« A Long Term Employment Picture - Go West for Opportunity | Main | Manufacturing, Technology, Canada and the Dollar. »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

"People won't want to consume 10 apples when young and 90 when old (which is what is implied by a high debt/GDP ratio) unless you offer them a very high interest rate to persuade them to defer consumption."

Nick, isn't lending to the governement voluntary? Did you go to the Delong link? Doesn't that show that the interest rate has not been kept high?

I said ealier that by keeping the interest rate *permanently* higher than the growth rate you create a prisoner's dilemma type scenario.

Oliver: You haven't gotten the intuition yet. You are hung up on exactly the same issue that everyone hangs up on. Just because C=100 in every year doesn't mean lifetime C=100 for every cohort. You really need to work through an example. Have another look at this one from my previous comment. Notice that C=100 in every year.

2. Lower lifetime consumption, but lifetime consumption asymptotically rising towards 100 as we go further and further forward in time. E.g. cohort C consumes {20,70}=90, cohort D consumes {30,65}=95, cohort E consumes {35,63)=98 etc. So each cohort has a lifetime consumption less than 100. But notice that 70+30=100, 65+35=100, etc.

You would get a pattern roughly like that if the government raised taxes to pay the interest plus about half the principal back each period.

I'm starting to think you guys declared victory early. That Krugman and Baker assume (perhaps intuitively) growth and interest are equal in the long run equilibrium. And that you guys only proved them wrong by assuming untrue conditions.

If interest rates are too high growth suffers, increasing the debt burden. If interest rates are too low inflation happens, reducing the debt burden. Under normal times monetary policy may be sufficient. But interest rates have a zero nominal bound, which may make fiscal policy necessary.

If governement spends wastefully future generations may be worse off. If they spend wisely future generations will be better off. But that's a seperate issue from whether there will be a monetary burden. (There must be a burden to the extent interest rates are above the growth rate. There can't be a burden to the extent interest rates are below the growth rate. There is no necessary burden to the extent the interest rate is equal to the growth rate, the relative "burden" depends on how the money is spent, but this is a seperate issue from the debt.)

Bob Murphy: "I.e. the government doesn't do investors favors by providing a debt market."

Ah, now that's a beautiful -- and off-topic -- subject.

If we were in MMT World, the government wouldn't issue debt. It would just spend, creating (electronic) dollar bills instead of treasury bills/bonds.

Just because the market wants a return on their safe assets doesn't mean the government is obligated to provide it.

@Nick:

"Yes, we have very good commenters here."

This is because our esteemed blogger is so generous in his explanations and discussions. Thanks for all your time.

"The kids' apples are eaten by their parents. But the parents are just being compensated for their apples that were eaten by the grandparents, when the parents were young."

By the "grandparents" here, are you talking about the first, deficit-spending generation?

If *not*, if you're talking about an intermediate generation, does this mean that in every period, older people eat some of their kids' apples? Does a government deficit today impose that dynamic -- the old eating more than the young -- until the end of time (or until taxation happens and short-circuits that dynamic)? Doesn't seem right but it does seem to be what's suggested.

I'm starting to think about this in terms of a wave: a disturbance propagated through a medium. The medium doesn't move; only the disturbance does. Eventually it hits the beach.

Interesting. So, the initially positive effect of borrowing (in this case +30 (50+80), if one assumes 50/50 before the 20/80 period) will, at the defined rate of interest and paying down net principle, affect following cohorts negatively while edging back to 50/50 asymptotically? I assume the cumulated losses in lifetime utility will add up to the initial difference of 30 in inifinity? And also I trust your part of the math is correct...

While this does beg the question of why one would pay down principle in such a way(as pointed out by JKH above), I do now see the theoretical possibility for such a mathematical artefact, but I can't quite grasp the relevance of it for the real world - yet.

Steve: the wave is good.

Start out with a flat surface. Everybody consumes 50 when young and 50 when old. Then we create a disturbance. The first generation consumes 50 when young and 60 when old (the government borrows 10 and gives them 10). The wave begins. The next half a dozen generations now consume 40 when young and 60 when old, as the wave propagates (the government increases taxes by just enough to pay the interest on the debt, so the debt is constant). Then the next generation consumes 40 when young and 50 when old (the government pays off the debt all at once). The wave stops. All subsequent generations consume 50 when young and 50 when old.

Steve, "By the "grandparents" here, are you talking about the first, deficit-spending generation?"

Yes, in Nick's model the gradparents lend apples to the government for interest. The government then gives the apples right back for nothing. Why did they do this?


"Does a government deficit today impose that dynamic -- the old eating more than the young -- until the end of time (or until taxation happens and short-circuits that dynamic)? Doesn't seem right but it does seem to be what's suggested."

I believe governments tend to fund direct transfers (that are not means-tested) with taxes. So you shouldn't see (in a non-corrupt system) the old eating more than the young. (I'm ignoring demographic and aging issues.)

Nick's model is confusing because he uses apples instead of money. And has no trade. In the real world even if the government gave the old a free bond-finaced transfer, the old would be forced to buy goods from the young. Who could raise prices.

Bob Murphy: "people alive at that time have a gross burden"

You keep saying this. Can you explain the significance of "gross burden" when you use it?

Maybe a silly question, but since when do transfers need to be financed? I thought transfers are just that - income transferred from one person to another. Whether they like it or not is a separate matter but I don't see any debt being incurred. And investing for one's own pension is another thing altogether. But in that case, no exchange between cohorts takes place, one can only run in to good or bad luck with one's own investment decisions. What am I missing?

@Nick:

"Too late. We'll all be dead by then. They won't be able to get back the apples we've eaten."

Don't you actually mean "the parents in that period won't be able to keep themselves from eating their children's apples"?

i.e: "The next half a dozen generations now consume 40 when young and 60 when old"

So it seems that you're saying: government deficit spending today *forces* future parents to eat some of their children's apples.

Toss a rock into a pool, and all the bugs floating on that pool must deal with the resulting waves. They have no alternative.

Until, taxation zeroes out that wave, like noise-canceling headphones.

Do you think that deficit spending today propagates incentives and constraints into the future in that way, effectively enforcing the behavior you describe?

