One type of simple Keynesian analysis would assume
that government expenditure produces nothing of value. An increase in
government expenditure financed by taxes has a positive effect on
aggregate demand. But this is accompanied by a dollar-for-dollar
reduction in (disposable) income combined with the same reduction
in consumption that would have occurred had there been an equivalent
decline in income from any other source. Since consumption
declines by less than the decline in disposable income, the net
effect on aggregate expenditure will be positive.
A more sophisticated Keynesian view would suggest that if the socially
optimal amount of government expenditure is being undertaken a
one-dollar increase in that expenditure should produce value
exactly equal to the dollar of taxes required to finance it.
Wealth, permanent income and consumption should not change. As a
result, increases in government expenditure should have greater
impact than the preceding Keynesian analysis would postulate because
there will be no offsetting reductions in tax-payer wealth and
consumption---unemployed workers will be put to work and will spend
the earnings so received.
Another interpretation would be the situation where the country is
engaged in a war. In this case wealth and consumption would be
reduced by the government expenditure in comparison with the
pre-war state, though not in comparison with the scenario where
the war is avoided by surrender to the enemy. But standard
Keynesian fiscal policy deals with government expenditure changes
designed not to fight wars but to smooth cyclical fluctuations in
output and employment. What is at issue is the timing of the
path of government expenditure to offset fluctuations in private
expenditure that cause variations in employment. Wealth effects
from misallocating resources by making the expenditure too early
or too late should be counterbalanced by the social gains from
smoothing out fluctuations in output and employment.
So the question ultimately turns on whether a temporary
expansion of government expenditure also leads to an expansion of
total private plus public expenditure. While the possible wealth
effects above may not be important, substitutions of public for
private consumption and investment expenditures may well be.
Suppose, to take a somewhat silly but informative example,
the government decides to hand out free soap and toothpaste to
the community to increase employment in the factories producing
these products. Since it must raise taxes to cover the expense
of purchasing these household items (we have already dealt with
the cases where bond and monetary finance are used instead of
taxes), the public has to reduce either its consumption or its
savings by the additional amount the government is spending. It
would seem reasonable to suppose that people would maintain their
original consumption of soap and toothpaste by simply substituting
the products provided by government for those that were
previously purchased in the private market. Total private plus
public consumption would be unchanged and the fiscal policy would
have no effect on the IS curve.
To have an effect on aggregate public plus private consumption the
government has to supply something that the private sector would not
otherwise have consumed. More generally, it has to induce an increase
in the fraction of permanent income consumed. This may be a quite
difficult undertaking.
Another form of government expenditure policy is the
provision of welfare benefits to individuals whose need has been
created by a decline in employment in a recession. Here, taxes
are levied on people with incomes to pay them and the funds given
as benefits to those with inadequate incomes. Aggregate consumption
will increase to the extent that those who pay the taxes
reduce their consumption by less than the additional spending by
those who receive the benefits. An overall increase in spending
might be expected to the extent that the recipients of the funds
are more liquidity constrained than those paying the additional
taxes. It might be easier for those paying the taxes to maintain
consumption by borrowing than for the recipients of the funds to
be otherwise able to borrow to maintain their consumption.
Another avenue for expanding government expenditure in
recessions is an increase expenditures on capital goods. To the
extent that the government undertakes investment expenditures
that do not substitute for private investment expenditures---for
example, building and fixing roads---total private plus public
investment will increase. The ordering of new jet aeroplanes by
the government airline to provide expanded service, on the other
hand, might well reduce the future returns to investment in
aeroplanes by competing private airlines so that the public-sector
investment will be offset by an equal contraction of private-sector
investment.
Also, if an increase in total public plus private investment is to
increase aggregate demand and shift IS to the right, it must not be offset
by a reduction in private consumption in response to the increase
in taxes necessary to finance the expanded public expenditure.
This will be assured if the level of wealth and permanent income
is not significantly reduced. Such wealth reductions can be
avoided if the government produces capital goods of value close
to what would otherwise have been produced with the additional
taxes raised.
A reduction in private sector expenditure in response to
an increase in government expenditure is called crowding-out.
Additional public expenditure can crowd out private expenditure,
leaving total public plus private expenditure unchanged. While
crowding-out usually will not be complete, some crowding out can
nearly always be expected. Thus, in analyzing the effects of government
expenditure changes on aggregate demand it is important to examine those
expenditures carefully to make a judgment about the amount of
crowding-out that might be expected in each particular case.
Another problem may arise even if the government attempts to offset the
unemployment effects of a recession by producing additional roads or
bridges or other public goods that it conventionally produces. The problem
is that the workers employed by this additional government production will
not typically be those laid off by the cyclical contraction of private
production---the government may simply end up paying overtime to workers
who have not been laid off.
It thus turns out that the government must be very careful as to the
particular goods it produces during counter-cyclical expansionary policies.
Given the planning requirements and the necessarily political nature of
government expenditure, the recession may be over before a proper form of
expenditure can be agreed upon and, if politically possible, undertaken.
Indeed, tax cuts and increases for the liquidity constrained would appear to
be the easiest, quickest and therefore best counter-cyclical policies
for governments to implement.
Its time for a test. Be sure to think up your own answers to the
questions before looking at the ones provided.
The standard Keynesian analysis also postulates that
an expenditure of the government on consumption or investment
goods will result in an equal increase in exogenous aggregate
public plus private spending on consumption and investment goods.
Here we deal with some modern refinements of the traditional
analysis---the conclusion that an expansion of government expenditure will
shift the IS curve to the right (at a given level of
the real exchange rate) remains, albeit in a substantially modified form.