01 Jun 2015

An Unexpected Twist in the “Debt Burdening Our Grandkids?” Debate

Debt, Economics, Shameless Self-Promotion 41 Comments

You guys think you’ve got me all figured out. Well you’re wrong. In my current EconLib essay, instead of focusing on Krugman, I design two thought experiments of war financing to show that even Ludwig von Mises missed something crucial in the “debt and future generations” argument. The intro:

Critics of government debt often argue that deficits are irresponsible and cowardly because the present generation is foisting the bill onto future generations, many of whom have no say in the political decision. This perspective resonates with the man on the street, but many professional economists—such as Paul Krugman and Dean Baker—believe that such thinking is completely fallacious because it confuses an individual household with the economy as a whole.1 According to these economists, today’s government spending is “paid for” by the present generation, period, and any talk of burdening our grandchildren is nonsense.

However, some economists—notably James Buchanan and those following his lead—have challenged this dismissal.2 They argue that when we account for the fact that generations overlap, there really is a legitimate sense in which government debt allows people today to enjoy higher government spending that is partially paid for by reducing the standard of living of taxpayers who have not even been born. In this article, I lay out the respective positions with a numerical example involving a hypothetical war. I use the example of war because that has a been a traditional cause of government debt. But the same analysis applies whether the government debt is to finance war or to finance anything else that government spends money on. I also show that this dispute is not simply an ideological one: I contrast Buchanan’s view with that of Ludwig von Mises, a champion of the free market and no apologist for government budget deficits. Even Mises, in his writings about financing war, overlooked the subtleties that Buchanan (decades later) would emphasize.

Incidentally, this is a very subtle matter. Krugman, Dean Baker, and (I believe) Abba Lerner all relied on the “we owe it to ourselves” notion in order to show that the public shouldn’t worry about big government deficits, and certainly shouldn’t view them as analogous to household or corporate debt.

Now, when I say that I used to subscribe to a version of this view, and so did Mises, of course I am just talking about the very narrow issue of whether the existence of a higher government debt, with the required taxation to pay interest, can directly burden future generations. I used to think this was impossible, that such payments were merely a transfer within the future generations. But after the great debate a few years ago, I now see that there is something else we have to keep in mind.

By bringing Mises into this debate, I am trying to defuse the ideological reactions. I want to show that there is a technical confusion here, beyond the fallout for policy.

In conclusion, and to avoid any confusion: Mises explicitly rejected the “we owe it to ourselves” argument for the BENIGNITY of deficits, and so did I back in 2006. But my point with this EconLib article is to show that even Mises missed the “overlapping generations” point that Buchanan (and then Nick Rowe and Don Boudreaux) nailed.

41 Responses to “An Unexpected Twist in the “Debt Burdening Our Grandkids?” Debate”

  1. Gene Callahan says:

    I think you mean to say, “Even Mises saw the irrelevancy of the overlapping generations point that others mistook for relevant.” 🙂

    • skylien says:

      How would it be possible to sell any government debt if generations would not overlap?

      You should ask yourself why people are not willing to make the transfers with outright taxes, but are ok with financing it by buying government debt. If you understand that, you understand why it is relevant. Either you need to argue that people are just delusional that they are only ok with government debt because it is completely the same for them with government debt or usual taxes, or you are wrong.

      • Gene Callahan says:

        “How would it be possible to sell any government debt if generations would not overlap?”

        Say what? The government sells short term T-Bills all the time. Do you think 91-day t-bills are be repaid by the next generation?

        “You should ask yourself why people are not willing to make the transfers with outright taxes, but are ok with financing it by buying government debt.”

        Because then the financing is done by people who are *willing* to give up consumption today!

        • guest says:

          They are indeed willing to give up consumption today.

          What’s missing, though, is that it’s being repaid by those who *were’nt* willing to do so – through taxes.

          It’s not like the government takes a bond-buyer’s money and tries to invest it in profit-generating businesses – a free market has those opportunities, and no government would be required for such investments.