I'd also really like to hear your response to JKH; I don't think Bob answered him. (JKH does accounts in his head better than anyone else I've read -- in his words, he's done lots of "pushups" -- so in my eyes his question merits serious attention.)

Does the post-tax generation "recoup" the consumption lost by the taxed generation?

IOW, does the fact of an intergenerational transfer at the point of taxation demonstrate that deficit spending today imposes a burden on all posterity?

JKH's other comment basically defends the MMT camp (though he's not a card-carrier himself), so is of less interest to me. But still, I think he's right, in a slightly tangential way.

Oliver: The government doesn't own any apples. If it wants to make a transfer of apples, it has to get those apples in a tax, or else borrow the apples. So the government borrows 10 apples from each A, and gives them straight back as a transfer. It is exactly as if the government just gave each A an IOU (bond) for 10 apples.

And if you want to "invest" in your pension, what assets are you using? You buy a bond from an old person.

Steve: "So it seems that you're saying: government deficit spending today *forces* future parents to eat some of their children's apples."

Not quite. The rate of interest must be such that the children *want* to buy the bonds off the parents.

Nick, I guess you're ignoring me. (I'm to you as you are to Krugman.)

I'm still hoping to get a response. (I'm sure I could be dead wrong.)

1. Wouldn't taxes still need to be raised if growth and interest rates rose equally? So, Krugman and Baker could assume no burden without Ricardian Equivalence.

2. Am I misinterpreting the Delong link?

3. You're now saying that if the interest rate is not high, people will not lend to the governement? Is this true in real life? Assuming it is, is what you're saying, "all borrowing needs to raise future growth by as much as the interest rate or there is a burden?"

(If borrowing fails to raise growth, should the government try to inflate away the burden. (Only to remove the burden on the young, not to put a new burden on the old.) Could they do this without raising taxes? Would inflation-indexed bonds constrain their ability to do so? Do inflation-indexed bonds constrain monetary policy and/or fiscal policy more generally?)

@Oliver: In my opinion, this clearly violates your own condition that there are always 100 apples consumed per period (assuming people don't live forever).

No that's wrong. Looking at a person's consumption over his lifetime, it is lower, even though in any given year, apple output is still 200 total (for two people). I mean, you are ignoring the whole point of Nick's original model. The reason it all had to come to a stop (thus leaving 4th through Nth generations out of it) in Nick's model, is that he he paid the debt off before it consumed all of GDP. But if we just tax each generation enough to cover the interest, while rolling over the principal, then it can last forever.

@Nick: We would have to think more carefully about what we're assuming in terms of bequests etc., but in principle why couldn't the apple-debt be 110% of GDP or more? As long as nobody minds passing it on to his heirs, then it's fine, so long as apple output eventually grows faster than the apple-debt principal. If old people don't want to leave any assets, and want to consume their estates in their last year, then yeah we ran into the practical constraint. But I was just arguing the general point about, "What if g=r in the long run?"

Anon wrote: The last part of your post seems to be invoking the invisible bond vigilantes. I believe Japan has debt to gdp of over 200%.

Yeah, and Japan also produces more than apples. Nick Rowe and I are idiots!

I picked 110% just to have a nice number. It could be 1 billion % of GDP. Do you think investors might worry then?

Again, neither Nick nor I am saying, "Debt will always and everywhere burden future generations, using mainstream economic analysis." No, we're saying that Baker and Krugman were wrong to try to prove the *opposite*, namely that in the absence of foreigners and distortionary taxes, debt can't possibly hurt our descendants.

@Steve Roth: You asked me to clarify the distinction I keep making between gross and net burden. OK here's the quick version:

Krugman and Baker kept trying to say that our great-grandkids (say) can't COLLECTIVELY be burdened by the taxation needed to service the debt. Sure, they conceded that some of our great-grandkids could be made poorer by tax payments for paying interest/principal, but thought this would be exactly counterbalanced by those interest/principal payments going into the pockets of the subset of our great-grandkids who were holding the Treasury bonds. That's why they kept saying it was important for Americans to "owe the debt to ourselves," because to the extent that those payments flowed to Chinese or whatever, then they agreed that American great-grandkids *could* be hurt even collectively.

So what they were trying to say is that, vis-a-vis the great-grandkids as a collective, the debt wasn't even a *gross* burden. I.e. you don't even have to tell me what the government today does with the money. Whether it spends it on heroin parties or a space station that will provide solar energy to our great-grandkids, our great-grandkids aren't hurt *as a group* by the existence of that debt.

So I'm (following Nick) saying that that is wrong. If the great-grandkids holding the Treasury bonds had to earlier reduce their consumption to induce others to sell them the bonds, then it's not a wash. The great-grandkids who are taxed (say) $100 billion to service the debt are clearly down $100 billion, but the great-grandkids getting paid the $100 billion aren't enjoying a pure $100 billion windfall; they're mostly just getting their principal back (or whatever).

So we see that if Ricardian Equivalence isn't 100% true, then the taxing needed to service/retire the debt imposes a gross burden on our great-grandkids, collectively. It's not merely that some win and some lose, but that the losses to the losers outweigh the gains to the winners--if we ignore the possible benefits of the government deficit spending.

To determine if our deficit-spending today is a *net* burden (not just gross), we have to ask what we're spending the money on. If we spend it on heroin parties, then it is a net burden too: Our great-grandkids are poorer (over the whole lifetimes) because of the taxes needed to service the debt we shoot down the pipeline at them, and they don't get any compensating benefits from our heroin parties today.

On the other hand, if we use today's deficit spending to build a space station that beams our great-grandkids really cheap solar power, then they will still be made poorer (on net) from the taxing to deal with the debt that we created, but they will be made richer by the solar power. And so on net, they could be glad we ran up the deficits today.

The important thing here is that Krugman and Baker weren't merely saying, "A deficit right now will help our great-grandkids more than it will hurt them." No, they were saying that if we ignore distortionary taxes and foreigners holding the debt, then there is no "hurt" at all that needs to be compensated. Considering the great-grandkids as a whole, they argued, it's all a wash--the gains of the winners (holding the Treasury bonds) cancel out the losses of the losers (paying the taxes). And this is wrong, once you think of an OLG model.

