Government borrowing is near pointless

Government borrowing is near pointless

Ralph S. Musgrave

Abstract:

The currently larger than normal national debts in numerous countries

makes this a good time to look at the rationale behind government debt. Four

of the main reasons are considered here. The conclusion is that they range from

hopeless to unimpressive. First there are political reasons which can be

demolished in a few sentences. Second there is having government ‘borrow

and spend’ with a view to stimulus. The main flaw in this policy is that where a

government issues its own currency and borrows units of its currency, it is

borrowing something which it can create itself in limitless quantities: similar

to, and as pointless as a dairy farmer buying milk in a shop. In relation to

stimulus, a zero borrowing system was set out long ago by Karl Marx, Milton

Friedman, and others: it is called ‘Functional Finance’. This is a viable

alternative to borrow and spend. Moreover, quantitative easing amounts to a

move in the direction of functional finance and an admission of the weaknesses

in borrow and spend.

Third there is borrowing with a view to the purchase of assets, like

infrastructure investments. The arguments for borrowing here are passable.

Fourth there is borrowing to smooth out the erratic timing of government

expenditure and income from taxation. The arguments for borrowing here are

poor.

On balance, the arguments for government debt are not impressive.

Hopefully the arguments below will assist those trying to curb such debt.

Note:

Athens Institute for Education and Research (ATINER) July 2010 Conference,

except for the last sentence of the conclusion (i.e. last sentence of the paper).

This sentence is a more recent addition, and it has two references to works by

William Mitchell which have been added to the list of references.

Copyright 2010 by Ralph S Musgrave. Excerpts of up to a thousand words

may be reproduced without permission as long as the source of such excerpts is

made clear.

2

This version of this paper is identical to the version submitted to the

Introduction

Four of the more important reasons for government debt are examined here.

The first reason, or set of reasons, is political. Second there is having

government borrow and spend with a view to economic stimulus, which

occupies the bulk of the paper. Third there is borrowing to fund the purchase of

assets, e.g. infrastructure investments. Fourth, is borrowing with a view to

smoothing out the erratic timing of government expenditure and receipts from

tax.

The political reasons are dealt with in a few sentences below, while the

others are considered in more detail.

The second reason, borrowing with a view to stimulus, is written just with

countries that issue their own currencies in mind. Thus care must be taken in

applying the points here to the Eurozone as a whole or to individual Eurozone

countries. The other three reasons are applicable to both countries which issue

their own currencies and those which are part of a common currency system.

The word government is sometimes used below to refer to ‘government and

central bank combined’. In contrast it is sometimes used to refer to politicians

and the bureaucracy as distinct from the central bank. The change in use is

indicated where it is not obvious from the context.

The words stimulus and reflation are used as synonyms: they refer to raising

aggregate demand.

1. Political reasons for borrowing

Politicians tend to think they can win votes by borrowing rather than raising

taxes. Amongst other reasons this is because it is obvious to voters who to

blame when taxes rise but less obvious when interest rates rise as a result of

increased government debt. This reason for debt is clearly unjustified.

A second political motive with ostensibly more justification because it

seems to involve social justice is that politicians probably think that when

borrowing rises it is the cash rich who supply the funds. Actually the evidence

is that funds become available primarily because debtors and potential debtors

reduce the amount they borrow thus releasing funds for government to borrow

(Musgrave (2010:9-11). This is just another way of saying that when

governments borrow, crowding out occurs.

Objections to crowding out

Some readers may object to crowding out for the following reason. Many

governments have substantially increased their borrowing in the last year or

two, yet their interest rates have declined rather than increased.

The answer to this is that crowding out as interpreted here is based on the

‘other things being equal’ assumption. That is, if government borrows and

3

spends the relevant money but makes no other changes, an interest rate rise (or

a reduced availability of funds for private sector borrowers) is almost

inevitable: government out-bids the private sector for funds.

In contrast, over the last year or two, governments have both borrowed more

and flooded the market with liquidity. In this instance, other things are not

equal.

2. Borrow and spend with a view to stimulus

A conventional view is that governments should borrow and spend so as to

bring stimulus. In contrast, there is a set of ideas called ‘functional finance’

which objects to the borrow part of borrow and spend. ‘Modern monetary

theory’ is a more recent name for functional finance. (The word government is

used in the sense ‘government and central bank combined’ here.)