          Words such as “we owe it to ourselves”, or “paid for by the current generation” simply hide the fact that the money is being taken from some individuals within the current generation, and given to other individuals within the current generation.

          It’s a superfluous aggregate. Go figure.

        • skylien says:

          „Say what? The government sells short term T-Bills all the time. Do you think 91-day t-bills are be repaid by the next generation?“

          Yes they are because they are rolled over every time, or aren’t they? With no overlapping generations this wouldn’t work out. If people would resist taxation without overlapping generations then they would also resist short term bonds in such a case, because there was no way not to be taxed for them. Where is the difference between taxation and bonds if generations don’t overlap?

          “Because then the financing is done by people who are *willing* to give up consumption today!”

          Gene the keyword is financing! Do you understand the difference between paying and financing?

          Or do you also believe that it is banks who pay for your house if they loan you money so that you can buy a house? No you are paying for it, the bank finances it and is rewarded with your interest payments for this.

          Just as the generation today finances bonds and is subsequently REWARDED for this as well. Bonds are finally paid by people who will not be rewarded sometime in the future for anything. Just as it is you who pays for your house and not the bank! With the difference that you decided to do this, future generations didn’t.

          • skylien says:

            Average Joe on the street is not saying, omg we are burdening our children because they will have to finance (by giving a loan) all of this. No, average Joe says they will *pay* for all of this, and no one will deny that this generations is financing it.

            At the end someone has to pay for an expense. The person/generation/bank who finances it by giving a loan doesn’t per definition! So who *pays*, Gene?

  2. Gene Callahan says:

    I don’t know why the smiley appeared mid-comment: I put it at the end.

    • skylien says:

      It always pulls the smiley in front. So if you want to have it at the end, put it into a separate line.

      I guess this shows that you never used smileys on Bob’s blog so far.

      😉

      • skylien says:

        The new wink smiley rather looks like it got punched in the face..

  3. Tel says:

    If Krugman is right, and the debt is no real burden then he should be OK about the idea of future generations defaulting and not paying back that debt. After all, in was never a burden for them in the first place, and “we owe it to ourselves” so default is not really default at all.

    • Gene Callahan says:

      A default would impoverish certain members of later generations (bondholders) to the benefit of other members (taxpayers). It certainly doesn’t transfer any wealth to that later generation!

      • Grane Peer says:

        Right, it is a wash. Therefore future generation neither loses or gains by defaulting, We owe it to ourselves is we don’t owe anything because we already have it.

      • skylien says:

        And yet in the aggregate bondholders and taxpayers are worse off (consumption-wise) than if there wouldn’t have been any bonds in the first place. How is that possible? Look at Bob’s example. The numbers don’t lie. And no you cannot replicate the same scheme with taxes.

        That would be like saying you can replicate violent robberies with people who freely give the money to the robbers, and beat themselves in the face…

      • Tel says:

        I’m willing to make a little wager (reputation at stake only) that Krugman will happily advocate for more debt, but never advocates for a default. If you think it’s a wash then can you suggest some motive for Krugman’s one-sided view of this “wash” situation?

      • Major_Freedom says:

        Who ever argued wealth can travel through time?

  4. Gil says:

    The question is what kind of debt is it? In other words, is it a case of a gambler father who screws over the next generation with a large debt they can’t hope to pay or an investor father who leaves them a portfolio that although has debt leverage is not required to paid off as such but maintained and the assets far outweigh the liabilities?

  5. Harold says:

    I must be missing some subtle point.
    “If we go that route, then we are forced to say that the war expense in Year 1 was paid for entirely by the Capitalists in Year 1. Yet this leads to a contradiction.”

    I am sorry but I do not understand the contradiction you mention. Or rather, I don’t see why we are forced to say the war expense in year 1 was paid for entirely by the capitalists in year 1.

    “Back in War Scenario A… the Capitalists did not really shoulder the war effort entirely because they knew they would be paid back out of future taxes for the loans they advanced.”

    In scenario 2 the capitalists were paid back by selling the bonds, so they did not shoulder the burden here either. They knew they would be paid back either by taxes, or by selling the bonds, or by leaving the bonds to heirs in which case “the analysis of Krugman et al. sounds more reasonable.”

