Matt another IOU.

1 is of course true. But 2 does not follow from 1. 2 is false.

We can't do anything about the existing stock of national debt. "It's no use crying over spilt milk", even if it was spilled down my throat rather than spilled on the floor.

But that doesn't mean there is no cost to spilling more milk. True, there's also a benefit, if I'm thirsty, and it's spilled down my throat rather than on the floor. But there's a cost too. If the government runs a deficit now, there is a cost to future taxpayers. Sure, there's a benefit too, depending on what the government does with the funds it borrowed. And we should compare those benefits to those future costs. But we shouldn't pretend those costs don't exist.

The existing stock of debt represents the costs of past deficits that cannot be undone. We can't turn back the clock. But the fact that we cannot turn back the clock on past deficits says absolutely nothing about whether we should run a deficit now. We need to compare the benefits to the cost to future taxpapers.

Matt Yglesias was defending Paul Krugman's recent posts on the burden of the past debt and the desirability of current deficits.

Here's Paul:

"Deficit-worriers portray a future in which we’re impoverished by the need to pay back money we’ve been borrowing. They see America as being like a family that took out too large a mortgage, and will have a hard time making the monthly payments.

This is, however, a really bad analogy in at least two ways.

.....

Second — and this is the point almost nobody seems to get — an over-borrowed family owes money to someone else; U.S. debt is, to a large extent, money we owe to ourselves."

This post is based on Noah Smith's post, and on my original post.

Comments

[Lord]

Is there really a cost to repayment? If the alternative to repayment is an increase in interest rates? If we choose repayment to raising interest rates?

[Jon]

Karl Smith likened debt to promises--promises that serve to bind independent actors into acting in concert.

There is something wrong with owing lots of debt to ourselves: the problem was captured in your plastic vs. glass thread. Government as large single borrower (on the order GDP) is a systemic risk. The same people (Krugman, Yglesias) who want to limit the size of large banks do not seem to want to apply the same rule to the government.

Okay, they claim that the government cannot default (Karl Smith claims the government should just take what it needs by force); its not clear that under a democracy, the people will support arbitrary taxation in preference to default. It's also not clear that 'the people' will support monetization of the debt over default either.

So, in short, yes there are costs to increasing the amount of gross debt even when the net debt is zero--we owe it to ourselves means net debt is zero. Krugman wants us to believe only the net matters, but surely we learned better after the financial crisis that has just come to pass.

(You can then add to this critique: Noah Smith's recent post about intertemporal decisions and deadweight loss implications.)

[Greg Ransom]

The key to the conversation is choice over time between alterntive production and consumption paths -- using goods to produce long period production goods creating far,far greater output much later later in time vs using goods for immediate consumption now or for short term processes than produce fair, less output much sooner in time.

E.g. ripping down a house and chopping it up and burning it to produce warmth right now vs planting forests, producing factories to make logging equipment and nail guns and rail cars and moving equipment and everything else needed to produce housings capable of providing warmth for centuries.

If the government takes money away from savers who,would invest in forest production, rail car production and housing production, and diverts that money via borrowing spent of tripping down houses and burning the chopped pieces of those houses for fuel, what is the consequence???

In other words, without capital theory -- engaging the choice theoretical fact that no one will use a longer production processs unless it promises superior output to a shorter process & that in fact people make this choice constantly -- economists cannot competently engage this question of "borrowing from future generations".

The logic of debt only has a purchase on the real world when considered in relation to the choice in the present across time between production processes producing greater or lesser output across time in the future.

Debt which generates greater consumption now and far, far less production involving more productive long period produnction goods is debt which enriches the current generation and necessarily make later generations worse off than they would otherwise be.

[JeffreyY]

I think your milk analogy implies that Noah is right. Specifically, say that IOU gets inherited by Matt's kids. Your drinking the milk had an aggregate cost to you and Matt, but since the milk would long since have gone sour, it has no aggregate cost to your or Matt's kids. Similarly, using debt to rearrange consumption today has costs today, but since the consumable resources we're rearranging generally wouldn't exist anymore in 30 years anyway, it doesn't have any implications for our wealth in 30 years.

On the other hand, as Noah points out, shifting wealth between consumption and investment today _does_ have implications for our wealth in 30 years, but it doesn't matter whether you use taxes or borrowing to shift that wealth. Krugman tends to be imprecise when it suits his purposes, but I think what he means here is that the deficit isn't any more or less of a burden on future generations than taxing ourselves the same amount today would be.

[Mustapha]

But there's a cost too. If the government runs a deficit now, there is a cost to future taxpayers.

I don't understand why that should be the case. Let's say the government credits $1 million to the account of a construction company to build a bridge, without raising taxes or issuing debt to offset that spending. That's $1 million created out of thin air, which never has to be paid back and on which no interest is payable. The only potential cost of such deficit spending is inflation. Am I missing something?

[JKH]

Nick,

I’m a bit confused about something which is probably basic and obvious to most people within this debate.

In your original post, you constructed a sequence whereby net apple consumption by the oldest generation was “funded” by net apple provisioning by the youngest generation, that being done through an apple debt sequence that ended up with apple taxation paying down the debt.

You indicated there that this sort of construction should be a very obvious thing to economists.

What isn’t clear to me now is how important that construction is to your entire argument in this series of posts. It seems to me that the essence of that apple sequence is that taxation of a future generation to pay down debt incurred by the current generation constitutes a burden for that future generation.

So what isn’t clear to me is the degree to which that equation (future tax = future burden) is accepted as obvious by everybody.

It seems to me that once you assume taxes, the burden is established, according to your sequence. So there should be no debate about that. It seems to me further that the real debate is about the necessity for such taxes, but that the equation itself (future tax = future burden) is assumed, I guess, to be obvious to everybody. Is this the case? You did take the time to set out the sequence in your example, which is why I’m wondering.

And I guess I interpret Krugman’s post as the “we owe the debt to ourselves” argument includes the assumption that debt is not paid down by taxes, for whatever reason. And I would say that “we owe the debt to ourselves” is an entirely valid point, provided you assume no future taxes to pay down the debt, for whatever reason. Conversely, if you assume future taxes, there must be a burden.

Do these questions make any sense to you?

I ask partly because, for my own satisfaction, I’ve developed a fairly formal accounting model that corresponds to your apples debt sequence, and replicates that logic in balance sheet accounting terms. I was thinking of posting it here, but I don’t want to do that unless you’re interested. And I don’t want to do it if the “future taxes = future burden” is a completely obviously fact to everybody, and if the discussion is and always has been well beyond that. I need not post something that’s obvious, and a distraction that will just clutter up the debate with a peripheral concern. Maybe I can’t see the forest for the trees here. But if you happen to be interested in some formal accounting for your apples sequence, I can show it to you (a bit lengthy, but not outrageously so), including asset, liability, and equity balance sheet components. If not, no problem, since I’ve done the accounting for my own purpose anyway.

[weary_and_worn]

Back in the real world Karl Denninger's view http://market-ticker.org/akcs-www?singlepost=2830228

[JKH]

My reaction to the Yglesias post:

Debt obviously doesn’t “drain real resources”.

But taxation is a drain on private sector monetary resources that can impact real resources compared to the counterfactual. So if for any reason, the government responds to its debt build up with taxation (to retire debt), that can have a relative effect on real resources. Taxation is contractionary at the margin.

Yglesias seems to contend that a balanced budget provides tax revenue to pay down debt. I don’t know what that means unless it means what it appears to mean, in which case it is wrong. A balanced budget provides for interest payments but not debt repayment. You need a surplus for that.

The repayment of debt with taxes transfers accounting equity from the private sector to the government. That is certainly a monetary resource drain and can be a real resource drain.

The effect of defaulting on debt depends on whether on its associated with anything else. If the government defaults but doesn’t tax, the net effect is exactly the same as if the government paid off the debt with taxes (ignoring distribution effects in either case).

I see the real resource question as being primarily unrelated to whether or not the debt is held domestically.

[Ken]

Both Krugman and Rowe make sense to me on this (although Krugman's attitude about transferring wealth, as often is the case, makes me cringe):

1. Taking milk from people does not reduce our aggregate capacity to produce milk (although it might reduce the incentives to do so). [Krugman]

2. Transferring milk from one person to another within a generation has no effect on aggregate utility. [Krugman]

3. Paying for milk with an IOU and passing that IOU down to a later generation will require some future people to drink less milk than they would have otherwise. [Rowe]

In aggregate, future generations will drink less milk than they might expect, because the expected value of the debt owned by savers can only be honored by increased taxation. This won't decrease future productive capacity, but it will make realized wealth different than expected wealth. [Hirsh]

[J.V. Dubois]

"But taxation is a drain on private sector monetary resources that can impact real resources compared to the counterfactual". I do not think so. Ignoring the disincentive/distortion effects, taxation in order to pay down debt cannot be drain on private sector monetary resources, because these proceeds will be paid back to private sector in form of bond repayment. So the net monetary position of private sector is actually zero.

If you would construct a model of representative agents with the same level of income and bonds in portfolio, then this operation would basically be equal to bond default. It does not matter if government increases taxes to use them for debt repayment or they will just order everyone to destroy the debt. Such a policy would have high impact on that particular generation, but some future generation (that was not overlapping during the time of debt payment) would be better off. They would start with clean debt sheet - but more importantly, they would not be forced to give their product to older generation in exchange of the bonds. They could even start this Ponzi scheme anew being again the first generation benefiting from it.