@JKH I think the base analysis can be translated into an MMT view without any change. An MMTer who wants to monetize the debt is still dealing with inter-generational burdening. Some old living cohort (whose members consumed less as children) cannot be made whole without taking from the now young, living cohort.

Thinking about the MMT "government bonds and cash are interchangeable" view has gotten me thinking that the issuance of bonds is not the *only* way to create burdens on future young cohorts. For example, suppose the government dropped cash on cohort A; the members of A decide to temporarily hoard the cash (rather than bidding up the price level); once cohort B is born and working and producing apples, A decides to buy those apples and eat them; cohort A dies.

This is basically the same as the bond example, except that the government does not borrow money and spend it (and therefore does not create immediate beneficiaries of fiscal largesse). The transfers are limited to the chain of "new cash holders" as the cash is spent through successive periods.

Unless I'm mistaken, the thing I've just described is an inter-generational Cantillon effect chain.

@Bob Murphy: Thanks, I get it: net is gross minus the benefits from debt-financed spending.

"Steve: "So it seems that you're saying: government deficit spending today *forces* future parents to eat some of their children's apples."

Not quite. The rate of interest must be such that the children *want* to buy the bonds off the parents."

'kay, incentives not constraints are propagated into the future.

But in any case, govt spending from deficits instead of current taxes creates a situation that ripples on: in every period, the old eat more, and the young eat less.

This is true, relative to the current-taxing otherfactual, even (especially) if G>R, and/or net (including benefits from the initial spending) is positive.

If the eventual taxing never happens, the old/young consumption pattern continues throughout eternity. 'zat right?

I guess old people are just smarter than young people. ;-)

Here's something that helped it click for me, for the 6 of you still reading at this point:

Suppose there had never been any deficit spending. Then the government in the year 2150 out of the blue issues $1 trillion in debt that the government owes to the holders of the bonds. It just hands those bonds out, to people it picks from the phone book randomly.

Then, in order to pay off the debt, the government imposes a one-time surtax of $1 trillion on all the taxpayers, and retires the bonds it just handed out 15 minutes earlier.

Did these operations hurt our descendants, considered as a whole?

No, of course not. Some of them were hurt--because they didn't get selected from the phone book to be given the bonds--but these people's losses were exactly counterbalanced by the gains to the winners.

On the other hand, notice that our descendants weren't helped, either. The gains to the people who were picked randomly, and got clear windfall gains from holding the bonds, were exactly counterbalanced by the losers who had to pay taxes on the debt.

OK, now tinker with the scenario: Instead of those people getting the bonds because they were lucky and were picked from the phone book, *instead* I tell you that they are holding those $1 trillion worth of bonds, because 20 years earlier they cut their consumption in half and gave the money to some older people, who then spent the money on sushi and cruises.

Now in this revised scenario, how do the people in 2150 fare, as a result of the taxation needed to retire the $1 trillion in debt?

Clearly these people as a whole have to be hurt. We already know that if the bonds were a pure windfall gain to the holders, then it's a wash. So if I now tell you that holding the bonds involved a preceding act of abstinence (in itself no fun), then clearly the losses to the taxpayers outweigh the (lifetime) gains to the bondholders.

Does that analysis do anything to help?

Bob, if it's 1 billion % of GDP and g=r in the long run then they should worry because of distortionary taxes/incentives, not because it's a burden on future generations.

Obviously, what we do now will either harm or benefit future generations. This is true even with no debt.

Krugman says explicity, "the problem is interest," "tax rates will have to go up," and "governments need to ensure debt grows more slowly than their tax base."

If the tax *base* is rising faster than debt (or debt service), growth is rising faster than (or equal to) interest.

Perhaps when he and Baker talk about "distributional problems," they're referring to a situation where one group owns all the debt. (Like in Nick's model.)

Also, Baker mentions the Fed holding assets. (I don't quite understand this, but might it be an alternative to paying bond holders interest.)

Perhaps both he and Krugman assume: 1. Responsible government. (Krugman mentions this, too.) 2. Healthy growth.

Maybe they downplayed the problems the debt can cause, but you and Rowe exaggerate the problem. (By using apples instead of money.) (With money) The oldies would have buy goods and services from youngies. The youngies could respond with higher prices.

Nick,

You might have not seen my arguments elsewhere (as Adam K). I will restate them here to increase the chance of taking the bait...

Summary: the whole reasoning is based on a flawed assumption that the government debt has to be repaid from taxes. Another flawed (gold-standard money era) assumption is that the interests on the public debt have to be high enough to allow for an unabated growth of the debt to GDP ratio.

I think that Robert Murphy’s example is a bit more sophisticated and better illustrated but he makes a similar point. I have several objections to the example and to the way it is interpreted.

NB I am not an economist. Things like “Ricardian equivalence” or even “rational expectations” sound like snake oil to me. I also can assume that I can travel it time if I go faster than light. If I start thinking like that in my work the most likely outcome is that I won’t be able to design a working solution (I integrate IT systems) and will lose my job. Nor you nor prof Krugman would hire me then because I have no degree in economics. Therefore I try to build and use models only when they are valid.

Let’s talk about the reality not about the (rotten) apples.

1. Why do you assume that the public debt has to be repaid? Because of the excessive burden of the interest rates leading to an excessive redistribution of the real income between the “99%” and the bond holders? So let’s reduce the interest rate on bonds (like in Japan). The Central Bank can buy the bonds.

Why during a recession when the (American) government wants to prop up spending, bonds (interest-bearing securities) are still issued (at a very low but still above zero, yields)? This is self-contradictory. I could understand issuing bonds in 1942 during WW2 to convince people to save more as a "PSYOP" and interest rate maintenance exercise but what are they needed for in 2012? Just because the government is incompetent and has been hijacked by the Wall Street?

You assumed that “[t]he rate of interest must be high enough to persuade them to eat fewer apples when young and more apples when old, otherwise they wouldn’t have bought the bonds from cohort A.”

What for? Apple bonds can be bought by the central bank for apple currency. The assumption that the market dictates the interest rates makes the whole example inapplicable to countries like the US, Japan, UK or Australia. This is the way to model the meltdown in Greece or Italy. But in the end ECB has blinked and they are buying bonds like hell.

