Effect of subsidies on
employment
[go
to accompanying
letter]
Assume [1]
(i) that the
total amount spent on investment is constant.
(ii) that
wage-earners and profit-makers save the same proportion of income, so
that the "expenditure multiplier" is the same in all circumstances.
Then with the given expenditure on investment there must be a fixed
expenditure of money on consumption goods,--i.e.
[b] that amount which the "expenditure multiplier"
demands.
(iii) that the
money wage-rate received by workers remains constant.
Suppose now
that a given tax per unit of profit made in consumption industries is
imposed and a subsidy per man employed in industries producing
consumption goods is paid, this subsidy per man being continually
adjusted so that it just uses the whole of the tax per unit of profit
(i.e. if employment increases the subsidy per unit of employment must
fall.)
We are now
faced with a situation in which (a) total expenditure on consumption
goods remains constant, (b) a tax per unit of profit in consumption
industries is being levied, and (c) a subsidy per unit of employment
in consumption industries is being paid. This must lead to increased
employment in consumption industries; for, if it did not, producers
would have an incentive to produce more. For if output remained the
same, the price of consumption goods would be unchanged because total
expenditure is constant, whereas the subsidy per unit of employment
will reduce marginal prime costs, and the tax per unit of profit will
not raise marginal prime costs. As output increases, the price of
consumption goods will fall, since total expenditure on consumption
is constant; total profits will fall and total wages will rise, since
total income is constant and there is greater employment at a given
wage-rate. As employment increases, therefore, the tax per unit of
profit will raise less revenue, because total profits fall; and the
subsidy per unit of employment will fall both because total revenue
falls and because there is a larger volume of employment to be
subsidized. But this means only that the increase in employment will
not be as great as might appear at first sight; there must be some
increase in employment [c] ,
for otherwise this counterbalancing fall in revenue and in the
subsidy per unit of employment would not occur.
Let us now
modify the above by assuming that the profit-tax &
employment-subsidy is applied to all industries (investment &
consumption), retaining the assumptions (i) that total expenditure on
investment is constant, & (ii) that the multiplier is unaffected
by transfers to labour so that money expenditure on consumption is
also constant, & (iii) that the money wage-rate is constant. Then
there will be an increase in employment in investment industries as
well as in consumption industries; for more machines will be produced
and sold at a lower price per machine for the same total amount of
money <for +> for the same reason that in the previous argument
more consumption goods were sold at a lower price per unit.
Let us now
modify the assumption that wage-earners & profit-makers save the
same proportion of income, retaining (i) the assumption that total
expenditure on investment is constant & (ii) that the money wage
rate is fixed. Since according to the previous argument the
tax-subsidy arrangement will transfer income from profits to wage
earners, & since wage-earners spend a larger proportion of income
than [d] profit makers, &
since a total expenditure on investment is constant, there will be a
rise in expenditure on consumption. Thus in this case the increase in
employment will be greater than in the previous argument, because
there will now be an increase in incomes. The price of consumption
goods will not fall as much as before, so that employment in these
industries will be greater than before; this argument is strengthened
by the fact that profits in consumption industries will not fall as
much as before because of the increased expenditure on consumption,
so that tax revenue will not fall off so much and a larger subsidy
per unit of employment can therefore be paid.
Let us now
modify the assumption that expenditure on investment is constant, and
assume instead that the rate of interest is constant, retaining the
assumption that the money wage-rate is constant. The question boils
down to this:--Will total money expenditure on investment rise, in
which case the increase in employment will be greater than by the
previous argument, because total incomes will rise still more? Or
will total money expenditure on investment remain constant, in which
case the previous arguments need no modification? Or will total money
expenditure on investment fall, in which case the increase in
employment will be less than according to the previous argument &
may indeed turn into a decrease of employment, if the incentive to
spend money on investment is greatly reduced? Let us assume that the
tax per unit of profit is also levied per unit of income obtained as
interest on debt fixed in terms of money (e.g. on interest on
government debt.) Then if the rate of interest is 5% & the tax
per unit of profit-interest income is 50%, this reduces the net yield
on government securities to 
% as well as reducing the net yield on a machine which is expected to
give a certain rate of yield on its cost by 50% as well. In other
words there is no decrease in the incentive to borrow money at 5% to
spend on investment unless the gross yield on a capital development
(i.e. the marginal efficiency of a machine without making allowance
for the tax on its yield) has fallen. Will the gross expected rate of
yield on the cost of a machine have fallen, i.e. will the gross
expected yield on a machine have fallen or will the present cost of a
machine have risen? The present price in the market (i.e. cost to the
investor) of a machine will have fallen, because, as we have argued,
there will be a greater number of machines selling at a lower cost
per unit, unless we can find other reasons for believing that this is
not true. This in itself will raise the gross marginal efficiency of
a machine and will tend to increase the expenditure of money on
investment. The gross expected yield on a new machine will not have
fallen, except because there is an increased output of machines which
may lead people to expect lower yields in the future. (This does not
allow for an irrational change in confidence due to the tax per unit
of profit; but there is no rational reason why the gross yield on
machines should be expected to be lower except because of their
greater rate of output.) We can conclude therefore that expenditure
on investment may be greater than we previously concluded, because
the present price of machines is lower; & may be less than we
previously concluded, because the present rate of output of machines
being greater may lead to an expectation of lower yields. But in no
case will the number of machines produced & so employment in
investment industries fail to increase, because the only reason
tending to reduce expenditure on machines at a given rate of interest
is the fact that their output is greater. There might, however,
conceivably be a fall in total expenditure on machines, because the
fall in the price at which machines will be sold due to the
profit-tax & employment-subsidy allows a slightly greater output
of machines to be sold for a smaller total amount of money. It is
therefore just conceivable that if the gross marginal efficiency
schedule is very inelastic, because a small increase in output of
machines assures a very large fall in expected yield on them, &
if the multiplier is very large, there may be a fall in money
expenditure on investment leading to so large a fall in expenditure
on consumption that there is a net diminution in employment. But for
this to be true the inelasticity of the marginal efficiency of
capital schedule has got to be so important that it offsets (a.) the
certain increase in employment in investment industries, (b.) the
rise in the marginal efficiency schedule due to the fact that the
price of machines is lower, (c) the tendency for consumption to
increase because of the effect of the tax-subsidy in transferring
income from profits to wages and (d.) the fact that with the
tax-subsidy in consumption industries the volume of employment in
these industries will increase even if there is no rise--or even a
slight fall--in total expenditure on consumption.
- 1. Refers to
Harrod, The Trade Cycle (
1936:8 ), pp. 230-32.
- a. AN,
five pages on five numbered leaves, is in file HP
IV-745-767/12.
b.
Ms: «ie.» (in all occurrences.
This is normalized throughout the note without further
warning).
c.
Ms: «increase employment».
d.
Ms: «that».
Welcome page
top of page
Return to index of this section
Go to previous page
Go to next page