[Nick Rowe]

Lord: "Is there really a cost to repayment? If the alternative to repayment is an increase in interest rates?"

We can't keep on raising interest rates as opposed to raising taxes. Because the debt simply grows faster and faster and eventually the next generation simply won't be able to afford to buy the bonds.

Jon: "Government as large single borrower (on the order GDP) is a systemic risk."

Good point, given uncertainty of taxing capacity. But my points stand even if you ignore your point (as I've been doing).

Greg: "Debt which generates greater consumption now and far, far less production involving more productive long period produnction goods is debt which enriches the current generation and necessarily make later generations worse off than they would otherwise be."

Greg, I'm surprised at you. You are putting a Hayekian gloss on Abba Lerner's analysis! As I showed in my original post, even in a world with no investment or disinvestment of any kind, and so no possibility for intertemporal transformation of consumption, it is perfectly possible for the debt to impose a reduction in the consumption of future generations.

It's much simpler. If there's an increase in the debt, there is a drop in the consumption of some (unknown) future generation of taxpayers. Any drop in the future output due to lower current investment is over and above that burden.

JeffreyY: "Specifically, say that IOU gets inherited by Matt's kids."

If you really do mean "inherited", as a gift, then you are assuming Ricardian Equivalence. If instead Matt *sells* his IOU to his kids, and if the government says that my kids have to pay my IOU, then the next generation (Matt's and my kids) will have to pay for the milk I drank.

Mustapha: "The only potential cost of such deficit spending is inflation. Am I missing something?"

Yes and no. If we are currently at "full employment" that money-financed inflation is undesired. (Plus, it amounts to a partial default on the nominal debt, plus, if anticipated, it requires nominal interest rates to rise one-for-one with expected inflation). And if the economy is at less than "full employment", and disinflation is the risk, that same money could have been used to buy back interest-bearing bonds and reduced the tax burden on future generations.)

JKH: "I’m a bit confused about something which is probably basic and obvious to most people within this debate."

You are totally wrong there ;-) Nothing seems to be basic and obvious to most people within this debate!

"Do these questions make any sense to you?"

Yes, absolutely. You are making perfect sense to me. And you are finding it hard to make sense of the opposing view. And you are damned right to find it hard to make sense of the opposing view.

Here is how someone who believes that opposing view would reply to you: "JKH, you are totally wrong because you are forgetting that those future taxes are simply a transfer to future generations of bondholders, who inherited the bonds along with the future tax liability!"

(Notice the weasel word "inherited"? They paid for those bonds, they didn't inherit them as a gift, unless you believe in Ricardian Equivalence.)

Can you post your accounting analysis as a comment (maybe on the 1,2,3,4 post)? I have no idea if it will make sense to me. But it might make sense to some readers. As I have said before, just for once, this really is (mostly) an accounting question.

[Greg Ransom]

In my judgment, in the real world, the drop in future output due to the shorting of the structure of investment vastly outweighs any other effect.

And note well, it is the time structure of production processes which matters, exactly equal investment dollar aggregates with completely different time structures of production would produce vastly different different output across time. Indeed, there need be no dollar drop off in total aggregate investment, the government might simply sift all of that total to low output short period production processes, per the demands of the rent seekers and bureaucrats controlling the payout system.

Nick writes,

"If there's an increase in the debt, there is a drop in the consumption of some (unknown) future generation of taxpayers. Any drop in the future output due to lower current investment is over and above that burden."

[Nick Rowe]

Greg: OK. I think that puts you in #2, along with Buchanan, except that you believe the indirect costs are empirically much more important than the direct costs Buchanan was talking about.

[Greg Ransom]

The key point, Nick, is that those using a simple aggregate conception of "K" have no way to imagine an intertemporal transfer of consumption via the lengthening or shortening of production processes, that is, changes in the time structure of production processes.

Keynes certainly completely failed to grasp this matter, as he reminds us repeatedly in his _General Theory_. And we see a repeat of this conceptual blindness to a causal key to expanding or contracting output in the writings of innumerable other economists, sort of like the mental blindness of folks who can't identify individual faces -- that part of the brain for perceiving reality just doesn't exist.

[Nick Rowe]

Greg: I think I sort of see it. Take a horribly crude and oversimplified example. You lay down a stock of raw whisky to age, with either a 5 year or a 10 year tamper-proof time-release lock on the barrel. "K" looks exactly the same in both cases, since the "historic cost" (yes, those are "scare quotes" around "historic costs" because we both understand that point) of the investment are exactly the same in both cases. But the time-structure of production is very different in the two cases. And it matters for planned future consumption.

[Adam P]

Nick,

I have to tell you, the whole time of been reading WCI I don't think you've had a post that was more indisputably correct as this series (starting with the "debt is too a burden" post). I mean it's a counter example for crying out loud, not a theorem!

As counter examnples go this is as good as it gets, perfectly standard OLG model, no tricky assumptions that secretly drive the results. How do you argue with that?

Yet argue they do...

[Nick Rowe]

Thanks Adam! And, given you are a tough critic to please, that means a lot.

But yes, all I need is one counter-example. And mine is about as simple as you can get in an OLG model.

I think I'm slowly winning.

[JKH]

Nick,

As requested, I’ve posted my accounting translation of your apple model here:

http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/01/the-30-years-debt-burden-non-war.html?cid=6a00d83451688169e20162fef52c26970d#comment-6a00d83451688169e20162fef52c26970d

Sorry, I was delayed a bit at the outset. And then I decided to add an addendum, showing the treatment of foreign bond purchases as well, for completeness.

Unfortunately, that roughly doubled the length.

And I probably didn’t proof it all perfectly.

But it’s accounting, so I know you’ll find it very exciting.

[Mustapha]

And if the economy is at less than "full employment", and disinflation is the risk, that same money could have been used to buy back interest-bearing bonds and reduced the tax burden on future generations.

By "tax burden," I presume you mean the new taxes that will have to be levied at full employment to contain inflation? That's the only cost to future generations I can see. But to call it a "burden" seems misleading. The new taxes would be taxes on money the government created by the push of a button at the Receiver General. I'm not sure where the burden to the taxpayer lies in getting $1 million of created money from the government and then having some of that $1 million taxed a couple of years later when the taxpayer's income is higher only thanks to that initial money creation. If anything, the burden would be having to do without the created money in the first place and making do with a lower income throughout the period.

In any case, it's possible to imagine ways of dealing with buyer's inflation besides raising taxes or borrowing from the public, if these things are seen as burdensome. Every month, for example, the government could declare the interest rate that banks have to pay on demand deposits -- in effect, a price control that the government would tweak by trial and error. This would probably require a zero interest rate policy because of the excess reserves that would pile up in banks. It would also require regulations to prevent banks from canceling out the higher interest payments with higher bank fees.

[Min]

"The existing stock of debt represents the costs of past deficits that cannot be undone. We can't turn back the clock. But the fact that we cannot turn back the clock on past deficits says absolutely nothing about whether we should run a deficit now."

Indeed. But in the current political debate, the politicos who are crying that the debt is a burden on our children are doing so as an argument against running a deficit. The implication is that a deficit would be bad because it would add to the current debt, thereby creating an intolerable burden on our progeny. They are not weighing costs vs. benefits. They are simply crying, "Debt bad! More debt worse!"

[Bob Murphy]

Nick I can't keep my head straight on this stuff. When I read Yglesias' post, I slipped back into thinking Krugman was basically right. And yet, I know you made the point in your apple example.

Can you tweak the apple example, so there is just an A and a B cohort? So in period 1, government induces B to give some of their apples to cohort A by offering an interest rate, then A dies. In period 2, there's just cohort B, and government taxes them collectively to pay off the bonds.

Does your story work there too? Because I think if so, that might be a better way to take Krugman on, head-on. Before I thought he was right, given that there are only two generations, and that your more realistic 3-generation model showed the problem with that simplification.

But since I think we can get your result with just two cohorts, now I'm thinking the # of generations isn't the issue.

[JeffreyY]

Nick: I did mean "inherited", but I agree with you that lots of assets are traded to younger people in exchange for consumption goods during retirement. I don't generally want to defend Ricardian equivalence, so let me re-think this...

Is there anything special about government debt in your set-up? Doesn't every single loan, private or public, create a tradable asset that, when sold to a younger generation, induces a transfer to the older generation? Maybe not, because it also creates a tradable obligation which, if "sold" to a younger generation induces a transfer in the other direction. But obligations aren't generally traded, and you can escape them by dying, at which point whoever owns the asset takes the loss. Is the difference in the term of the debt? (i.e. that government debt is likely to live longer than individual people, while loans usually get repaid?) Or maybe in that the government usually won't default, so when the younger generation buys its debt, they don't write it down for default risk?

[K]

Bob,

I've kind of slipped towards agreeing with Krugman, though I have no idea whether for the same reasons as him. To me, A pulled a fast one on B, but that in no way means that they've pulled a fast one on C and all subsequent generations. And since B was around for the original "deal" it's arguably fair. Whether B wants to try the same trick on C, or just pay off the debt that they got stuffed with is up to them. But it's a real stretch to argue that A stuffed C. If there's a "problem" it's in the way that political power is distributed between (currently alive) generations since votes aren't distributed according to productive output and older generations are therefore economically over-represented. But there's nothing preventing C from ending the charade and just defaulting on liabilities that they were never a party to creating in the first place. After all, all they lose (if anything) is the ability to play the same tricks on future generations, an ability which arguably they've already lost anyways.