Karl Marx was the first to set out the basics of functional finance, though

there is nothing left wing about it. The next and much the most influential all

time advocate of functional finance was Abba Lerner (born: 1903, died: 1982).

Keynes said of Lerner, ‘Lerner's argument is impeccable, but heaven help

anyone who tries to put it across to the plain man at this stage of the evolution

of our ideas’. (See Landes (1994:220). Friedman (1948) also backed functional

finance, though he did not often use the phrase functional finance. Another

economics Nobel Laureate who advocated functional finance was William

Vickrey: see Colander (2003).

The basic idea behind functional finance is that the fact of government

spending exceeding income (or vice versa) is not important. (The word

government is used here in the sense ‘government and central bank combined’)

That is, there is no merit in government balancing the books just for the sake of

it.

Under functional finance, the only important considerations are the

‘functional’ effects of government income and expenditure, much the most

important being the effects on inflation and unemployment. That is,

governments should use their income and expenditure to bring the maximum

level of employment that is compatible with acceptable inflation. For example,

given high unemployment and subdued inflation, government spending should

exceed income. To that extent functional finance is similar to conventional

policies. But the big difference is that functional finance says governments

should not borrow to make up the difference between income and expenditure.

One reason for not borrowing is that when government borrows, it borrows

something (money) which government itself has created and which it can

create in limitless amounts. Thus for a government which issues its own

currency to borrow units of its currency is similar to, and as pointless as a dairy

farmer buying milk in a shop.

A second objection to borrowing is that borrowing involves crowding out.

Crowding out would not matter if there were agreement on the extent of the

problem. But there is a lack of agreement. Thus introducing crowding out

4

firstly introduces uncertainty. Secondly, if crowding out is a serious problem -

say 90% of borrow and spend is nullified by crowding out - the expansion in

the national debt for given stimulus will be about nine times the expansion in

the monetary base required for the same stimulus. The latter outcome is hardly

desirable, particularly in view of recent concerns about the size of national

debts. Indeed, it is possible that the recent large increases in national debts

combined with resulting increases in demand which have been scarcely enough

to counter the recession are explained by crowding out.

Incidentally, Friedman (1948) dismissed the need for government debt in

about two hundred words. The main valid reason for government borrowing he

could think of was a scenario where there was a sudden and large need to

constrain private sector spending, like in war time. In contrast, as he rightly

pointed out, in peace time, government spending is more or less constant from

year to year. In the latter case, there is little need for government borrowing.

To summarise, functional finance essentially consists of government

creating and spending extra money when unemployment is excessive and doing

the reverse when inflation looms. By ‘reverse’ I mean destroying money, that

is raising taxes and extinguishing the money collected.

The irrelevance of the structural deficit

A nice example of the irrelevance of the actual size of the deficit is the socalled

structural deficit: much discussed recently, at least in the U.K.

Government net spending in 2012 or thereabouts will clearly be less than

was planned prior to the recession, even if an economy returns to near normal.

The difference between the deficit in these two scenarios is the structural

deficit. But what is the practical relevance of the structural deficit concept?

Suppose that in 2012 the private sector is still deleveraging. That would

justify a larger deficit than normal. Conversely, if the private sector is

exhibiting irrational exuberance, to use Alan Greenspan’s phrase, that would

justify a smaller deficit, or even a surplus.

To summarise (and repeat) the size of the deficit or surplus in any given

year should be whatever looks like optimising the inflation unemployment

relationship. The actual size of the deficit or surplus is almost irrelevant.

Advocates of functional finance disagree with each other

Economists are famous for disagreeing with each other, a fault found

amongst functional finance advocates as much as with any other group. Thus

the version of functional finance set out here inevitably has the author’s stamp

on it. So for those new to functional finance, this paper should be taken as an

introduction to the subject, with further reading required.

5

The advantages of functional finance

Functional finance has several advantages over conventional policies. One

is simplicity: it cuts out some well-paid middlemen in the world’s financial

centres and for the following reasons.

Conventional stimulus consists of the following. 1. Treasury borrows $X. 2.