Will “cheap money” or extra reserves lead to another credit bubble when the economy recovers? This is what Paul Krugman claimed a few months ago. Banks do not lend reserves and the money multiplier model is invalid. If there is a risk of excessive credit creation leading to inflation then the government should apply brakes on speculative lending such as 50% deposit on new mortgages or even a tax on net loans levied on banks. This is the correct solution anyway as cranking up interest rate will damage the productive economy and may not stop the bubble on its tracks.

If you don’t repay the debt there is no issue of overtaxing a certain cohort in the future. There still can be a redistributional issue between the different cohorts living at the same time in the future if the real interest rates are high. NB in the real world the redistributional flow has never exceeded 3.5% in the US (source: Bill Mitchell's blog)


2. (An Old-Keynesian argument partially addressing the redistributional problem) The government does not need to run primary surpluses in order to service the debt as long as in the long run the rate of growth of nominal GDP is greater than the nominal interest rate on the debt multiplied by the ratio of debt to GDP. As long as this condition is true, the ratio of debt to GDP will not increase. That condition was true for the most of the immediate post-WW2 period in the US - there are nice graphs on Bill Mitchell's blog - and the debt was not a burden in the sense demonstrated in the example with the apples. Of course, on the margin there will always some trade off unless the interest rate is zero.

3. You haven’t distinguished between the case when the economy is already running at full capacity and the government merely tries to redistribute the GDP by inducing higher saving (like during WW2 – issuing bonds was just a part of the operation, strict price/wage controls and rationing of products was also relevant) and the more common case when the economy is running below the full capacity. In that case the Skidelsky’s argument is relevant as extra government spending (offset by bonds sales to the non-government sector or merely to the Central Bank) will most likely lead to higher GDP in the future.

4. (merely a comment not a critique) I have to admit that the original example itself is possibly incomplete as we don’t know whether there is anyone who is going to eat apples when “the government decides to pay off the debt by imposing a tax of 121 apples on each young person in cohort C, which it uses to buy back the bonds from cohort C. Each member of cohort C eats 121 fewer apples.” This has been fixed by Robert Murphy. I am assuming that you wanted to say something similar to what he said. The JKH's comment is an attempt to correct the issue. It may appear that you have agreed with this statement by saying that "Then everything goes back to normal."

5. We have to distinguish between the redistribution of income between :
5a. the different time periods (impossible unless someone ignorant first props-up the economy in period A and then throttles the demand on a purpose by excessive taxation in period B), the original Lerner's argument applied to this case
5b. the different cohorts of people living during multiple time periods
5c. the different groups of people in general due to taxation and transfer payments

I would argue that 5b (described in detail by Robert Murphy) is not caused by the issuing or public debt itself. It is merely a result of transfer payments. I can attempt to build a formal proof debunking that claim if the statement related to causation is written in a more formalised language.

Let's modify your example. Let's assume that in period 1 the evil Stalinist government of USA does a transfer payment of 100 apple-money units to the lazy parasitic unemployed, members of the Occupy Wall Street and White House movement. The in period 3 the government of the Chinese province of South Vancouver (ex-USA) reduces the money supply by 133 apple-money units by levying taxes on the descendants of the original unemployed working as serfs on the cotton plantations owned by the apparatchiks of the Chinese Communist Party.

What's the difference? That no bonds were used? But the net effect is the same. So there is a re-distributional issue within the population living at time period 1 and 3. Why do you claim then that this re-distributional issue a result of issuing debt and then taxing back the people? I can envisage an example where the government issues 100 apple-money units of debt in period 1 to finance public infrastructure during a recession thus enriching everyone living in that period of time (and allowing for saving of some of the income by generation A) and then when the members of the generation C want to spend more than they earn - redeems these bonds and offset the extra spending by levying appropriate taxes. As long as the share in the total (aggregate) amount of goods and services which can be bought for the bonds does not change, nobody loses and everyone wins. This would require moderating interest rates. Again - can't we do that on domestic currency-denominated debt?

Sorry, looks like I have multiple @JKH posts. My browser freaked out, so I posted once more, albeit with slightly different wording.

[marris: I unpublished the more recent one. Let me know if it's not OK. NR]

Regarding MMT, below is Mosler’s direct response to the debate. But it’s obvious he actually hasn't read any of the detail of it, and therefore hasn’t understood the nature of it. If he had, he would have responded along the lines of my earlier comment regarding MMT.

The only thing he’s done is address the standard boiler plate MMT issue of the debt not being a burden, in the sense that insolvency is never an issue for a fiat currency, and that the government can always repay debt by crediting new bank reserves to the system. That's a reasonable point, but not at all the one that this marathon has addressed, which is that of the net effect of taxation on future generations, when taxes are imposed to pay down the debt.

http://www.multiplier-effect.org/?p=3192#comment-10715

@Bob Murphy

No that's wrong. Looking at a person's consumption over his lifetime, it is lower, even though in any given year, apple output is still 200 total (for two people). I mean, you are ignoring the whole point of Nick's original model. The reason it all had to come to a stop (thus leaving 4th through Nth generations out of it) in Nick's model, is that he he paid the debt off before it consumed all of GDP. But if we just tax each generation enough to cover the interest, while rolling over the principal, then it can last forever.

Yes, it sunk in afterwards as you can see in my subsequent reply to Nick.

@Nick

The government doesn't own any apples. If it wants to make a transfer of apples, it has to get those apples in a tax, or else borrow the apples. So the government borrows 10 apples from each A, and gives them straight back as a transfer. It is exactly as if the government just gave each A an IOU (bond) for 10 apples.

I see, so taxation is disappropriation of real goods whereas borrowing is also disappropriation but with a promise to repay (in real goods) + interest (also real) attached. That certainly fits with both the original meanings of the words and with your moneyless model. I'm not at all sure it still applies in the same way if we reintroduce money, though. How can you be comfortable with a model that allows one kind of financial claim (debt) but not another (money)?

I'll play through my own version quickly.

First off, there is no world in which everyone consumes 50 / 50 (or any other non 0 combination) without some kind of transfer because neither children nor pensioners typically produce / have income. But that's a side issue.