[Nick Rowe]

Bob: Yep, I know the feeling. If you've spent decades seeing the thing as a rabbit, it's hard to see it as a duck, and you keep slipping back to the other way. I've been there.

"Can you tweak the apple example, so there is just an A and a B cohort? So in period 1, government induces B to give some of their apples to cohort A by offering an interest rate, then A dies. In period 2, there's just cohort B, and government taxes them collectively to pay off the bonds."

Yep, that works fine. I only used 3 cohorts for completeness, to show that cohorts in between the transfer recipients and the taxpayers are neither richer nor poorer, and just have a delayed time-profile of consumption. If you have 3 cohorts you can extend it to any number. Or cut it down to 2.

[Nick Rowe]

Min: "But in the current political debate, the politicos who are crying that the debt is a burden on our children are doing so as an argument against running a deficit. The implication is that a deficit would be bad because it would add to the current debt, thereby creating an intolerable burden on our progeny. They are not weighing costs vs. benefits. They are simply crying, "Debt bad! More debt worse!""

Unfortunately, those politicos are right (except for the fact that they are forgetting about the benefits of whatever it is the deficit is spent on, plus, if you are a Keynesian, those benefits include escaping the recession). The existing debt is a burden on whoever is going to pay the taxes on it, and there probably are increasing marginal costs of that burden. So the bigger the existing burden, the bigger the marginal cost of an additional deficit. Yes, we already own the bonds reflecting the existing debt, but some of us paid for those bonds by cutting our consumption when young so we could consume more when old.

[Nick Rowe]

JeffreyY: If I borrow money the IOU has my name on it. If the government borrows money, the name on the IOU is left blank. It will be filled in by some later government which decides to increase taxes. Alternatively, each generation can issue 25 year bonds with the next generation's name on the IOU. And that next generation can issue more bonds, to repay those bonds, with the third generation's name on them.

[Bob Murphy]

Man this is just blowing my mind Nick, to think of your apple example with just two cohorts. It really *is* like a time machine, isn't it?

In fact, it's so much "as if" there's a time machine, that I keep wondering if maybe you're doing something wrong--since we know there's not a time machine.

Anyway, this whole thing is great. I haven't been this frazzled since I spent 8 hours in grad school thinking I had come up with a counterexample to Arrow's Impossibility Theorem.

[Adam]

"If the government runs a deficit now, there is a cost to future taxpayers."

Really? Why? I mean, I guess that's true in nominal terms, but it is it true in any real way? Government don't repay their debt, and to the extent that government debt decreases relative to GDP it's usually because of growing GDP, not shrinking government debt. It's not true that a deficit today mean higher tax rates in the future (even if higher tax receipts are needed). So is there any actual cost to future taxpayer without some sort of assumption about what happens to interest rates or what future fiscal policy will be?

[Audrey]

Nick is making a ridiculous argument.

As Noah says, it is absurd to think about debt burden without looking at what the debt pays for and who benefits.

The debt is a side issue. In a closed system, debt makes no difference to aggregate welfare.

The same applies intergenerationally. If the receipts of the debt are used to make a park, that park benefits future generations who don't have to pay for it. If they're used to "spill milk" today (how is THAT a useful real-world example?), then of course future generations have a burden imposed.

Nick's entire argument is ridiculous because it's absurd to think about this issue (indeed, the issue has no meaning) without considering how the debt is used.

[JKH]

Nick,

It’s clear that taxes are the thing that matters here.

At the micro level, if a bond maturity is paid with taxes, it is a burden.

And it becomes a net burden to any future generation that happens to hold that bond as the result of purchasing it (as opposed to inheriting it). Conversely, there’s obviously no future generation burden if the current generation both buys and matures the current bond. The latter case would correspond to a single stage apple model.

At the macro level, the same holds if the government pays down any part of gross debt with taxes – i.e. any time it runs a surplus, pretty much.

But arguably, it’s not the case that there will be a macro level burden so long as the government is growing the debt outstanding.

Sound right as a summary?

[Nick Rowe]

Bob: here's another example, without a time-machine in it. Instead it has a machine that converts rabbits into apples.

There are three groups of people, in an endowment economy:

A are herbivores who only eat apples, and who have an endowment of 1,000 apples.

B are omnivores who have an endowment of 500 apples and 500 rabbits.

C are carnivores who have an endowment of 1,000 rabbits.

Question: can the government tax C to give A more apples?

Wrong answer: of course not, since C neither own nor consume apples.

Right answer: of course it can. C sells rabbits to B in exchange for apples, which it gives to the government to pay the tax, and the government gives the apples as a transfer to A.

A eats more apples. C eats fewer rabbits. B eats fewer apples and more rabbits.

[JKH]

above, clarifying:

"And it becomes a net burden to any future generation that happens to hold that bond at maturity as the result of purchasing it (as opposed to inheriting it), and where the bond is paid off with taxes."

[Nick Rowe]

Remember that trade theory example of how to convert wheat into cars? Put the wheat on a ship, send it out into the Pacific, and it magically returns full of cars. Same thing.

I know the feeling Bob. 30 years ago I had to get a big pad of paper, write "1 apple" on some sheets, and "IOU 1 apple" on other sheets, and lay them all out on a big table.

JKH has just got his head around it by doing the accounting very thoroughly. (See my previous post). We all have our different ways of getting to see it.

[Jacques René Giguère]

Nick:"Right answer: of course it can. C sells rabbits to B in exchange for apples, which it gives to the government to pay the tax, and the government gives the apples as a transfer to A."
You just transformed apples into currency.

[Nick Rowe]

Audrey: "If they're used to "spill milk" today (how is THAT a useful real-world example?), then of course future generations have a burden imposed."

Thank you. You have just agreed with me. Many economists will say that what you just said ("....of course future generations have a burden imposed.") is impossible.

And BTW, with nearly all government spending, there are 2 questions: who gets the benefits?; and who pays the cost? This argument is about the second question. I am disagreeing with many economists who say the future generation can *never* bear the cost.

[Adam]

"If there's an increase in the debt, there is a drop in the consumption of some (unknown) future generation of taxpayers."

Relative to what? To what that future consumption would be without the increase is debt? Aren't there a lot of other potentially confounding factors that need to be accounted for? And not to get all Keynesian, but what if that newly accumulated debt is the difference between an economy that grow between now and that unknown future point in time and one that stagnates?

[Nick Rowe]

Adam: fair point. Relative to what it would have been with the same GDP. I would put any Keynesian effects on future GDP on the benefit side of the ledger. But if you want to subtract them from the cost side, the answer should be the same.

[Audrey]

Nick: "I am disagreeing with many economists who say the future generation can *never* bear the cost."

Ok, fair enough. But point me to an economist who has said this WHILE dismissing the benefit issue as irrelevant?

Do you think there are economists out there arguing to borrow a bunch of money to spill milk?

Because this is certainly the way you're coming across in your discussion at large.

[Mustapha]

I'm still confused.

If the Receiver General credits a private bank account in the amount of $1 million, and if there is no accompanying increase in taxes or issuance of new debt, then the Receiver General has just created $1 million out of nowhere. It has not even borrowed the money from the central bank. In a time of less-than-full employment, the only consequences of such spending are some numbers on a computer screen at the Receiver General and the real output purchased with the money. The Receiver General does not have to get the money from anywhere; it literally creates it by keystroke.

Where is the cost to future taxpayers? The only cost I can see is future inflation (again, we're at less-than-full employment here). To be clear, there is no IOU anywhere in the picture, not even a blank one.

[Mustapha]

In the example above, I'm starting from a balanced budget position.

[JKH]

“If the Receiver General credits a private bank account in the amount of $1 million...”

Obviously an MMT motivated argument - not that there’s anything wrong with that.

(“Keystrokes” gives it away).

That transaction represents government deficit spending.

Whether the deficit manifests itself as Treasury debt, or a Treasury overdraft at the central bank, or whatever, is immaterial to Nick’s argument.

The issue is not the accumulation of debt or the accumulation of deficits manifested in whatever financial form.

The issue is whether the accumulated deficit is repaid even in part with taxes at some point.

There is nothing in MMT that prevents the possibility that some part of the accumulated government deficit will be repaid in taxes at some point. MMT fully acknowledges the long term risk that fiscal tightening may be required at some point – such point depending on the utilization of real resources and the emergence of inflation risk.

At that point, if the government runs a surplus in order to shrink the accumulated deficit, and if the generation that pays the taxes is different than the generation that originally benefited from the earlier deficit spending, the generation that pays the taxes will bear a net burden, analogous to Nick’s apple example.

[K]

JKH: "The issue is whether the accumulated deficit is repaid even in part with taxes at some point."

I don't agree. If it's repaid and then shortly afterwards reborrowed, how does that change anything? The issue is whether it is *necessarily* repaid: voluntary, unnecessary repayments don't count. That's why the interest rate vs NGDP growth is so important.

Nick:

On the topic of *necessary* repayment, I made a comment about the relative rates of interest vs NGDP growth on the other post (just before JKH's monster comment) (just in case you didn't notice).