Treasury gives $X worth of securities to lenders. 3. Treasury spends $X. Under

functional finance, stage 1 and 2 become obsolete. (This simple illustration

assumes that the Treasury incorporates central bank functions, of course.)

A second advantage of functional finance is that the likelihood of owing

money to other countries is reduced. Third, less national debt means less

interest paid to creditors. Fourth, and to be cynical, the simplicity means fewer

economic illiterates in high places talking nonsense about national debts, and

doing immense economic damage in the process. Fifth, government is less in

thrall to ‘bond vigilantes’ because there are fewer government bonds or none at

all.

Quantitative easing is a move towards functional finance

Returning for a moment to the above three stage illustration in which the

Treasury borrows $X, if it then quantitatively eases the $X worth of securities,

this amounts to reversing stage 1 and 2. That is, assuming it is government debt

rather than private sector debt that is quantitatively eased, then QE comes to

the same thing as stimulus functional finance style. Indeed, QE amounts to an

admission of the limits of borrow and spend and the merits of functional

finance. And just to rub it in, borrowing combined with dropping interest rates

(mentioned above) amounts to a similar admission, because to force interest

rates down, the central bank may well have to engage in QE.

Inflation

The inflationary dangers of functional finance are obvious, but this is not a

strong criticism because conventional policies, if taken too far, can be just as

inflationary. Four reasons are set out below for discounting inflationary

dangers.

1. Dollar for dollar, ‘print and spend’ is doubtless more potent than ‘borrow

and spend’. But that does not of itself mean inflation. If a more potent fuel is

fed to a car engine but constant power output is required, it does not take a

genius to work out what to do: use less fuel! Similarly for given stimulus, print

and spend requires less expenditure than borrow and spend. Incidentally and in

relation to money creation, the word ‘print’ is used in this paper as a synonym

for ‘create’: that is, it is not just the physical printing of say dollar bills that is

referred to.

6

2. It is debatable as to how big a difference there is between what might be

called ‘formal’ money (monetary base, M1, etc) and various assets which are

effectively used as money, like government debt, as Klein (2009) makes clear.

3. There is the argument that expanding the monetary base enables

commercial banks to then lend some multiple of the monetary base expansion.

This sort of idea still appears in some text books, but it is obsolete and for the

following reasons.

i) One of the main constituents of commercial bank reserves is monetary

base and there are no reserve requirements at all for Canadian banks.

ii) There was an astronomic and unprecedented increase in the monetary

base of the U.S. and U.K. in 2009. The effect was certainly not a dramatic

increase in bank lending.

iii) If commercial banks do lend more simply because they have money to

lend, that suggests that banks are lacking in competence. It suggests they are

not too good at performing one of their basic functions: distinguishing between

credit worthy and non credit worthy customers. This calls for tighter bank

regulation e.g. banning 100% or 90% mortgages.

iv) Plenty of articles and papers which explain the unimportance of reserves

can be found by typing the phrase “banks are capital constrained not reserve

constrained” into a search engine.

4. On converting national debt to monetary base, there is no reason it has to

be done on a dollar for dollar basis. Indeed, a dollar for dollar conversion

would probably be too stimulatory, and thus inflationary. But the solution is to

obtain some of the money for buying back the debt from increased tax (or

reduced government spending) the effect of which is deflationary. (The word

deflation is used in this paper to refer to damping economic activity rather than

to falling prices – though of course the former sometimes causes the latter.)

This means there are two methods of buying back, one being stimulatory

and/or inflationary, and the second deflationary. They are respectively ‘buy

back with printed money’ and ‘buy back with money from tax or reduced

spending’.

This in turn gives two levers which can be adjusted to bring any rate of

national debt repayment desired, plus any stance on the reflation – deflation

scale that is desired. For example, for a faster rate of national debt reduction,

apply more of both levers. And for a more deflationary method of debt

reduction, apply more of the second relative to the first.

Extreme circumstances

The differences between functional finance and conventional policies are

more pronounced in extreme circumstances (e.g. a recession) than in normal

circumstances.