Now, if we add money to the equation, your example changes in the following ways:

Before you can ask what effect a reduction in income (through taxation) or an expansion of financial wealth commensurate with an expansion of financial debt (borrowing) will have on real consumption of both A and B, you need more information about the motives of all parties for engaging in these transactions.

It is, and you will recognise this as the typical MMT story, perfectly plausible to assume that those who buy into government debt are not, as the classical story implies, being enticed to forgo present consumption through an offer of attractive interest rates. Rather, it makes sense to assume that they are actively seeking to save anyway because they know whey will run out of income at some point, and will thus buy into more or less any offer made. So, in the counterfactual world with neither borrowing nor taxation, the money would not have been spent by young B, it would have been saved!

Reality may be a mixture of the classic model and the above, but my observation of the world around me, especially wrt interest rates and saving behaviour, certainly adds credibility to the latter interpretation. And, looking at economics from the outside for a moment, I would say it comes down to an anthropological and political difference in approach. The classic world is full of incompetent and cowardly government officials, heroic, benevolent capitalists and a proletariat in dire need of guidance. The Keynesian world is much more modern and optimistic in its trust in democratic institutions and in the equality and maturity of all members of society. Again, reality may, sadly, be somewhere in between but I very much want to believe in the superiority of the latter approach as a way forward.

RSJ has two good posts on the topic real vs. nominal for those who are interested:

http://windyanabasis.wordpress.com/2011/03/31/nominal-versus-real-part-1/

http://windyanabasis.wordpress.com/2011/04/07/monetary-versus-real-part-2-why-monetary-exchange/

Ad4mk4: your comment got caught in our spam filter. I just retrieved it. This happens from time to time. Sorry.

Steve Roth: "If the eventual taxing never happens, the old/young consumption pattern continues throughout eternity. 'zat right?"

Correct.

If the interest rate is bigger than the growth rate, the wave gets bigger and bigger (relative to GDP), as interest accumulates on the debt, until we get to a cohort that consumes {0,100}, at which point it is physically impossible for the wave to continue, so the government must increase taxes or default.

If the interest rate is less than the growth rate, then the wave gets smaller and smaller relative to GDP, and asymptotically dies away, so there's no burden (in fact there can be a benefit to future generations).

JKH:
I agree with your comment regarding MMT's take. The funny thing in all of this is that Professor Rowe considers that he has been able to show that under some very specific circumstances, there is a *possibility* that past budget deficit will be a burden to a future generation to the extent that this future generation is taxed (MMTers like me would note that this taxation is only necessary to the extent that inflation is a problem).

Concomitantly, some MMTers like me are are happy with the fact that Professor Rowe has recognised that a budget surplus is a cost to the future generation (the one that is taxed) since the budget surplus results in a confiscation of net financial assets [from] the non-government sector. (MMTers would note that this confiscation was not necessary if inflation was not a problem).

[typo fixed NR]

Jos: Is there a typo in your above comment? Did you mean: "...since the budget surplus results in a confiscation of net financial assets [from] the non-government sector."

If it was a typo, and it's OK with you, I would like to edit your above comment to fix it. Because it is a useful clear comment for me to build on.

Nick Rowe:
Correct. Please go ahead.

Oh dear: I'm feeling so clear then I come up with what seems like a crucial question:

Does this whole model apply to privately issued debt as well?

It seems like the answer would depend on how you think about money creation -- by treasury through deficit spending, by the Fed through reserve issuance, and/or by fed-licensed (and shadow?) banks through lending.

If Robert Waldmann fails too, that would be very bad for the credibility of economics blogging.

I just find this statement...odd, reading it a second time. Why would what appears to be a marginal issue (a counterexample to something that if made as a general law might be wrong, but holds true most of the time in the real word) be "very bad for the credibility of economics blogging", when there are far weightier reasons for questioning the credibility of economics blogging? Is it "very bad" in the sense of a straw that might break the camel's back, as the saying goes?

Jos,

That's exactly right (with typo corrected).

And MMT'ers typically don't worry about the generational distribution of the potential tax; they just don't like surpluses period.

And again, the usual MMT point regarding generational burden of the debt has nothing to do with the question of tax incidence - its about the irrelevance of the solvency issue.

Oliver,

In Rowe's model and the real world, the past cannot be changed. You made the following comment earlier:

"Krugman is right in the sense that there is nothing that the initial borrowing sets in motion that cannot be rectified - assuming time doesn't end suddenly."

This is exactly the point Krugman was making, but it is exactly where he and Baker were flat out wrong. The future distribution of lifetime consumption cannot be rectified because the bondholders in that future have already underconsumed their production in order to buy the bonds which allowed the bond sellers to consume more than they produced. Anyone in the past who has already consumed more than they produced and died cannot be made worse off to rectify this inbalance. Thus, to make future bond-buyers whole or less worse off by the tax requires their contemporaries be made worse off than they would have been otherwise, but those are also our future generations, so Krugman was completely wrong- a burden must be recognized by someone. All that can be done in the future is to spread that burden out.

Jos: "Concomitantly, some MMTers like me are are happy with the fact that Professor Rowe has recognised that a budget surplus is a cost to the future generation (the one that is taxed) since the budget surplus results in a confiscation of net financial assets [from] the non-government sector. (MMTers would note that this confiscation was not necessary if inflation was not a problem)."

OK. Let's start with this. Here's my take on the rest of the story.

1. Suppose we have deficient demand unemployment (with no fear of inflation). OK, there's a free lunch here for monetary policy, so let's eat it. Loosen monetary policy/cut interest rates to increase demand. (No burden here on future generations).

2. Suppose we are at full-employment, and the rate of interest is permanently below the growth rate of the economy. There's a free lunch here for fiscal policy, so let's eat it. This is the Samuelson 1958 model. Cut taxes and/or increase government spending, then let the debt rollover forever without raising taxes. As long as the rate of interest is permanently below the growth rate, you can run a sustainable Ponzi scheme and make all generations (current and future) better off.

3. (I am deliberately leaving aside the case where the interest rate is temporarily lower than the growth rate, but may be higher than the growth rate in future. It's too complicated for me.)