Also, in the other post I made the argument (which I repeated above for Bob Murphy's sake, that each generation is free to choose to accept or reject the liability which it has no moral obligation to bear by virtue of it not having been a party to the original agreement, and that no one can therefore burden future generations. What do you think?

[Paul Rogers]

Dean Baker brings some sanity to this debate:

http://www.guardian.co.uk/commentisfree/cifamerica/2012/jan/03/climate-change-real-bequest

[Nick Rowe]

K: "Also, in the other post I made the argument (which I repeated above for Bob Murphy's sake, that each generation is free to choose to accept or reject the liability which it has no moral obligation to bear by virtue of it not having been a party to the original agreement, and that no one can therefore burden future generations. What do you think?"

Each cohort burdens itself (or future cohorts) by being willing to accept the bonds sold to it by the previous cohort. But it would require a *collective* decision by the cohort to do that. Each individual member of the cohort would still be rational to buy the bonds. When cohorts are continuous (rather than the discrete cohorts of my model) it would be very hard to draw the line and say "we will not pay your debt". Who's "we", and which debt is "yours"?

Even when there's a regime change, the new regime will often honour the debt of the old regime.

[Mustapha]

JKH: "At that point, if the government runs a surplus in order to shrink the accumulated deficit, and if the generation that pays the taxes is different than the generation that originally benefited from the earlier deficit spending, the generation that pays the taxes will bear a net burden, analogous to Nick’s apple example."

Ah, thanks for the clarification. It looks like I was talking past Nick. That's the sense in which I understood the cost to future generations, as well. When I said "the cost of future inflation," I included in it the cost of fiscal retrenchment to contain inflation (or its alternative, the cost of price increases). I should have made this clear.

What I still don't understand, though, is why deficit reduction constitutes a "burden" on the future cohort if the deficit has resulted in a higher GDP for them. That possibility seems crucial to me, because if the future cohort enjoys higher incomes, they shouldn't have to sacrifice much at the margins to pay the higher taxes.

[JKH]

Mustapha,

“What I still don't understand, though, is why deficit reduction constitutes a "burden" on the future cohort if the deficit has resulted in a higher GDP for them. That possibility seems crucial to me, because if the future cohort enjoys higher incomes, they shouldn't have to sacrifice much at the margins to pay the higher taxes.”

The situation is that of comparing the effect of a tax to pay down debt, versus the counterfactual of no tax and the continuation of that debt level. That’s all. The rest of the economy should be held ceteris paribus in this analysis - however it has performed up this point.

If you think of this from an MMT perspective, taxation reduces net financial assets held by non-government. That’s contractionary. It’s a reduction in non-government income and wealth, again compared to the counterfactual of no tax. So in that sense, there is clearly a monetary burden associated with the tax. Again, from an MMT perspective, the purpose of the tax isn’t to put the government on a “solvency” footing. It’s to impose tighter fiscal policy in the face of real economic constraints and inflation pressure. And to the degree that the tax holds back spending power, it has an associated effect on the real economy. In Nick’s model, this is manifested in the taxpayer forgoing the apples that he would have enjoyed in the counterfactual. And that is a real burden.

[Bob Murphy]

Nick wrote:

JKH has just got his head around it by doing the accounting very thoroughly. (See my previous post).

Sorry, but which post? Who's JKH?

[JKH]

One of the interesting things of working through this stuff is that Nick’s model would first appear on the surface to be not only counterintuitive to the MMT view of inter-generational government finance, but something that MMT might reject outright.

“No grandchildren involved” is a favourite MMT refrain on this subject.

But in fact I think it’s quite reconcilable, as per my last comment. I think the MMT approach generally assumes continuous government deficits – largely of the basis of rejecting any type of insolvency argument, while promoting the feasibility of deficit spending right up to the point of full employment. And in that context, the “we owe it to ourselves” logic works very well. It works so long as you are rolling over debt and/or expanding it. It’s only at the point of taxation for the purpose of repaying debt that Nick’s model is activated at an operational level.

And the debate about deficit management trajectory etc. is entirely separate from the issue of what actually happens when you pay down debt with taxes, as per Nick’s model.

The interesting corollary to all this as well as that government surpluses are a burden by this definition, assuming they are used to pay down debt, since they arise from net taxes.

And that sort of thinking about surpluses is in fact very much in synch with the way MMT thinks about them in a more colloquial way.

[JKH]

"Sorry, but which post?"

http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/01/the-30-years-debt-burden-non-war.html?cid=6a00d83451688169e20162fef52c26970d#comment-6a00d83451688169e20162fef52c26970d

[JKH]

"Who's JKH?"

unsub

[K]

Nick: "But it would require a *collective* decision by the cohort to do that."

The collective action is for example via debt repudiation by the government. Or cancellation of the pensions/health care of A or, if B was OK with the scam, the pensions/health care of B. There's always recourse. If the old man was a drunkard, stick it to him!

[Greg Ransom]

What's missing here is a strong sense of the superior output aspect, although it exists (in the realm of those with differential tastes in whiskey).

Let's consider a different example -- and the output of shade. Your production good is a small poplar tree (grows 8 ft plus per year) on the south side of the house which requires no maintenance. You can cut the tree down and use the limbs to construct temporary shading over your windows -- until the leaved soon dry and fall off.

Or you can wait for the tree to grow and eventually provide shade for the whole house year after year after year.

There is a choice set before the tree owner between a small output now, or a far greater output later.

A Keynesian might say, lets boost output now by using idle labor and idle small trees, in conjunction with the the diversion of other resources to help with the tree cutting, etc. If we borrow from Jack to pay for the additional diverted resources taken away from other uses, we can simply pay Jack back later, an in total we we all be in the same state of total wealth, simply transferring resources temporarily from Jack to us, and then later back the other way. And we'd be ahead the short term shade provided temporarily by the leaves on the cut limbs.

But over the long term, in fact, we out out the massive shade provided by the grown poplar tree, producing output continuously for decades.

Nick writes,

"Greg: I think I sort of see it. Take a horribly crude and oversimplified example. You lay down a stock of raw whisky to age, with either a 5 year or a 10 year tamper-proof time-release lock on the barrel. "K" looks exactly the same in both cases, since the "historic cost" (yes, those are "scare quotes" around "historic costs" because we both understand that point) of the investment are exactly the same in both cases. But the time-structure of production is very different in the two cases. And it matters for planned future consumption."

[Greg Ransom]

Consider another example. You have 3 crappy iron hammers and everything else you need to forge and shape knives.

Where are to scenarios, assuming the same beginning production goods, inputs, and the same labor time:

(1) You use the crappy hammers to quickly turn out a set of six dull and crappy knives that don't last more a week or so -- by the time the 6 knives are made all 3 of your hammers have cracked and detonated.

(2) You use the 3 crappy hammers to build a very strong hammer that won't crack -- a process with takes greater time heating the coals, waiting for cooling, etc. (When not working, you use your labor time for other tasks, as you did also under the other scenario). You end up destroying the crappy hammers in the process, you then turn to using your remaining labor time assign to this iron working to carefully hammer out 5 very strong and very sharp knives that won't crack and last for decades.

In the first case, you have the output earlier, because your production process is shorter -- but your output is far inferior.

In the second case, the output is far superior, because your production process is longer -- although you have to sacrifice waiting till a later time to have your output.

Now, we can borrow from Peter & Paul to finance either process (1) or to finance process (2).

When we pay them back later after financing process (1) in the aggregate we are worse off -- in the long run we have few goods and inferior goods.

When we pay them back later after financing process (2) in the aggregate we are better off -- in the long run we have more goods and superior goods.

If incentives tend to channel private borrowing in a private allocation system toward production process (2), and incentives tend to channel government borrowing in a government allocation system toward production process (1), government borrowing now & government taxing later will make us worse off than the private alternative process / system.

[JKH]

Nick,

You probably saw it, but this week’s post by Fatas and Mihov seems to get it right:

http://fatasmihov.blogspot.com/2012/01/debt-does-not-matter-spending-and-taxes.html

[Ramanan]

Nice discussions.

I checked some SFC literature and language of national accountants and seems they have always measured burden as a flow, not stock.

So one sees burden used for households as the interest plus principal payments in any given period such as a quarter or a year. (i.e., repayment of principal paid in one period, not the whole principal) And similarly the tax burden.

In the international case, the burden is the total interest paid to foreigners in any given period.

Paul Krugman had a recent post on the US net investment income of the balance of international payments, which to everyone's surprise (actually a few who track it) has remained positive and that there is no debt peonage in his way of putting it. I disagree with him because he is talking of the potential future as well as present -- faster the attempt of the US to bring unemployment down (by either fiscal or monetary means), the quicker this will turn negative.

[JKH]

Ramanan,

I think the issue being discussed is, in effect, the net burden – including the case of using the definition of burden that you cite.

The key question relates to the interpretation of “we owe it to ourselves”.

To some, that implies zero net burden – which it does on the surface, including the case of using your definition.

But that fails to consider the relationship of the debt to future taxation.

So the analysis has to focus on whether “we owe it to ourselves” is the right or wrong interpretation of net burden. And it’s not necessarily the case, because it depends on the incidence of future taxation.

BTW, all of the discussion in this debate that relates to the effect of government investment in NPV positive projects, etc. is actually peripheral to the central question of taxation. The central question is the connection between debt and future taxation – regardless of how government NPV projects may slow the growth of debt, other things equal, for example.

The net burden is the tax.

Etc.

[JKH]

P.S.