For example, given the extensive deleveraging and cash hoarding by the

private sector that has taken place in the aftermath of the recession, there is no

alternative, if demand is to be maintained, to a

7

That is if households hoard cash instead of spend it, then total spending

declines, which raises unemployment. The authorities, certainly in the U.S.

have failed to grasp this point. They expanded the monetary base by around a

trillion dollars in 2009, but as Galbraith (2009) pointed out, a trillion was not

enough.

This difference between normal and extreme scenarios may help explain

why Keynes and Milton Friedman approved of functional finance, but did not

devote huge efforts to advocating it: they may have thought that the sort of

credit crunch we have recently experienced was unlikely.

very large deficit.

Central banks under functional finance

The rules governing central banks under functional finance would be similar

to those that currently obtain. The rules would be approximately as follows.

(Government and central bank are treated as separate entities here.)

Rule 1. Politicians must work on the assumption that a small annual deficit

is permitted. The reason for this stems from the widely held view that inflation

of around 2% is optimum. The reasons are thus.

i) 2% inflation means a 2% reduction in the real value of the monetary base

per year other things being equal. Thus to maintain the value of the base in real

terms, a deficit equal to 2% of the base per year is required.

ii) Where an economy expands by X% a year in real terms, and assuming

the value of the base relative to real GDP is to be maintained, a further amount

of deficit is required equal to X% of the base.

iii) A very similar point applies to the national debt, if there is one.

The above three items amount to significant amount of annual deficit: a

deficit required even where neither stimulus nor deflation is required.

(Incidentally, exactly the same deficit is required under conventional policies:

it is just that, to be impolite, advocates of functional finance are clearer on what

size deficit is required and why than those with more conventional views. Put

another way, the contribution to the debate from some conventional sources

has consisted, at worst, of little more than getting hysterical about the large

numbers involved.)

Rule 2. The central bank shall be responsible for holding employment as

high as is consistent with acceptable inflation. If unemployment is too high and

there is room for reflation (i.e. stimulus), the central bank shall inform

government of what extra sums it (the central bank) will allow government to

spend without any corresponding tax collection or borrowing. Conversely, if

inflation looms, the central bank may have to inform government that the latter

rein in spending or collect more in tax, or even collect more in tax than it

spends. This surplus money collected is remitted to the central bank where it is

extinguished.

8

‘Borrowing’ so as to extinguish money

Under functional finance, where inflation looms, government (in the sense

that includes central banks) needs to rein in money either by raised taxes or

reduced spending.

However the purpose of this so-called borrowing is definitely not to

spend

the money: the objective is to extinguish or shred the money, or render it

temporarily inoperative. This activity is not borrowing in any normal sense of

the word.

Abba Lerner favoured borrowing, while Friedman (1948) did not and nor

does Mosler (2010 – see second last paragraph). However, Lerner did not

favour ‘borrowing as a source of funds’ (Lerner (1943:355). His attitude to

borrowing was the same as his attitude to government income or spending, that

is, ‘functional’. In other words, the only important question for him was ‘what

is the effect of borrowing?’. He favoured an interest rate policy so as to bring

about the optimum level of investment. But that is a very questionable idea and

for several reasons.

i) Does anyone actually know the optimum level of investment?

ii) Investment suffers from diminishing returns. In other words the most

worthwhile investments in any economy have been made long ago. While the

additional investment that comes from dropping interest rates from say four

percent to two percent brings negligible benefits.

iii) The returns on investments vary from plus a hundred percent and more

p.a. to minus a hundred percent and more p.a. Again, in the light of this, the

above change of interest rate from four to two percent is of negligible

importance.

iv) There is the argument that government can control the price of money

or the quantity, but not both. Lerner seems to think governments can control

both. The view taken in this paper is that governments can control only one,

and that should be the quantity. (As to interest rates these would tend to be near

zero - as obtains at the time of writing. See Mosler (2005) for more on this.)

To summarise, to the extent that Lerner favoured borrowing, it was

borrowing not in any normal sense of the word. And even this latter form of

borrowing is questionable because the aim is near impossible: optimising the

amount of investment.

Is monetary base a debt owed by central banks?

A possible criticism of functional finance is that it involves replacing

national debt at least to some extent with monetary base, and the latter appears

on the liability side of central banks’ balance sheets. Thus it could be argued

that not much has changed in that monetary base is borrowing or a central bank

liability of sorts, much like the securities that governments issue.