4. Now assume the interest rate is equal to or greater than the growth rate. And that we start at full employment, with a balanced budget. If we cut taxes on the current generation, they will want to spend more, and the central bank will need to raise the interest rate to prevent inflation accelerating. So, even if the rate of interest was initially equal to the growth rate, it is now above the growth rate. If the government does not increase taxes to at least pay the interest on the extra debt (and run a primary surplus) the debt will grow at the rate of interest, and so the debt/GDP ratio will be increasing over time. In order to rollover the debt, and persuade the young to save an ever-increasing share of their income to buy the bonds off the old, the central bank will need to keep on increasing the interest rate still further to prevent accelerating inflation. This is clearly unsustainable. Therefore the government eventually will be forced to increase taxes (or cut spending) on some future generation to at least pay the interest on the debt.

Those taxes are a burden on that future generation. Individually, they choose to buy bonds, because they want the interest. But collectively, that interest comes out of their own pockets, when they pay taxes.

Loosen monetary policy/cut interest rates to increase demand. (No burden here on future generations).

This (credit expansion) is where MMT does see a potential burden on future generations because of the vagaries of private sector indebtedness.

"This (credit expansion) is where MMT does see a potential burden on future generations because of the vagaries of private sector indebtedness."

That's interesting, but I am not sure if MMTers would 100% agree. After all, Mosler advocates a permanent 0% interest rate (the MMT "natural rate"), and so doesn't see low interest rates generating an unstable debt dynamic. I think they're more likely to think unstable debt dynamics could be created if fiscal policy isn't supplying enough NFA, as is what they believe happened after the Clinton surpluses. As far as Nick's number 1), it doesn't make much sense in MMT paradigm, which would say you should use fiscal in that scenario, but I realize this is a gap we're not going to bridge any time soon :).

Well, I guess, you might be thinking - if you rely only on monetary policy, you are therefore foregoing fiscal policy, and because fiscal policy isn't kicking in, you risk unstable debt accumulation. But this might be happening independent of the interest rate and not necessarily encouraged by it.

Probably should say "doesn't see low interest rates generating an unstable debt dynamic everything else constant," but I'd want him to speak for himself.

Nick Rowe:
Your point #4 is precisely why I was questionning your assumption that bonds always sell at their nominal value between cohorts. If cohort 2 wants more attractive yield in order to be willing to buy government debt, then cohort 2 will bid down government bond price to the desired yield and cohort 1 will incur a capital loss on their bond holdings (selling it below nominal value to cohort 2). Cohort 2 does not have to save a greater portion of its income if it does not want to, all it needs to do is to bid down bond price.

Bond price (yields) matters.

You have to note something here- let's say in the future, at t2, all non-bondholders decide to boycott purchasing the bonds so that those present holders can't consume the proceeds, but must bequeath the bonds. That changes nothing but the distribution of the burden- it just moves the recognition of the burden back one step, but that one step back is still our descendants being burdened.

Mosler today on debt, link below.

He’s talking about no generational burden, in the sense of the Fed being able to repay debt with reserve balances.

This is the standard MMT interpretation of the issue.

No reference to taxes.

Taxes are the essence of the Rowe/Murphy model.

MMT doesn’t deny that taxing to repay bonds is a burden.

It maintains that bonds can always be repaid with non-tax proceeds (Fed reserves), and that this is not a burden.

And it denies that bonds must be repaid with tax proceeds.

But it allows that bonds may be repaid with taxes, should inflation ever become severe enough to warrant that sort of tax increase.

But the purpose of such a tax increase is not to repay bonds. That’s just a by product. The purpose is to curb aggregate demand and inflation pressures.

So says MMT.

http://www.cnbc.com/id/46148704

I'm a non-economist living in the UK, doomed to suffer our government's austerity measures. Help.

Anyway, what a fascinating set of posts. Just as interesting as the topic itself is the fact that Nobel prize winners, economics professors and the rest can argue about, or cause debate about, this question for weeks. I don't know what the answer is. It's not easy. To those who say,“I don't know why this is so hard for people to accept” -- well, that's because the conclusions drawn here by Nick are hard to accept.

I understand that Nick's apple economy as originally described in his post of 28th December is an economy where the burden of government apple debt is passed to future generations. It's a B economy.

I have one question. Is Nick right when he says: "The only case where it does not create a burden on future generations is where Ricardian Equivalence holds?" Is that the only case?

I have an apple economy almost exactly the same as Nick's, except the government has an Infinite Apple Generator Mark 2.0 (IAG2.0). Made by the MMT organisation, the IAG2.0 is a machine which can create any number of apples required at any time. (Note that the IAG1.0 didn't work properly -- it created 42 apples then stopped permanently).

Amazingly, just as in Nick's economy the government borrowed 100 apples from each of a cohort A, then gave each person in cohort A a transfer payment of 100 apples. Unfortunately the IAG2.0 was not up and running at this time. So, just as Nick says: "it is exactly as if the government had simply given each person in cohort A an IOU for 100 apples. That IOU is a bond."

A few years later the, now working, IAG2.0 created the 121 apples (ie, 100 apples plus interest) to buy back the bonds from Cohort C. No taxes were levied on Cohort C. It's an economy where the burden of government apple debt is not passed to future generations. It's an NB economy.

By the way the person who worked out why the IAG1.0 didn't work -- and who created the IAG2.0 that did work -- that person was part of Cohort A and only did the amazing work they did after a moment of inspiration caused by eating an additional 100 apples. Each member of Cohort C eats no fewer apples than they would have done without the initial apple loan to Cohort A. In fact thanks to the IAG2.0 there is now an unlimited supply of apples. Cohort C is grateful, but is now getting a bit fed up with all the apples.

Jos: when I say "interest rate", you should perhaps interpret that as "yield". (There's a terminological difference here, that JKH has mentioned to me in the past).

Yancey Ward,

(I think.) Future generations are only necessarily worse off if you assume Robert Murphy's utility function. (In the apple model.) A debt of 100 apples or less could be rolled over indefinitely.

Nick,

On point #4, you have the government cut taxes at full employment, then not raise taxes when rates rise. If they do raise taxes, couldn't they (roughly) keep interest equal to inflation over time. Or do interest rates have to always be above inflation rates to prevent accelerating inflation?