The fact that the government owes the debt to the foreign sector is of no consequence to the question of whether or not there is a net burden for a future generation of the domestic sector (which IS the question).

The government comes to a point where it taxes the domestic sector to pay down the debt.

It swaps the proceeds of the tax with the foreign sector in exchange for extinguishing the debt at maturity.

The bond maturity transaction in and of itself has no effect on the domestic sector.

But the tax that is paid to fund the bond maturity transaction is a net burden for the generation that exists at the time when the debt is paid down – just as it is in the case where the maturing bond is owned by the domestic sector.

[Ramanan]

JKH,

Yes I figured the points from reading all the comments.

I guess both saying debt is a burden and debt is not a burden are misleading.

[K]

Here is how I see it (now). If B was insufficiently powerful (too young, whatever) to be properly represented when A enriched themselves at the governments expense, then B is under no *moral* obligation to recognize the incurred liabilities. The solution is to repudiate the bonds. This is not unlike how we traditionally treat stolen goods or proceeds of crime (in this case fraud). If you have bought stolen goods (the bonds) then you are likely to take a loss even if you were an unwitting participant. If B does not repudiate the bonds then they are likely to end up taking the loss when C does repudiate them. The moral of the story is that if you are the victim of a scam, the solution is not to perpetrate the same crime against some other helpless victim. The solution is to blow the whistle and try to reclaim your losses. Since nobody is *forced* to victimize the next generation there is no way that A can burden future generations other than B who has an obligation to take recourse as soon as they are capable.

[Ramanan]

JKH,

The fact that the government owes the debt to the foreign sector is important in my view. More importantly, it is the net indebtedness of a nation but in many cases, the public sector picks up an important piece of this.

For to service the net debt to foreigners, interest has to be paid and as far as interest on government securities held by foreigners is concerned, taxing and paying it off.

(I understand an MMT ultra-quibble which may enter in all this - saying taxes are destroyed and a narrative along those lines :) - that is an NB for someone who may want to interrupt)

Just one example: A poor nation faces high unemployment: The government can relax fiscal policy to counter this but because of this relaxation, and the increase in income due to this the nation may suffer a widening of the trade deficit and as a result a higher net indebtedness to foreigners. Since the government of this poor nation cannot keep doing this, it has to adjust demand sometime in the future - such as by higher taxation. So in this sense, there is a burden entering the game.

[JKH]

Good points, Ramanan.

But a few moving parts that should be separated out in my view.

Expansive fiscal policy can affect the current account deficit, yes.

But the current account deficit doesn’t necessarily NEED to be funded with government debt.

So the fact that government debt may end up being held by foreigners is separate issue from the current account deficit per se.

And we’re back to the question of whether or when the government will need to tax in order to pay down its debt.

And that is the issue here.

Not that fiscal policy may affect the current account deficit, although that is a separate and very important issue that you have hammered home on the MMT blogs, to your credit - (although not necessarily from the MMT guys :) )

[JKH]

K,

That does illustrate the potential burden on B.

I suppose B could organize a boycott or buyers’ strike on the bonds - A can’t even square its own financial position unless B agrees to buy its bonds.

That would force A either to will the bonds to B, or tear them up out of spite.

As far as organizing is concerned, B has a lifetime to put pressure on the government to stop spending so much, as opposed to being forced to raise taxes to the point of running surpluses to pay down debt.

[Greg Ransom]

The net burden is the massive lost output.

Opportunity costs rule.

"The net burden is the tax."

[Paul Rogers]

I agree with Audrey:

"Do you think there are economists out there arguing to borrow a bunch of money to spill milk?

Because this is certainly the way you're coming across in your discussion at large."

Until and unless Nick wants to make the argument that Krugman and whoever else he's trying to attack are arguing in favor of borrowing to spill milk, his whole argument stinks of a Mankiw-like disingenuity.

[Adam]

I guess my point is that yes, all else being equal, more debt now means less consumption in at some unknown point in the future. But in reality we can say with pretty strong certainty that all else will not be equal in the future. If you're going to buy Keynes at all, at minimum there will be multiplier effects from the deficit spending. But even if you aren't, what happens with the business cycle may very well not mean higher tax rates in the future (#4 in Nick's prior post). But even without that, the future will be influence by so many other fiscal, tax and monetary policy decisions that I just don't see it as at all credible that deficits now mean taxes later so there's no point in fiscal stimulus.

[Ramanan]

JKH,

Yes I understand very well that "But the current account deficit doesn’t necessarily NEED to be funded with government debt." and I have an example which isn't surprising - Australia.

It certainly makes the argument more involved - cases like this :)

It could as you probably pointed out, also be the case that foreigners hold a lot of government bonds independent of net indebtedness. I assume that would have been the case with foreigners holding a lot of US Treasuries well before the US turned a net debtor nation.

And of course, the reasoning does depend on behavioural assumptions - my point being that external debt has all the characteristic of intuitive debt under this assumption. The government has to keep taxes high to keep domestic demand low and in this way from the households' point of view, higher taxes are burden and from a nation's viewpoint, interest paid is the burden. In the special case when foreigners' holding as a proportion of net debt is high, one can give the intuitive explanation that taxpayers are taxed in part to pay interest to foreigners.

[Nick Rowe]

Paul: "Until and unless Nick wants to make the argument that Krugman and whoever else he's trying to attack are arguing in favor of borrowing to spill milk, his whole argument stinks of a Mankiw-like disingenuity."

Google "No use crying over spilt milk", "bygones are forever bygones", "sunk costs"; understand the important role of this idea in economics; then retract your accusation.

[A Facebook User]

We can certainly pass a real burden onto later generations by using up irreplaceable resources, such as fossil fuels. They will then have to find ways to do the same things with other resources, such as sustainable energy or even hydrocarbons from the solar systems gas giants. Either may or may not require them to spend more time and effort obtaining energy, which, if the latter is the case, is definitely a burden on them. They have to spend more time and effort getting energy and have less time to do other, more enjoyable, things.

But it is easy to show that the idea of passing a monetary debt on to future generations is nonsense. Assume that we can. Generation one passes it's debt to generation two. Well, what stops generation two from passing on it's bigger debt to generation three, and so on forever? Nothing! Sure the debt will keep increasing but it's only a nominal debt, not a real debt. And any generation can pass it to the next one forever. So since it never has to be paid it's not a burden and that contradicts the original assumption.

As for taxation to hold inflation in check, so long as there is enough purchasing power for humanity to purchase all it can produce, how is this a burden? You are just getting rid of money that is of no use to you because you can't buy anything new with it once everything has been bought. All you can do is bid up prices and that can be stopped by simply removing enough purchasing power by taxation so that you can't bid up the prices any more, but still you can buy everything you want.

[JKH]

Ramanan,

"In the special case when foreigners' holding as a proportion of net debt is high, one can give the intuitive explanation that taxpayers are taxed in part to pay interest to foreigners."

Maybe - “taxed in part” arguments get a little more complicated. However, unless the government is running a budget surplus (indicating “net tax” or “net burden”), the debt is expanding. And as long as the debt is expanding, the “we owe it to ourselves” argument rules, I think.

E.g. the “Ponzi condition” of the interest rate less than the growth rate means in theory the debt could expand forever. So suppose the debt does expand forever. And suppose it does turn out that the foreign sector buys all the bonds issued by the government. And suppose the current account deficit exactly matches the expansion of the debt. That means the foreign sector is funding the entire deficit, de facto. So the foreign sector in this case has literally taken over the “we owe it to ourselves” paradigm of deficit financing. And under the assumptions, there will be no need for taxes from the domestic sector to pay off the debt. So there is no net burden according to Nick’s model.

I think that example means that as the interest rate on the debt approaches the growth rate, the interest on the debt approaches accounting for the entire current account deficit - which would mean that the current account deficit approaches the size of the services deficit (interest) without any goods deficit. Not sure. But it works so long as foreigners don’t won’t to sell their bonds back to the domestics in exchange for goods.

[JKH]

AFU,

"So since it never has to be paid it's not a burden and that contradicts the original assumption."

It doesn't contradict the original assumption. It's a different assumption.

The original assumption (a scenario really) was that the debt would be paid down with taxes. The conclusion from that is that the debt is a net monetary burden (or real burden in Nick's apple model). And Nick's model (and my accounting model in the monetary case) shows how its a burden.

Assuming a scenario of indefinite debt expansion doesn't contradict any of that. It's just a different assumption. And its completely complementary in terms of the conclusion.

[JKH]

Ramanan P.S.

"I think that example means that as the interest rate on the debt approaches the growth rate, the interest on the debt approaches accounting for the entire current account deficit - which would mean that the current account deficit approaches the size of the services deficit (interest) without any goods deficit. Not sure. But it works so long as foreigners don’t won’t to sell their bonds back to the domestics in exchange for goods."

i.e. still no net tax burden, obviously

and if the foreign sector is in effect financing the entire interest on the debt, its difficult to invoke the "taxed in part" argument - because all taxes then must be directed toward domestic uses, by construction

[Greg Ransom]

I visit Matt Yglesias' grand Coolee Dam. I take charge of the dam.. I write Matt an IOU for one Grand Coolee Dam

1. If Matt subsequently tears up that IOU, then I am richer and he is poorer. Taking the two of us together, in aggregate we are neither richer nor poorer if Matt tears up the IOU. Tearing up the IOU doesn't eliminate the dam, which still stands.