One answer to this was given by Buiter (2008) who said ‘These monetary

base ‘liabilities’ of the central bank are not in any meaningful sense liabilities,

9

because they are irredeemable.’ (Willem Buiter is a former member of the

Bank of England Monetary Policy Committee.)

Does borrow and spend increase private sector assets?

In the above illustration where the Treasury borrowed $X, the net result

appears to be that the private sector is $X better off, which should be

stimulatory. However part of the reason that Treasury securities have value is

the interest they earn. But this interest comes out of the pockets of taxpayers,

thus the creation of these securities involves the creation of a liability as well as

an asset! Thus the size of the net increase in private sector assets is debatable:

yet another uncertain aspect of borrow and spend.

In contrast, print and spend undoubtedly gives the private sector additional

assets.

Borrow and spend with a view to stimulus: the conclusion

Borrow and spend involves uncertainties of which crowding out is just one.

For this is and for other reasons given above the zero borrowing alternative,

functional finance, should be given serious consideration.

3. Borrowing to purchase assets

A third reason governments borrow is to purchase assets, for example

infrastructure investments. The points made here apply both to countries which

issue their own currency and to countries that are part of a common currency

area.

The first flaw in this form of borrowing is that governments just do not need

to borrow in order to fund capital projects. This is for two reasons. First, the

total of all such investments in any given country is much the same every year.

Thus if a country

devoting much the same proportion of its budget to these investments every

year (in the form of loan repayments, interest, etc) as if it pays for the

investments out of income.

Second, governments do not face a problem that micro economic entities

like firms or households often face. This is that the latter sometimes make

investments worth several times their annual income. A family buying a house

is an example. Here, there is normally no alternative to borrowing.

Conclusion: the need to fund capital projects does not justify government

borrowing. However, the case for borrowing here is stronger than is the case

with the other reasons for borrowing considered in this paper. Indeed, there is

not much difference between funding a public sector investment and debiting it

10

with a nominal sum for interest, and on the other hand, actually borrowing the

capital sum and paying interest to lenders.

does borrow, to fund this expenditure, it will just end up

Spreading the cost across generations

A plausible sounding argument for government borrowing to fund capital

projects is that this spreads the burden across generations because future

generations ostensibly have to repay some capital and/or pay some interest.

The flaw in this argument is that while various members of the next generation

inherit the above liability, others inherit a corresponding asset, namely

government debt or bonds. These two net to nothing.

Put another way, the brute physical reality is that it is impossible to build a

road this year (2010) without diverting concrete, steel and so on from other

uses in 2010 (assuming constant aggregate demand). That is, people living in

2010 have to sacrifice the consumption of other products to have the road built.

Or put a third way, it is a physical impossibility to build a road in 2010 with

concrete produced in 2050.

The only way to have future generations in one country to pay for today’s

investment is for the country to borrow the capital sum from some other

country, and gradually pay the money back, as pointed out by Musgrave (1939:

269). But if every country adopts the latter policy, it becomes self-defeating.

Moreover we all benefit from investments made by previous generations

where the relevant debt has long been paid off. An example is education:

capital costs apart, these costs are normally born and paid off annually (e.g.

teachers’ salaries). Thus allocating the costs of investments to each generation

in any sort of accurate way is a bureaucratic nightmare. It is probably simpler

just to accept that each generation inherits huge benefits from previous

generations, and that each generation should ‘leave the world as it found it’,

that is fund and pass on a roughly equivalent amount of investment to

succeeding generations.

To summarise, having future generations pay for the investments they

inherit is, first, almost a physical impossibility, and second, even were it

possible, the idea is more trouble than it is worth. Thus government borrowing

aimed at the latter objective is also more trouble than it is worth.

4. Does the erratic timing of tax receipts and expenditure within each

year justify borrowing?

The quick answer to this is that given a short term lack of income,

government can easily print money; then, a few months later, when income

exceeds spending, government can do the opposite, that is rein in money and

extinguish it. It could be claimed that the latter money printing would be

inflationary. However, there are flaws in the latter claim.

11

First, inflation is unlikely to take off just because the money supply is more

than optimum for three months and then an equivalent amount below optimum

for the next three months.