(Dean Baker talks about the Fed holding assets and collecting interest. He and MMT talk about interest on reserves to control inflation. If MMT is right and the government does not need to sell bonds to spend, (I think they say the government could basically just provide bonds as a guaranteed savings accounts), could they then effectively keep interest and inflation balanced over time? Could they deficit spend at zero interest and keep interest on reserves high at the same time? Or balance interest paid to the Fed and interest paid to bond holders?)

Nick,

I've done my bit, and have failed.

From my point of view, you succeeded. I learned a great deal by carefully thinking about your apples example. Also, I appreciate very much that you graciously answered a question I asked when I was obviously confused.

I find this issue fascinating because of its importance for public policy, and because the arguments on both sides, even the ones that are wrong, are so plausible (at least to someone like me with no particular training in economics).

@The two people who have explicitly criticized me for saying “I don't know why this is so hard for people to accept...

Let me be clear: I'm not talking about the full-blown, general analysis and all the different permutations etc. Yes, that's a fiendishly difficult thing, and even Nick and I are disagreeing on minor points here and there, or are saying, "I'm not sure about that little issue, but what I *can* say is blah blah blah..."

But that's not what I was talking about. When I complained, "I don't know why this is so hard for people to accept," the "this" was the fact that Krugman and Baker were using bad intuition when they were telling their readers it was "especially nonsensical" to worry about the debt. They were flat-out wrong, as even Baker implicitly admits at this point.

Look, suppose Dean Baker said, "I don't believe in unicorns, so it's nonsense to worry about an asteroid destroying the planet." Can we all agree that would be a silly thing to say? And yet, to carry through the analogy, if Nick Rowe objects to Baker saying that, then people would take 100+ comments to ask Nick, "Are you saying you *believe* in unicorns??" and "Let's be fair to Dean Baker, I think he had in mind the data on asteroid trajectories, and most scientists think we're in no immediate danger...."

Baker and Krugman said "they owe it to themselves so it's not a burden" and I think everybody here agrees that *that* is a bad argument. Period, end of story on that particular issue. So Baker and Krugman were wrong, since that was the foundation of their argument. It's why they kept talking about foreigners holding the debt, which would be a totally superfluous comment if they were saying the valid things that some people here are arguing that Baker and Krugman may have had in mind. (For example, if no taxes are ever levied because g > r, then it doesn't matter if Americans or Chinese hold the bonds. So clearly Krugman wasn't relying on g > r for his statements to go through.)

Bob,

According to Baker and Krugman China keeps their currency artificially low. And, even if they didn't Chinese and American citizens are not free to switch places. And even if they were, there are trade offs. So, Baker can think a trade deficit is a problem, but think national debt is not a problem.


Anybody who is taxed for the purpose of repaying debt is bearing a burden. And it doesn’t matter who holds the bonds or how they got them.

If the taxpayer doesn’t hold the bonds, he’s worse off than not being taxed.

If the taxpayer does hold the bonds, he’s worse off than not being taxed. And that applies whether he purchased or inherited the bonds. If he purchased the bonds, and he pays the tax with his bonds, he’s lost the principal value (paid away in apples earlier) that he can no longer recover by selling the bond in the future. If he inherited the bonds, and he pays the tax with his bonds, he’s lost the principal value that he can no longer gain (apples) by selling the bond in the future.

Whoever the taxpayer, he suffers a marginal hit to his net wealth because of the tax, regardless of wherever the bonds are held.

By extension, it doesn’t matter whether the bonds are domestic or foreign held. The domestic taxpayer is worse off either way. The foreign taxpayer is unhurt. And whether either of them holds the bonds or not is irrelevant.

More generally, all domestic taxpayers are hurt, whether they hold bonds or not. And all foreign taxpayers are unhurt, whether they hold bonds or not.

The burden is the tax to pay interest and/or repay the debt, not the debt per se. From there, it only remains to show under what conditions a tax is likely, etc.

You can’t make an unconditional statement about a future debt burden without assuming something conditional about expected future taxes. That’s where you get into (g, r) arguments and the rest.

The debt burden is the tax burden. And the tax burden itself is always contingent. So the future debt burden, generational or otherwise, is always contingent on taxes.

The relative advantage of having inherited bonds is a marginal advantage, regardless of taxation.

Conversely, the relative disadvantage of being taxed to service debt is a marginal disadvantage, regardless of the state of holding bonds by purchase, holding by inheritance, or not holding at all.

Any generational argument can only apply to a defined cohort that happens to be taxed to service debt. If the “current generation” is taxed, but a “prior” one isn’t, the current generation bears the burden exclusively. It is always worse off at the margin, because of the tax, whether or not it holds bonds, or how it got them.

KRUGMAN loses on the domestic/foreign issue on all counts. The issue of “debt burden” has to do with the nature of the debt as a liability. It has nothing to do with who actually holds the debt, as noted above. This is easy to demonstrate in the case of foreign held debt. Krugman et al maintain that foreign held debt is a special case. Well, suppose it is. For example, China holds tons of US Treasuries. Now, suppose China swaps all of its Treasuries for US corporate bonds and equities tomorrow. What happens? Well, according to the Krugman criterion, that debt can no longer be a burden, because “we owe it to ourselves”. That’s nonsense. Absolutely nothing has changed with regard to the mechanics of any of the interest payments, principal repayments, or rollover of that government debt. How can the nature of the liability have changed simply because the debt has been sold back to the US? Whatever the interpretation of the international exposure, it’s not the particular debt instrument that’s a specific burden, since instruments can be switched on a dime. The international exposure is the same, whatever the instrument that fills the funding. Moreover, international exposure is a hodgepodge of gross and net exposures, much larger than corresponding current account imbalances themselves. Much of the alleged “burden” could be interpreted just as easily as asset swaps as net funding plugs. E.g. the bilateral US Chinese cumulative current account imbalance is much smaller than China’s holdings of US Treasuries.

But KRUGMAN et al are wrong on this in the more fundamental sense of the lack of connection to the US taxation threat. As noted, there is no marginal burden due to debt per se, wherever it is held. And the marginal nature of the actual tax burden doesn’t depend on who holds the bonds. Those who hold bonds, whether domestic or foreign, are never worse off because they hold the bonds. They get paid in apples, whether they’re domestic or foreign. It’s the domestic taxpayers who are always worse off, regardless of whether or not they hold the bonds, or how they came into holding the bonds if they do happen to hold them. And there is no taxpayer burden in the case of foreign held debt when there are no associated taxes, although the contingent gross exposure depending on future taxes does get larger.