2. Therefore there is no aggregate cost to me and Matt of me taking ownership of Matt's Bonneville Dam, and writing Matt another IOU.

1 is of course true. And 2 is equivalent to 1. Therefore 2 is just as true as 1.

[Greg Ransom]

I visit Matt Yglesias' grand Coolee Dam. I take charge of the dam. I write Matt an IOU for one Grand Coolee Dam

1. I notice Matt's 1,000,000 idle hand shovels. I take I take charge of Matt's idle hand shovels. I write Matt an IOU for 1,000,000 hand shovels. I notice 1,000,000 idle workers. I put the 1,000,000 idle people to work filling in the reservoir behind Grand Coulee Damn making use of the 1,000,000 previously idle hand shoves. The economy returns to full employment. All idle capital goods are put to use -- and used up -- all of the hand shovels are eventually broken or worn out.

2. If Matt subsequently tears up the IOU's for Grand Coulee Damn & the 1,000,000 shovels, then I am richer and he is poorer. Taking the two of us together, in aggregate we are neither richer nor poorer if Matt tears up the IOU. Tearing up the IOUs doesn't eliminate the dam, which is sits there -- but now with no reservoir standing behind it. Nor does tearing up the IOUS eliminate the physical remains of the shovels, which exist as broken parts littered all about Grand Coulee Dam.

3. Therefore there is no aggregate cost to me and Matt of me taking ownership of Matt's Bonneville Dam, and writing Matt another IOU.

1 is of course true. So is 2.

But notice what matters here. 1 and 2 are trivially true. Substantively, what matters is that,in aggregate, taking the two of us together, we are much poorer if we have destroyed the electrical generation and water storage and distribution capacities of Grand Coulee dam -- and broken our shovels to boot -- all because of the efforts which were made possibly by my ability create full employment in labor and capital via the appropriations made possible by the issuing of government enforced IOUs allowing for the legal transfer of resources.

As for 3, we don't know what the aggregate cost is of your taking ownership of Bonnerville Dam might be, unless we know what you do with it, and compare the to what Matt would have done with it.

[Ed Seedhouse]

"The original assumption (a scenario really) was that the debt would be paid down with taxes. The conclusion from that is that the debt is a net monetary burden (or real burden in Nick's apple model). And Nick's model (and my accounting model in the monetary case) shows how its a burden."

But of course that's an assumption that is irrational in a modern financial system. The government can always pay off the entire debt simply by creating money. Now that's probably a bad idea and it's probably a good idea for the Government to issue debt so people who want to save safely can do so. But it is wrong to say that the government funds it's spending with debt. It issues debt as a service to the economy but when the economy is not producing anywhere near it's capacity it is silly to say that the necessary deficit must be paid by issuing debt.

Of course we are talking about the national sovereign government, not Provinces or Cities. They have to earn revenue in order to spend it. But the Canadian government does not and there are situations in which it shouldn't. The real problems we face are not financial but physical, and the physical problems are long run while the financial "problem" can be cured quickly and relatively painlessly.

We then still have to figure out how to arrange an economy that is sustainable into the distant future, but we'll never be able to do that if we keep being enslaved to monetary illusions.

By the way I am the "FBU" you were talking to. I don't know why I thought I should do that - I have no desire to remain anonymous.

Ed Seedhouse (Just in case my name doesn't get attached to the response this time)

[Paul Rogers]

--Nick Rowe comments:

--"Google "No use crying over spilt milk", "bygones are forever bygones", "sunk costs"; understand the important role of this idea in economics; then retract your accusation."

No, Nick. YOU'RE the one making the accusations of major economists being clueless. Don't deflect.

So YOU google "Krugman wants to issue bonds and use the money for spilt milk/whatever..." and if you find anything remotely credible, I'll happily retract.

Until then, I remain convinced you've just set up a big fat straw man.

[Bill Woolsey]

In the real world, we do have capital.

If government debt crowds out the production of new capital goods, then there is less output in the future.

This effect exists if we assume a complete absence of markets, money, contracts, etc.

If Robinson Crusoe spends more time gathering coconuts and less time repairing the hut, when Crusoe Jr. inherits, he has to fix up the hut and has less time to gather coconuts.

Rather than Crusoe and Crusoe Jr., consider generations living in an "ideal" socialist commonwealth where mankind has taken its destiny into its own hands and directs all production and consumption (present and future.)

Everyone agrees that something like this is possible, and that maybe government debt in a market economy would could have an effect something like this--less capital accumulation.

The Lerner position is that if we assume that this does not happen, then there is no way that government debt can create a burden for future generations.

One way to see the Buchanan position is that if Lerner is right, then we are saying that debt financed government spending makes today's bond holder worse off and the beneficiaries of the government spending better off.

Buchanan's view is that no, the bondholders are not worse off. Presumably, they are slightly better off. It is an exchange. It isn't a gift or a confiscation of income or wealth.

Rowe's story is that the bondholders sell the bonds and consume later and they receive consumer goods produced by young people. When those young people are old, they will sell the bonds and get consumer goods from younger people, and so on.

I think the Lerner/Krugman assumption is that the bondholders just give them to their children. (You know, bondholders are rentiers--the exploiting capitalist class.)

Anyway, the beneficiaries of the government spending and those who bought the bonds are all dead now. And so, we just have the children. Some of the children just have these bonds that were a free gift. If there is a default, then they just don't get the gift. If, instead, interest and principle payments are made, then it is a tranfer, with the taxpayer (the working class) worse off, and those capitalst kids inheriting the bonds better off. (Doesn't it make punative taxation of interest income look just?)

Of course, if we look back to the original bond buyer, who we are assuming just wants to give a bequest of bonds so that his children can collect this future transfer, then no longer can we assume that the bondholder is really better off if this future transfer is conditional.

For example, suppose the bond contract was personal--you literally cannot transfer it to your heirs. Suppose that when you die, all government bonds are subject to a 100% wealth tax. Under those conditions, our capitalist class bond buyers would be giving gifts to fund the current government spending. Or.. maybe they wouldn't have bought the bonds.

Now, suppose we lie, and tell the bond holders that these are transferable promises that you can give to your children, but really, we don't plan on paying.

Well, there are benefits from government services, and the bond buyer is happy like before (he,he,he.. sucker.) And then, the spoiled capitalist children fail to obtain surplus value from the working class, so they are worse off than they would have been if the tranfers were made, and the workers are better off.

The Buchanan view is that treating government debt in this way is unreasonable. The bond buyer is no worse off and a bit better off because, in this situation, the children of the capitalist will earn income at the expense of the working class. Or, more generally, the future bond holders will earn income from future taxpayers.

If they really aren't contracts, then.. what are we talking about? Again, what is the burden of government spending when it is funded by giving people thank you notes? Well, in that case, the those giving away the wealth to fund the government spending sure enough bear the burden and there is no burden to future generations. Oh, unless we start imagining that they might have given the money to their children instead of to the government (capital accumulation.)

[JKH]

Ed,

“But of course that's an assumption that is irrational in a modern financial system. The government can always pay off the entire debt simply by creating money.”

It’s not irrational.

I don’t know if you’re coming from the MMT perspective or not (doesn’t really matter, since most of MMT is just accounting), but your interpretation lends itself to some observations from what I would refer to as some of the better insights of MMT.

We know that there’s no effective issue of solvency from a fiat currency issuing perspective. That’s another way of saying what you’ve said about the operational capacity for money creation to replace debt, which I agree with.

But the reason for paying down debt with taxes doesn’t necessarily have anything to do with the issue of solvency.

It may have everything to do with the issue of real resource capacity and inflation pressures.

So paying down debt with taxes is simply a form of fiscal tightening for those reasons.

And that’s a very rational proposition unto itself.

Now how likely is it that a fiat currency issuing government need go into surplus, (which is the precondition for paying down debt)?

Well, it’s not exactly the favourite dream of MMT’ers, but it happens. It happened in the US, and it happened in Canada, in a very meaningful way.

The point of Nick’s post is that it demonstrates the nature of the burden when it happens.

Now as far as the government paying off the debt with money creation is concerned, that’s irrelevant to the issue of the potential future tax burden. That does absolutely nothing to change the fact of the cumulative deficit. The cumulative deficit is now just intermediated into the financial system by banking system deposits instead of debt securities. The cumulative deficit remains. It’s just not accurate to label it as debt, because there are no debt securities outstanding.

And the fact that there are no debt securities outstanding in your scenario again does not change the fact of the existence of a cumulative deficit. And it does not change the fact that potential real capacity and inflation pressures may cause the government to tighten fiscally, to levy taxes, and to pay down the cumulative deficit or at least some of it (Nick’s logic only requires that a non-zero portion of the debt be paid down). And when the government runs a surplus to pay down the cumulative deficit, it will remove deposits from the banking system instead of removing debt. That’s the only difference, and it’s an irrelevant difference in terms of the logic of Nick’s model.

[JKH]

Bill Woolsey,

Regarding Rowe/Lerner/Krugman:

Agreed - that Nick’s model depends on the final cohort buying rather than inheriting the bonds.

If the final cohort inherits the bonds as a gift, that final cohort will end up whole and experience no net monetary or real burden in Nick’s model.

The inheritance event effectively transfers the burden back to the previous generation, which will absorb the net burden unless they too inherit the bonds from the generation before them.