The next flaw involves Ricardian equivalence of a sort: that is, where a

private sector entity knows there is a $Y tax liability in the pipeline and has

sufficient cash to meet this liability, that entity is almost certain to use the cash

to meet the liability, rather than spend the cash on something else in the

interim.

This argument gains support from examining what might be called the

counterparties to the regular annual government shortages and surpluses of

income. These counterparties are very definitely not the sort of people who

might blow a temporary excess of cash.

For example a significant proportion of people living on social security

might spend any temporary excess cash; however a large majority of these

people (at least in the UK - pensioners or those on invalidity benefit) receive a

fairly regular income from the state and thus do not cause erratic changes in

government net income.

As to a more responsible or talented or lucky section of the population,

employees, this section of the population also has a regular income. As to

income tax, this is deducted from wages and salaries, both where wages are

paid weekly and where paid monthly (at least in the UK). Thus this section of

the population does not contribute much to changes in government income

either.

Finally, the sections of the economy which really cause the irregularities in

government income are corporations and the self employed. These sections of

the economy (at least in the UK) pay tax on profits once or twice per year.

However, it is precisely this section that is least likely to blow a temporary

excess of cash. Any corporation or self employed person with a propensity for

this sort of behaviour will not last long in business.

Conclusion: the variations in government net income through the year are

not a reason for governments to borrow.

Main conclusion

Of the four alleged justifications for government borrowing examined here,

three are feeble. The only one without serious flaws is borrowing to fund

capital projects like infrastructure investments. But even here, the merits of

borrowing are debatable.

But that is not a message that Goldman Sachs, other banks or the well-paid

employees in the world’s financial centres want broadcast. They have an

interest in seeing governments heavily in debt. They devote millions and will

devote further millions to furthering their interests: foisting debt on

governments. Hopefully the arguments in this paper will help governments

counter banks’ sales pitch. Indeed there was one particularly odd case in

Australia in 2002 where banks managed to foist extra debt on the Australian

12

government despite the fact that the latter was running a surplus and happily

paying off its national debt – see Mitchell (2002 and 2010).

References

Buiter, W. (2008). ‘Wanted: tough love from the central bank.’

Times,

Colander, D. (2003). ‘William Vickrey’s Contributions to Economics.’

Middlebury College Economics Discussion Paper No. 03-23. Vermont, U.S.A.

Friedman, M. (1948). ‘A monetary and fiscal framework for economic

stability.’

Galbraith, J.K. (2009). ‘Navigating the Jobs Crisis: Old Mistakes Die Hard.’

Franklin and Eleanor Roosevelt Institute. Available at

http://www.newdeal20.org/2009/11/23/navigating-the-jobs-crisis-old-mistakesdie-

hard-6556/ [23

Klein, M., Lantz, C., Sweeney, J., & J. Wilmot, (2009). ‘Long Shadows,

Collateral Money, Asset bubbles, and Inflation.’ Credit Suisse.

Landes, D.S. (1994). ‘Abba Ptachya Lerner.’ In:

Volume 65,

Lerner, A. (1943). ‘Functional Finance and the Federal Debt.’

Research,

Mitchell, W. & Mosler W. (2002). ‘Public Debt management and

Australia’s macroeconomic priorities - Submission to the Review of the

Commonwealth Government Securities Market.’ Newcastle, Australia: Centre

for Full Employment and Equity.

Mitchell, W. (2010). ‘Market participants need public debt.’

http://bilbo.economicoutlook.net/blog/?p=10404#more-10404

Mosler, W. & M. Forstater (2005). ‘The natural rate of interest is zero.’

Financial22nd March.American Economic Review, 38(3): 245-264.rd November 2009].Biographical Memoires,208-231. Washington: National Academy of Sciences.Social10: 38-51.

Journal of Economic Issues

Mosler, W. (2010). ‘Proposals for the banking system.’

, 39(2): 535–542.Huffington Post,

21

13

Musgrave, R.A. (1939). ‘The Nature of Budgetary Balance and the Case for

the Capital Budget’.

Musgrave, R.S. (2010). ‘Government borrowing is pointless where a

country issues its own currency.’ Munich Personal RePEc Archive paper No.

20057.st January.The American Economic Review, 29 (2): 260-271.