    As long as the first the first generation is paid off one way or another, why is there a contradiction?

    If the bonds are inherited by their heirs, then the young capitalists inherit the bonds. They in turn die and the youngest capitalists inherit, now worth $4350 billion. We are left with year 70 onward exactly the same as scenario 2. If Krugman’s analysis is reasonable here, why is it not if the various swaps go on in the intermediate generation?

    • Bob Murphy says:

      Harold, you’re saying that there’s no contradiction, because in both cases, the capitalists in Year 1 didn’t really pay for the war. OK fine, but then that means in scenario 2 that the people who really paid for the war are the grandchildren in year 70 onward. Krugman would say that’s impossible. People in year 1 can’t make people in year 70 pay for their expenditures.

      • Harold says:

        Yes, I see the problem now. I keep looking at this, and coming back to the start again. It remoinds me of the unexpected hanging. At the moment I think the crux is here:
        “Back in War Scenario A, it was crystal clear that the Misesian worldview—which I also think Krugman would have endorsed—said that the Capitalists did not really shoulder the war effort entirely because they knew they would be paid back out of future taxes for the loans they advanced.”

        However, the capitalist did actually pay for it at the time – they had to do without. Nothing short of time travel will alter that. They will be paid back sometime in the future by taxes on workers. In scenario 1, these workers were from the same generation, yet we still say they paid for a war that happened years earlier even without a time machine. Why should that be any different if the workers are further in the future? So realistically we must say that the capitalists shouldered all the war effort unless we have a time machine.

        I have a few worked scenarios, but no time to sort them out now.

        • Tel says:

          However, the capitalist did actually pay for it at the time – they had to do without.

          Which is why a default coming up in the future would sting them badly. Then suddenly they would really be paying for the whole lot.

      • Gene Callahan says:

        Kurgman AND Mises rightly point out that a war can only be fought with current resources and currently foregone consumption. We can’t fight in Iraq now using bullets from 2085.

        The rest is just spreadsheet tricks.

        • Tel says:

          The rest is just spreadsheet tricks.

          The normal way of saying it would be “The rest is promises” so for example if I promise you a roast chicken next Christmas, I don’t need to really have a chicken in order to promise that. I may or may not deliver come the day, but the promise is real enough, as an instrument of agreement between people.

          We could write that on a spreadsheet if you like, or on illuminated vellum, or on magnetic media in the mainframe at the central bank. How we write it seems kind of irrelevant to the principle of a promise.

          Are promises just “tricks” ?

          Good question, they do assist trade between people, so at times promises can be useful, but they can lead to trickery if (for example) I pretend to promise something I knowingly have no intention of delivering, or (as another example) I make a promise to myself on behalf of children not born yet,

          • guest says:

            “Are promises just “tricks” ?

            “Good question, they do assist trade between people, so at times promises can be useful, but they can lead to trickery …”

            Trickery leads to trade, too, if by trade you mean “activity” or “output”. Ponzi schemes lead to trade.

            All Acting Man needs to act is the *belief* that his act will result in the alieviation of some felt unease.

            The question then becomes: Are all trades profitable?

            Keynesians mistakenly believe so, which is why they think that an increase in money velocity necessarily equals economic growth.

            • Tel says:

              The wheels of industry are greased by blood, sweat and tears. The wheels of commerce are greased by purest bullshit.

              Or as Ezekiel would say, “Wheels Within Wheels”.

      • Grane Peer says:

        Dr. Murphy,

        I get that the war was effectively paid for, money exchanged for goods, transaction complete. But, there is one transaction that is not complete- the repayment of the bond.

        The bond is inter-generational. Effectively, the bond is an incomplete transaction reaching into the future for fulfillment. I think this means someone in year 70 is making someone else in year 70 pay for something from year 1.

        The bond and thus whoever is holding said bond remains constant from past generation to future generation. It doesn’t matter if future generation is year 2,30,or 100. To my thinking the bond holder is essentially the generational overlap.

        If the bond is paid out in real terms(meaning not paid for with an additional debt) then there can’t be a neutral effect on future generation because the bond holder represents the past generation.

        The only way I can see the debt remaining neutral from one generation to the next is if a) the past generation is exactly the same as the future generation and the bond is paid(not floated by new debt) or b) If future generation is different than past generation and the bond is floated or never realized.

  6. Julien Couvreur says:

    I think I got it on third reading. I’ll have to mull it over.

    It seems to me there is a simpler way in which I would argue debt is burdening future generations: in order for government to sell its bonds, the bonds must be more attractive than other market option. This may be in improved terms (good interest rate) or reduced risk (due to captive taxpayer base). Such a deal, where the capitalists consented, but the “borrowers” were only represented very indirectly by the political system) must ultimately come at the expense of people repaying the debt.
    The politicians have less incentive to negotiate a good deal than the capitalists. So the capitalists or their descendants are gaining at the expense of the general population in the second generation (or turn) of the scenario.

  7. Julien says:

    I think I got it on third reading. I’ll have to mull it over.

    It seems to me there is a simpler way in which I would argue debt is burdening future generations: in order for government to sell its bonds, the bonds must be more attractive than other market option. This may be in improved terms (good interest rate) or reduced risk (due to captive taxpayer base). Such a deal, where the capitalists consented, but the “borrowers” were only represented very indirectly by the political system) must ultimately come at the expense of people repaying the debt.
    The politicians have less incentive to negotiate a good deal than the capitalists. So the capitalists or their descendants are gaining at the expense of the general population in the second generation (or turn) of the scenario.

    [Apologies if this is a re-post, the first time didn’t work, so I’m trying my other mail address]

  8. Bill B says:

    And if the debt is sold to foreigners, isn’t it clear that a future generation will pay that debt?

  9. DMS says:

    I posted the following on EconLog as well:

    The confusion on this issue has always baffled me.

    There are two sides to a loan – the obligation to pay and the right to receipt. With government borrowing and the power of taxation, these two sides are not treated symmetrically across generations. The obligation to pay is BEQUEATHED down the generations (through the power of general taxation), while the right to receipt has been SOLD down the generations, transferring wealth from the future to the present.

    It can be shown very simply. A son lends his father money. Dad spends it all drinking himself to death. Dad dies, and the son inherits Dad’s balance sheet, including the obligation to pay the loan. The son now literally owes the money to himself, so according to Krugman, that means nothing has been transferred and there is no concern. But this is obviously wrong – the son is impoverished by the loan amount, just as if the money had been stolen.

    Perhaps Dad actually borrowed the money not to drink, but because his father had pulled the same trick on him. Since he didn’t want to be out the money, he borrowed from his son instead. And perhaps Dad’s son does this to his own son, i.e. Dad’s grandson, whom Dad may never even have lived to see. And so on.

    The parallel to government borrowing, i.e. kicking the can down the generations, is unmistakable. A future generation can indeed be impoverished by a prior generation’s profligacy, notwithstanding “owing the money to ourselves”. Of course, the prior generations might have invested the money well, allowing the later generations to also inherit assets as well as liabilities. And those assets may even exceed the liabilities. But that is not the argument made by the “we owe it to ourselves” crowd, and is in fact a separate issue.

    Personally, I think this confusion is at the heart of Keynesianism. Hayek said that “Mr. Keynes’ aggregates conceal the most fundamental mechanisms of change” – I think this is why Krugman et al are so blatantly in error and yet cannot grasp it.

    • Nick Rowe says:

      DMS’s thought-experiment is very clear and right on the nail.

      The problem is not so much aggregation, but 2 different ways of aggregating:
      http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/02/teaching-olg-models-and-the-phenomenology-of-perception.html

      • DMS says:

        Exactly. Your post shows the need to decompose through time as well as through space, i.e. the “owe it to ourselves crowd” cannot grasp different constituencies over time (vertically vs. horizontally in your jargon). Still, I was wondering why Krugman gets stuck, and I think he fixates on aggregate GDP not sloshing through time, so therefore claims there can be no generational wealth transfer. He is of course wrong.

        So I googled “Paul Krugman debt we owe to ourselves” and it is shocking. He not only doesn’t get it, but he keeps doubling down on his misconception. He must be stopped, but I fear your example is too complicated for people to get over Krugman’s simple thinking. I think that’s true for Bob’s work as well, as good as it is. I think it may be because people get confused by bond finance. How about this example instead (without bonds and also using services instead of goods):

        The government institutes a new program where the young give backrubs to the old. Overall economic output is unchanged (say GDP), as increased elder consumption (the backrub) is offset by decreased youth consumption (foregone leisure, foregone work output, etc.) Current senior citizens clearly are benefitted. The young are repaid when they become senior citizens (perhaps with two back rubs as interest, but that’s beside the point), so they are not disenfranchised. As the program unfolds through the generations, GDP remains unaffected. One day, the program is cancelled. This benefits current youth, who are relieved of their backrubbing obligations, but hurts current seniors who do not get the previously promised service, though they had to provide it when they were young. But again, no affect to GDP, as the benefits and costs offset.

        So macroeconomic aggregates aren’t touched at all throughout this example, but there has clearly and demonstrably been intergenerational wealth transfer, from today’s current batch of seniors back THROUGH TIME to the original batch of seniors in place when the program was initiated. Just as in your beer model.

        Krugman’s blinkered focus on unchanging GDP makes him fail to see the wealth transfer; in fact he outright denies that the latter is true. At least I think that is what is going on in his thinking.

        By the way, it is quite easy to map this backrub economic model onto a bond-financed government spending model, i.e. it may seem silly, but it can be morphed into the very situation Krugman claims cannot exist. After all, this is what bonds are for – to allow for increased current consumption in exchange for decreased future consumption. In a government debt and taxation model, government borrowing is BY DEFINITION a transfer of wealth from future generations to current ones, or an imposition of future debt on our unborn grandchildren (again, putting aside the possibility we also give them assets as well). Krugman is right that GDP is unaffected in this transfer, but that is a non-sequitur.

        • Grane Peer says:

          DMS,

          Krugman is clear that he is talking about the economy as a whole. So, when he says we owe it to ourselves that must mean the whole economy owes it to the whole economy. One instance of the whole economy can’t owe anything to the same instance of the economy. It should read; one instance of the whole economy owes it to another instance of the whole economy. More simply put;

          The future owes it to the past.
          🙂

      • Gil says:

        You’re still assuming the government’s frittering the away money like a problem gambler.

  10. Silas Barta says:

    It should always have been clear why capital consumption hurts later generation; I don’t know how we got to the point that this has to be proven with increasingly complex numerical examples.

    That is, if I burn a wrench for heat, society gets a benefit now (a few people heated) but then has to work harder to build the same things later because they have to do it without wrenches.

    No one would think to say, “That’s just a *distribution issue*. Sure, some people have to get their work done without that wrench now, but other people benefit: the workers that now get paid to do the work that the wrench would have obviated. No *net* loss.”

    Likewise, when you consume resources as a result of borrowing, you have a lesser stock to work with later.

    Is there a major part of the debate I missed that addressed this?

    • Bob Murphy says:

      Silas, everybody was conceding that if the present generation consumes capital and thus bequeaths fewer machines (or less oil) to their grandkids, then the grandkids are poorer.

      However, even if there is a pure endowment economy with no physical saving even being possible–e.g. 200 apples harvested every period, no matter what, and they rot if not eaten right away–then we can get the OLG stuff. That was why Nick (and then I) focused on those simplistic endowment economies of apples back in the day, to isolate that this was a completely separate point.

  11. Yancey Ward says:

    Skylien and Tel both make the closing argument here (though DMS’s two comments help in the set up)- if Krugman is right, then he should recognize there are no problems with replacing all government borrowing with taxes today, and that any debt defaulted on now or in the future is a meaningless act, also. I seriously doubt Krugman agrees with either of those, which contradicts his statements on this matter. Just read what Callahan writes above.

  12. guest says:

    “According to these economists, today’s government spending is “paid for” by the present generation, period, and any talk of burdening our grandchildren is nonsense.”

    But returning to a gold coin standard today would somehow be barbarous to our grandchildren.

    (To be sure, it’s *supposed* to be barbarous to those businesses which were stimulated by the fraudulence of artificial credit creation.)

  13. Harold says:

    Still trying to work out the problem. I have a way to describe it that makes sense to me. Assuming no interest, and in year 70 the Govt taxes only workers to pay off the debt.

    1) Each generation dies and bequeaths the bonds. First capitalists never get their money back. In the final generation, the youngest workers are taxed to pay the youngest capitalists. First capitalists lose by $150 billion, youngest capitalists gain by $150 billion, youngest workers lose by $150 billion. Looking at it per generation, first generation capitalists lose by $150 billion, last generation is a wash. This is as Krugman says.

    2) First generation sells bonds to subsequent generations as per scenario B.
    First capitalists spend $150 billion, are $150 billion down. Sell bonds to young workers yr 30. First capitalists now all square, young workers are down by $150 billion.
    Next generation, young workers get $150 billion back from youngest capitalists. Young Workers all square, youngest capitalists down $150 billion. Then Govt. taxes youngest workers to pay youngest capitalists. Youngest workers down $150 billion, youngest capitalist all square.
    Looking at it by generation, first generation is a wash, second generation is a wash, third generation Youngest workers down $150 billon, youngest capitalists all square. Sure looks like the youngest workers have paid for the war.

    Now I understand the importance of the overlapping bit. If we look at time periods instead of generations we get a different picture.
    Time period 1: First capitalists down by $150 billion
    Time period 2: First capitalists paid $150 billion by young workers.
    Time period 3: Young workers paid $150 billion by youngest capitalists
    Time period 3: Youngest capitalists paid $150 billion by youngest workers (via tax).

    The first time period is down $150 billion, each subsequent time period is a wash.

    Because the generations overlap, the re-distribution which by definition is a wash in each time period, is able to redistribute between generations.

    In one sense this should not be a surprise, because even in Scenario A the re-distribution occurs at a later date than the war. There is no reason why this re-distribution has to stay within one actual generation of people.

    So the confusion occurs because we use generation to mean different things. If we mean a generation is everyone alive during a time period, we cannot redistribute between generations without a time machine. If you remain fixed in the generation of the time you were born, then we can re-distribute between generations because they overlap.

    This now seems to make perfect sense, so I suspect I have got it horribly wrong.

  14. Bala says:

    This bit from Gene bothered me and I wanted an answer that didn’t involve models like OLG. So I tried thinking it through and came up with this “explanation” to myself and would be glad to have it evaluated.

    Kurgman AND Mises rightly point out that a war can only be fought with current resources and currently foregone consumption. We can’t fight in Iraq now using bullets from 2085.
    The rest is just spreadsheet tricks.

    As per Gene, since only current resources can be drawn away from other uses, it is only the current generation that pays for the war and not future generations.

    However, that statement seems to contain the implicit assumption that it is just consumers’ goods that are being drawn into the war. That is not true. A war needs producers’ goods just as much as it needs consumers’ goods simply because it takes time to wage a war and war materials get consumed in waging the war.

    Clearly, if a line of production has to emerge to feed the war, this is clearly productive resources that are being drawn away from other production processes that would have churned out other consumers’ goods in the future. So, while present consumers’ goods are being diverted into the war, consumers’ goods that would otherwise have been produced in the future in a world that did not experience the war are being, in a very real sense, diverted. It is just that they are being diverted from the counterfactual world and we are not able to see them.

    But isn’t seeing the unseen that others fail to see the most distinguishing feature of the economist, especially of the Austrian School?

    • Harold says:

      Is this essentially the same argument as Silas Barter’s above?

      • Bala says:

        Thanks for pointing me in that direction. Looks like it is quite similar.

Leave a Reply