And so on, regressing back to the generation that bought the bonds at tender and who also received the transfer payment. That first generation, which ended up with a net benefit in Nick’s model, would also end up flat instead if they bequeath the bonds to their descendents, instead of selling them.

In summary, if the inheritance chain is complete from start to finish, there is no net burden for any generation from the issuance of the bonds to their repayment with taxes.

I suspect this was all obvious to you, but I thought I’d write it out anyway.

[vimothy]

Re MMT's taxes and "real resources" angle, did any MMTers ever comment on Landsburg's "man who can't be taxed" article, which drew a lot of heat from the DeLong/Krugman axis last year?

http://www.thebigquestions.com/2011/04/18/the-man-who-cant-be-taxed/

[Ramanan]

JKH,

The interest rate less than growth rate condition has a lot of loopholes in my view. It was first used by someone named Evsey Domar and I don't know why most Post Keynesians believe it as if it is obvious (though not all). I know we have been in part of the discussion - as in not one-to-one but generally with many commentators around. I guess the discussions digressed because there are so many issues around this which are related. It's a bit difficult to summarize what I want to say, but in such "proofs" it is assumed that the primary balance does not get out of hand, i.e., it is bounded.

In the extreme case think of a 1% rise in CAD every year. i.e., 3% now, 4% next year, 5% the following year etc. Assuming a positive private sector balance, both the fiscal deficit and the current account deficit rise forever. If growth is somehow increased (with the aim of bringing ratios down or keeping them from growing), it is self-defeating because a domestic led growth will lead to an even more widening of the CAD. Hence an export-led growth is required.

How does this logic "fail" in a closed economy? A widening of deficit means a widening of private sector surplus and hence more consumption and higher national income and as a result higher taxes and lesser fiscal deficit- automatically stabilizing. For the case of the open economy if foreigners do not exchange their assets for goods, debt ratios can keep rising. However, there are realistic (very) situations in which they keep rising and domestic demand has to be controlled if its needed to prevent them from rising. Or else other institutional means.

Of course the stand that the process (i.e., rising debt ratios) can continue forever without an issue is a slightly different matter. I once had an argument with someone and he sorta figure out my points but my point was why bring in the condition r

That still leaves open the possibility that exchange rate market forces will lead to a "balance". However it is mentioned in "MMP" (no typo) that it may not work but it is not an issue since the imbalance is not thought of as an imbalance. And this is presented as a dissenter view and I dissent with this dissenter view.

[Ramanan]

"but my point was why bring in the condition r"

Second last para - didn't appear correctly for some reason.

but my point was why bring in the condition r less than g.

[Rob]

How relevant is what happens to the bonds between when they are issued and when the government pays them off to the overall story?

Unless I have oversimplified this then the important thing seems to me is that the government gives the apples raised by the initial bond sale to the first cohort, and eventually taxes the later cohort to pay off the bold holders. The story would as work as well even if the bonds were initially sold to outsiders. The important thing is that it is possible that a different group of people may benefit from government borrowing than the ones who have to pay it off, and these different groups may be in different cohorts.

[JKH]

Ramanan,

"but in such "proofs" it is assumed that the primary balance does not get out of hand, i.e., it is bounded"

not a complete response here, but the algebra assumes that the growth rate g generates a growth in spending and tax revenue of g, etc. and that growth trajectories overall are smooth extrapolations - so that's an ideal type of assumption relative to the real world

probably wouldn't help the debt math either if the United States did a leveraged buyout of the Eurozone

:)

[Ed Seedhouse]

"Now how likely is it that a fiat currency issuing government need go into surplus, (which is the precondition for paying down debt)?

Well, it’s not exactly the favourite dream of MMT’ers, but it happens. It happened in the US, and it happened in Canada, in a very meaningful way."

MMT, as I understand it anyway, and as stated on Bill Mitchell's blog several times IIRR, does envisage times when the Government sector should go into surplus, mainly to reign in inflation.

The surpluses in Canada and the USA were not for that reason and MMT blames them in large part for the crash. If the government runs a surplus then the combined private sector and external accounts must be in deficit. This is not in doubt as it is a simple accounting identity, true by definition.

Now Steve Keene ascribes the crash to hugely increased levels of private debt, which is far larger than the public "debt". As far as I can tell he advocates a different "get out of jail fast" solution but MMT agrees that it is the levels of private debt that caused the problem.

But, if the real economy is running at full capacity (including the external sector) then inflation starts up and at some point you take money out of the economy by running a surplus.

A fiat currency government does not have to run a debt to fund a deficit. The debt is a service as Government bonds provide the safest possible means of saving for those who wish to save without risk.

[rabbit]

Arguing over whether huge government deficits and debts are dangerous and undesirable is like arguing whether jumping out of a window from the 10'th story is dangerous and undesirable.

We can argue about the theory of gravity and general relativity and air friction and terminal velocity and whether people have wings till we're blue in the face. Or we can ignore all of that and look at what actually happened to people who jumped out of 10'th story windows.

[Greg Ransom]

This debate turns on the same issue Hayek's debate with Lerner over collectivist planning turned -- the constant re-evaluation of alternative costs and gains involving alternative uses of production goods. And Lerner's conceit in both cases is the same -- relying on the fantasy that resource allocations directed by government on the basis of backward aggregated sums of stipulated valuations stocks can magically retain and communicate meaningful future oriented valuations across time. E.g, we are to image that simply because we have stipulated that "no production of capital goods has been crowded out" based on _past_ aggregated valuations, that this prior "stock" of valuation / production goods magically retains its "stock of valuation" into the future despite the fact that the government has re-arranged all valuational relations -- especially in the domain of production goods and processes -- based on BACKWARD LOOKING and NONRELATIONAL measures of "value". In fact, when the government borrows, it _disrupts_ the valuational relations of production goods -- and even if according to a backward looking valuational measures of "aggregate output of new capital doors" there has been no loss of output, in the only perspective meaningful to economic valuation -- the future relational significance of goods -- there will have been an alternative cost loss because forward looking private entrepreneurs would have re-arranged production relations to fit the new relational conditions, while the assumed perspective that sees past valuations as fixing a non-crowded out "stock" of new capital goods will leave us with goods which are out of valuational sync with the newly created, post-debt production relation situation.

And this doesn't even get into how government would take resources away from greater valua creating alternative production processes -- which only Lerner's assumption of impossible magic prevents from taking place.

"If government debt crowds out the production of new capital goods, then there is less output in the future ..

Rather than Crusoe and Crusoe Jr., consider generations living in an "ideal" socialist commonwealth where mankind has taken its destiny into its own hands and directs all production and consumption (present and future.)

Everyone agrees that something like this is possible, and that maybe government debt in a market economy would could have an effect something like this--less capital accumulation.

The Lerner position is that if we assume that this does not happen, then there is no way that government debt can create a burden for future generations."

[Greg Ransom]

"The Lerner position is that if we assume that [loss of new capital production or "crowding out"] does not happen, then there is no way that government debt can create a burden for future generations."

But this BEGS THE QUESTION at the HEART of economic science -- what gives production goods their valuational significance?

Production processes get their valuational significance via a constant future looking re-configuration of alternative production combinations and pathways. If massive government borrowing and non-forward looking / non-economic government resource allocation throws a massive monkey wrench into the gears of the production system, there is no way to calculate a comparison of what is or what is not a comparative "loss of capital production" or is or is not a "crowding out" of production or investment, matching up some aggregate valuational benchmark of past "output" and some aggregate valuational measure of post-borrowing "output" any more than the old Soviet Union had any measure besides the changing quantity or weight of "nails" in different time periods as a measure of constant, increasing or decreasing "output" of new capital goods (I.e. nails).

[Max]

"Well, it’s not exactly the favourite dream of MMT’ers, but it happens. It happened in the US, and it happened in Canada, in a very meaningful way."

Last time the U.S. was in surplus I don't remember anyone cursing previous generations for their profligacy. Everyone was pretty happy about it.

[Greg Ransom]

I hope folks know the history of what happened in the Soviet Union, re nails.

If the government requested increasing output in nails according to the metric of weight, nail factories would produce giant, very heavy nails.

If the government requested increasing output in nails according to the metric of quantity, nail factories would produce incredibly tiny and thin nails by the billions.

In each case, according stipulation for "greater output", the Soviet economy would be counted as having "increased production" -- irrespective, of course, of the RELATIONAL significance of that "output" to all other trains of consumption and production goods in the economy.

"any more than the old Soviet Union had any measure besides the changing quantity or weight of "nails" in different time periods as a measure of constant, increasing or decreasing "output" of new capital goods (I.e. nails)."

[Ramanan]

"but the algebra assumes that the growth rate g generates a growth in spending and tax revenue of g, etc."

Nice point JKH, my point being that sustainability analysis requires one to show that tax revenues grow at g and expenditure doesn't grow faster than g, instead of assuming it in the proof.

[JKH]

“Last time the U.S. was in surplus I don't remember anyone cursing previous generations for their profligacy. Everyone was pretty happy about it.”

The cursing of MMT’ers about that same surplus now probably makes up for it.

[Bill Woolsey]

Ransom:

I don't get it.

Is the burden of government debt on future taxpayers anything other than an alternative framing of the allocation of present production between current and future goods? Lerner says no. Buchanan says yes.

Whether or not we can, as a pratical matter, determine what is happening to the allocation of resources between future and present goods in response to anything, including a budget deficit, is another issue.

My way of thinking about budget deficits (and according to Murphy, Mises' view as well,) balls these things together. The way I would put it, the opportunity cost for the government bond buyer is purchasing capital goods. Neither they, nor their heirs benefit much from the debt. (I think there is a change in real interest rates, and so some benefit.) Those receiving the benefits of the government programs benefit, and those who pay future interest and principle bear the burden.

Rowe (and Murphy has been doing this too,) has constructed an all consumption, fixed capital example that shows that there can be a shift of consumption from the future--basically, step by step. This is an important argument.

Buchanan, on the other hand, had something else in mind. My interpretation was that treating the payments of interest and principle during the current generation as just an arbitary transfer means that the deficits in past generations weren't really being financed by debt. Buying bonds isn't giving gifts or paying taxes. They weren't worse off, and really, were a bit better off. But that is only true if the interest and principle payments occur in the future. And there are people who lose from the transfers in the future. I am pretty sure that Buchanan wasn't thinking about shifting consumption by an overlapping generations model.

Much of the discussion here has been of the ponzi business. One of the comments I liked best was the question of which generation gets to enjoy the benefits of the ponzi scheme? I would think that each generation should share it equally. Since the benefit is finite, how much doest that provide for each generation going on to eternity?

By the way, will human society as we know it exist forever? In particular, can any particular nation state be assumed to exist forever? In particular, if the U.S. one day becomes a wilderness, or is mostly under the sea, what happens to these bonds? Default? When the Americans all emigrate (to outer space) do the debts follow them? What if they die off because couples choose a one child policy?

To get more fantastic, when Americans start turning into beings of pure energy, how is the national debt settled?

The ponzi scheme assumes there is always a future generation.. forever.

I suppose if the world ends suddenly, then we have default, but the bondholders don't have anything to lose, having no futher opportunity to consume and no decendants to receive inheritances.

Too science fictiony? Well, mabye the "ponzi" plan is really just as unrealistic.

[Ed Seedhouse]

"I hope folks know the history of what happened in the Soviet Union, re nails."

This is just name calling. No one in the MMT camp wants to eliminate markets. Personally I am not and never have been a socialist, and think markets are a great human invention up there with the wheel, fire, and bureaucracy. To imply that I want a Soviet system is just odious.

But the evidence is clear that markets are unstable unless properly regulated, and MMT offers a way to keep them stable at least so far as the monetary economy goes. Moreover it is neutral on the question of how big the government sector could be. Even if the only thing government did was supply money MMT would still work from their viewpoint.

I prefer government to be "right sized", doing things that government does best and leaving the rest to markets. Natural monopolies are best run by Government, population wide insurance schemes work best if run by government as a straightforward consequence of the law of large numbers.

Even my libertarian friends agree that the police and the army need to be run by government. A "free market" system cannot operate when the mafia is free to intervene and no one is there to stop them.

You may agree or disagree with my viewpoint on what and what not should be run by governments, but MMT, if it works, works regardless of the size of government.

[Greg Ransom]

"Rowe (and Murphy has been doing this too,) has constructed an all consumption, fixed capital example that shows that there can be a shift of consumption from the future--basically, step by step. This is an important argument."

To me, this seems trivial -- although it is patent that Krugman and Lerner often gets even trivial matters wrong (see Hayek's "Two Pages of Fiction" on Lerner).

If a thief steals my bank account and spends in on the last 3 days of my mothers life trying to keep her from a natural death, there has been a generational transfer of wealth.

If the government borrows my bank account and spends it on the last 3 days of my mothers life trying to keep her from a natural death, there has been a generational transfer of wealth -- and then later taxes me again to retire that debt, there has been a generational transfer of wealth.

The magicians sleight of hand from Krugman and Lerner is the attempt to hide this transfer by hiding the transaction in unearned premises that "capital investment has remained the same" -- somehow we a suppose to simultaneously (1) increase government promoted consumption now and at the same time (2) maintain or increase future private production output later, and do this without transferring goods inter-generationally. The mind boggles at how _any_ "re-allocation" mechanism could magically pull that one off, give the fact of scarcity and heterogeneity and on-going inter-relational valuational and production process changes over time, etc. (Facts which Krugman & Lerner constantly ignore).

They are simply imbedding logically -- and mathematically --incompatible premises while telling everyone that "by assumption" this incompatibility has been ruled out in advance (a statement which itself is among the logically incompatible statements).

One of the magicians slight of hands used by Lerner and Krugman et all is to pretend that there is a cardinal or time-univocal metric which would provide an accounting metric to show us by summing up "stocks" that debt financed government spending can both shift current consumption within the current generation and maintain future productive output for later generations.

But no such metric exists within the economic way of thinking (changing marginal relational valuation across time and subjective alternative cost evaluations by entrepreneurial actors) -- or within the real world of real people in the actual private property world making use of local knowledge and changing relative prices while coordinating their consumption and production processes across time.

Individual choosers can compare their expected outcomes involving relations between particular alternative production and consumption choices -- using the economic way of thinking -- but governments and economists cannot aggregate stocks using a univocal and cardinal metric across goods and across time in any sense that has a meaning from the economic point of view, the only point of view that allows us to coordinate alternative production processes across time in an economically relevant fashion.

[Greg Ransom]

Lerner here, and in his work on collectivist economic planning, assumes away the problem re-evaluating and re-coordinating production goods processes given scarcity, limits to knowledge, heterogeneity -- while at the same time asserting the false pretense that he doesn't assume these away.

These things don't go away because he asserts as a premise of his model that these problems don't exist by assumption.

"Is the burden of government debt on future taxpayers anything other than an alternative framing of the allocation of present production between current and future goods? Lerner says no."

From the economic point of view (marginal valuation, alternative costs, etc.) the most important thing that government debt and the burden of government debt do is make a economically meaningful or coherent "alternative framing of the allocation of present production between current and future goods" strictly impossible.

And from a pragmatic point of view utterly improbable to the point where it is impossible any any non-magical world we will ever live in.

But the intent of Krugman and Lerner is to make the for-all-purposes completely impossible world seem to be the world we actually live within.

If you look under the hood the assumptions are magical. If you accept impossible cardinal accounting metrics in violation of the economic point of view, the impossible becomes magically "tractable" and imaginable.

[Greg Ransom]

Wow, Ed. I didn't imply or suggest any such thing. What you write has nothing to do with what I wrote.

[Ed Seedhouse]

"The cursing of MMT’ers about that same surplus now probably makes up for it."

Well, the prediction by MMT that big government surpluses would lead to trouble happens to have come true. That might, of course, be just luck. But the prevailing orthodoxy among economists at that time was that a crash like the 1930's just could never happen again. Reality disproved them, yet will they recognize that there was something terribly wrong with their belief system and they need to adjust it to fit? Nope.

If MMT is wrong, fine. Prove it. Show where the fallacy is. But you won't be able to do that even if it is fallacious until you understand what it actually is saying, and the comments on this thread appear to me to show that this has not been done by it's "critics" who are not actually criticizing it but instead criticizing their incorrect beliefs about what it says. Well, straw-man beating is nothing new, and it doesn't seem it will go away soon.

Even Krugman dismissed MMT by claiming it says something that it simply does not say. Steve Keene has some objections which may well be pertinent, but no one is listening to him, either. He has, if I follow his reasoning and I think I do, shown that neoclassical economics prevailing model has a built in contradiction. If it doesn't, then show why it doesn't.

Until someone does that I will continue to believe that some paradigm like MMT or Keen's gives us a chance to understand how economies actually work, unlike the current orthodoxy.

[Greg Ransom]

Buchanan says you can only think about these things using the logic of subjective marginal valuation by individual evaluators/choosers. Lerner says you can think about these things -- e.g. no loss of production output across time -- assuming cardinal metrics and aggregates "stocks" that have meaning across time irrespective of individuals constantly re-coordination production processes on the basis of individual subjective evaluations as the margin in the context of changing relative prices and changing local knowledge. This is the only way that Lerner can say that the forced re-distribution across time and between production and consumption processes via massive borrowing and taxing by the government can be assumed as output neutral.

Bill writes,

"Is the burden of government debt on future taxpayers anything other than an alternative framing of the allocation of present production between current and future goods? Lerner says no. Buchanan says yes."

[Greg Ransom]

Shifting costs via force -- i.e. the promise of forced taxation in the future and present borrowing allowed by the promise of that force -- SHIFTS the time costs of consumption and production across time.

It's going to NECESSARILY alter ALL of these relative relations between time structure production and consumption plans.

And what it will unavoidable do is make consumption now less costly and provision for future output more costly -- and only by "magic" will the government no how to re-coordinate all production and consumption processes to avoid this fact.

[Greg Ransom]

Similarly, it is equally true that only via magic would private individuals be able to re-coordinate production and consumption processes. But of course,

I wrote,

"And what it will unavoidable do is make consumption now less costly and provision for future output more costly -- and only by "magic" will the government no how to re-coordinate all production and consumption processes to avoid this fact."

The assumption of that public debt and future tax increases changes nothing in terms of output and consumption across time can be true only on the assumption of "and then a miracle occurs" -- and this miracle is made to appear by non-economic thinking using cardinal stocks and aggregate modeling shifting "value" across time in boxes full of univocal "value" unrelated to changing marginal relations between production processes and consumption processes taking place through time.