Conclusion: servicing debt without a related tax is not a burden for any generation. The tax is the burden, not the debt. And it’s only a burden to the taxpayer.

(All of this holds in the context of Nick’s original model. And the MMT view of the issue is consistent with this. Nick, I now agree that taxes paid to service interest payments on bonds are a burden as well. So the tax to pay/repay either or both interest and principal is a burden. This is the only point I’ve changed my view on.)

I reckon we are all converging. Which is good.

Alan T Thanks!
"I find this issue fascinating because of its importance for public policy, and because the arguments on both sides, even the ones that are wrong, are so plausible (at least to someone like me with no particular training in economics)."

And are equally (or even more?) plausible to us with economics training!

anon: "On point #4, you have the government cut taxes at full employment, then not raise taxes when rates rise. If they do raise taxes, couldn't they (roughly) keep interest equal to inflation over time. Or do interest rates have to always be above inflation rates to prevent accelerating inflation?"

1. That's not what I said. It's the relation between interest rates and the growth rate (not inflation) that matters.

2. I already explained it it the comment.

Nick Rowe, JKH:
Ok, let's forget about the difference in interest and yields. Let's assume we have a super central bank that is able to set the interest so that quantity demanded for net financial assets in the non-government sector (i.e. government bonds) always match the quantity offered (so the Central Banks always adjust the interest so that Cohort(t+1) always find it appealing to buy bonds from Cohort(t)). And lets further assume that conditions as set out in point #4 of Nick Rowe's comment @ January 26, 2012 at 03:21 PM holds (in a nuttshell I accept ALL Nick Rowe's assumptions). Then this is what I think would be MMTers' take on this scenario:
MMTers: "this scenario demonstrates how bad a tool is monetary policy to control inflation even assuming that we have SUPER Central bank that is always able to set interest rate perfectly. By injecting ever more interest income into the private economy, the Central Bank keeps on increasing the size of the net financial assets of the non government sector, which means that ever increasing interest rate is necessary in the future to make sure that these "out of control" net financial assets do not become inflationnary. This is really a bad policy as the Central Bank IS BOTH THE SOURCE AND THE REMEDY TO THE PROBLEM! We need a definitive solution: tax (that is, confiscation of net financial assets from the non governmental sector through a budget surplus). And no need to wait at Cohort 5 to tax, as soon as government realises that interest rate needs to go higher to control inflation, government should tax! This would mean taxing Cohort 1 at the end of their life."

So the above is what I think would be MMTers' response (again, this is accepting all Nick Rowe's assumptions... but as Nick Rowe may have noticed, MMTers like me would first typically take a shot at his assumptions :-).

I also agree with JKH's last comment regarding the role of taxation although I prefer to say in typical MMT fashion that "net financial assets is not the burden" rather than government debt is not the burden. :-)

"its importance for public policy"

But IS it important for public policy? Everyone so far seems to have agreed that it is possible and even likely to have debt overmatched by the gains from the investment...

Jos,

I think MMT might be more closely in synch with Nick's approach than appears on the surface.

Nick describes taxation as a requirement to pay interest on the bonds, in the face of unsustainable debt math, where the interest rate exceeds the growth rate. Taxes are necessary to avoid default due to unsustainable interest rate/growth math

The key is why this math relationship has emerged in the first place – inflation.

MMT would say that the CB can control the interest rate – through the short term policy rate plus expectations – but that in order to ensure that control in all environments, it needs to have a fiscal policy option available in its hip pocket, in the event inflation risk become sufficiently severe.

So MMT would explain such a tax as a firewall response to inflation risk rather than what is required to pay interest on the debt. Once inflation is tamed at the margin by fiscal policy, the debt math returns to a comfort zone, and the central bank can continue its interest rate control without fiscal policy.

I think this gets you to roughly the same place as Nick’s argument, but with a slightly different nuance on the reason behind the decision to tax.

The MMT “zero rate, zero bonds” policy option is an extreme case of this thinking. Rates are set permanently at zero, and the cumulative deficit is captured within the banking system balance sheet. That puts all the onus on fiscal policy to conduct what in essence is an indistinguishable fusion of fiscal and monetary policy. (Whether or not such a theory could ever work in practice is a different question; I think Nick believes strongly that it’s impossible.)

JKH: "(Whether or not such a theory could ever work in practice is a different question; I think Nick believes strongly that it’s impossible.)"

A zero real interest rate would, if a permanent policy, almost certainly require a very tight fiscal policy for a very long time. I think the national debt would have to go strongly negative. And it would almost certainly be dynamically inefficient, because it would mean the real interest rate would be less than the real growth rate. I would then bring in Samuelson 58 to argue for a looser fiscal policy and higher real interest rate.

A zero nominal interest rate is possible. Milton Friedman argued for it in "The Optimum Quantity of Money". But as I argued in that old post about a year back: it's a recipe for communism. The central bank would have to own all the assets in the economy to make it feasible.

Mandos: Here's anothwer way of thinking about it: I've forgotten what it's called, but there's a way of doing the accounting that subtracts government assets (including real assets) from the government debt.

Nick: "That's not what I said. It's the relation between interest rates and the growth rate (not inflation) that matters.

I understand that. I'm asking when g is less than r can the government or central bank offset it. Could they do write downs? What about the Fed holding assets?

"I already explained it it the comment."

I'll assume you meant g must be less than r in the long run. Then aren't cyclical burdens inevitable? (Are these similar to Minsky moments?) Are private sector interest payments a burden? Is there a difference between a government debt burden and a balance sheet recession?

JKH: It's not the principle that's a burden, it's the interest. Tax-financed transfers are "burdensome," too.

"I think the national debt would have to go strongly negative".

Unless you mean by this that the national debt would have to decrease substantially, isn't this mathematically impossible?…

bigmac: the government could be a lender, not a borrower. It pays off all the debt, and then starts buying private bonds.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad