Summary. Nick Rowe’s theory relies on an increase in private sector net financial assets (psnfa) during and after the initial bout of government spending. That increase in psnfa takes the form of bonds. However, assuming the economy is at capacity throughout the process, and assuming zero growth, psnfa cannot increase, else aggregate demand would rise, which is incompatible with the “at capacity” assumption. I.e. the very existence of bonds (or extra psnfa in any form) after the initial bout of government spending has finished just isn’t allowable.
Nick Rowe produced a theory about government debt involving a burden that is passed to future generations a few months ago. And the argument over this theory flared up again in the last week or so. E.g. see here or here. And many of us (me included) have been reduced to near nervous breakdown in the process. Even Nick is near the end of his tether: in the comments section here, he said, “I'm getting old, and my brain is going.”
Readers acquainted with Nick’s theory can skip to the section below headed “The flaw”.
The Rowe Theory.
Seriously unsophisticated folk think that government debt is a burden on future generations.
Those with a little grasp of economics think that no burden is passed on for the following reasons. While a portion of those in future generations have imposed on them the obligation to pay taxes to fund interest on national debt and the obligation to pay more taxes to repay those holding government debt, there are also those who inherit government debt holdings and the right to RECEIVE interest. So the two arguably cancel out.
Continuing on up the “sophistication” scales, there is Nick Rowe who recently produced a new theory. Incidentally, the following is MY DESCRIPTION of his theory. If you want the theory in his words, see here or here and the various links he gives to earlier expositions of his ideas.
Assume to keep things simple that government borrows to spend on CURRENT EXPENDITURE items rather than capital items. Or put another way, if government IS SPENDING on capital items, let’s consider just the effects of a given tranche of borrowing to fund current spending.
So in period 1, government borrows and spends. Pensioners (the first cohort) tend to sell assets to fund their retirement, so they will tend not to buy the newly available bonds. In contrast, it will tend to be those of working age (the second cohort) who buy the newly available bonds. And that means pensioners benefit from the extra government spending while those of working age lose out. I.e. it’s those of working age who sacrifice personal or household spending in order to make room for the extra government spending.
In period 2, the above working age lot become pensioners and in their turn sell their bonds, which of necessity are bought by a new lot of working age people (the third cohort). Again, pensioners benefit and working age people lose out.
Next, you can stipulate as many periods as you like, but eventually there is a generation or cohort which buys bonds but does not benefit because it cannot sell bonds in its old age. So that generation loses out. As to the intermediate generations, they neither lose nor benefit since they both buy and sell bonds.
So in effect, a burden appears to be passed to a future generation: that is, the initial lot of pensioners gained and the final lot of pre-retirement people lost out.
Let’s make three assumptions in order to keep things simple. First let’s assume that the economy is a capacity throughout. Second, I’ll make the perfectly reasonable assumption that aggregate demand is related to private sector net financial assets (psnfa). (The latter assumption is actually just another way of saying that Keynsian paradox of thrift can be a problem.) Third, I’ll assume no growth and zero inflation.
When government borrows $X and spends it back into the economy and gives $X of bonds to sundry private sector lenders, psnfa rises by $X. That is one reason why Keynsian “borrow and spend” works. Another reason is that the latter borrow and spend policy takes money away from the cash rich (those with a low propensity to spend) and channels it into the pockets of those with a higher propensity. I.e. the latter effect augments the psnfa effect.
That being the case I won’t refer to the “propensity” point again: I’ll just assume it’s included in the psnfa effect.
A diversion – negative bonds.
Incidentally, it could be argued that when government bonds are given to the private sector (which is what is effectively involved here), negative bonds are also dished out at the same time. That is, if one lot of people gain the right to receive interest and eventual payment of a capital sum, there must be another lot with the obligation to pay taxes to fund that interest, and repayment of the capital sum. I.e. it could be argued that for that reason, psnfa does NOT INCREASE.
However, what is important here is PERCEPTIONS. That is if the above first lot feel richer, while the second lot don’t realise they’ve been had (to put it crudely), then aggregate demand will rise. And it’s the effects on AD that are all important. So to keep things simple, I’ll assume the “negative bonds” don’t exist. But if you want to claim that the net worth of “positive bonds” are reduced by some proportion as a result of the “negative bond” effect, that doesn’t make any difference to the basic argument here.
By the way, after a little Googling, I can’t find any reference to “negative bonds”, so that makes that an original idea (Nobel Prize, please).
Back to the argument.
Anyway, after that little diversion and to pick up the thread of the argument again, when government borrows and spends, there is a stimulatory effect, but we DON’T WANT that here, because we’ve assumed the economy is at capacity throughout. So government would have to counter the psnfa effect somehow. I’ll assume for the moment that it does this by borrowing MORE THAN $X: let’s say $X+. That way, the private sector is starved of cash, which makes room for the extra government spending.
However, the latter borrowing is only needed WHILE the extra government spending takes place.
That is, when the extra government spending is finished there is a problem: the private sector is still cash starved. And cannot be allowed, else aggregate demand would fall. So government really needs to buy back the bonds. I.e. those bonds that get shuffled between cohorts under the Rowe theory just aren’t allowable.
But of course in the REAL WORLD governments have a very defective grasp of the nature of deficits, national debts etc. (There is an article by John Cochrane on this point.)
So what would happen in the REAL WORLD given the above bout of government spending is that when the spending finished, government and central bank would find to their amazement that the economy was stalling. Worthy articles would appear in the Financial Times or Wall Street Journal purporting to explain the mystery. Actually the articles wouldn’t explain the phenomenon at all.
To summarise, on the assumptions made above (economy at capacity, etc) if the economy is at capacity when psnfa equals $Y BEFORE the bout of government spending, then all else equal, psnfa will need to be $Y afterwards. Plus the make up that psnfa will need to be the same (ratio of cash to bonds in particular).
And now I’m off to see my psychiatrist.
Also I should have said something about the Ricardian effect as this as it looms large in Nick’s argument. The empirical evidence, far as I can see, is that there certainly are Ricardian effects, but they are not of major significance (thought that varies a fair bit depending on the scenario – tax cuts, pay increases, etc etc. ). So let’s just say the simplifying assumption above is: no Ricardian effect.