\

Consumption

Consumpt,on
. .,s Important for a••regate
""' demand, the total demand for all goods and services in the economy,
since according
. to Keynesian
. theory , aggregate demand determines the level of output and employment In an
economy; the more we demand , the more we produce and the more we create employment. Also, income
that is not consumed •,s saved and savi ngs have a large Impact on the growth of an economy. Thus ,
consumption• .,s 1mpo rt•nt
• to undernand savings, capital stock, Investment, employment, and income growth.
But, there .,s more.. the effec1iveness of economic policy is also closely related to the nature of the
consumpl"ion funct ion.

Consumption depends on rurrent di5posable income, that is current income minus taxes . In turn, the marginal
propensity to consume (lo wit, the change in consumer spending due to a change In income) det ermines the
magnitudes of government e•penditure and tax multipliers. To put it simply : how much GDP Increases or
decrea~es when the government increases or decreases spending In the economy, depends on how much we
spend on consumption.

INTERTEMPORAL CHOICT AND CONSUMPTION FUNCTION

Rational consumers are a~ umed to behave as selfish utility maximizers. More specifically, In consumer theory
the rational consumer ob1?ys the axioms of rational choice and maximizes his/her utility function subject to the
budget constraint. Consequently, Individual demand curves are derived from this standard model of utility
maxlmltallon. The same model is then extended to include intertemporal choice, that Is the choice between
consumption today versus consumption at some point in the future, which forms the basis of mainstrea m
consumption function.

This theory Is given by lrvlng Fisher.

In this framework, It Is assumed consumers are rational, forward-looking agents and choose consumption
levels for the present and future so as to maximize lifetime satisfaction.

Consumers' choices are subject to an intertemporal budget constraint, a measure of the total resources
available for present and future consumption.

In a simple two-period model, there are two periods, period 1 (which is the present) and period 2 (which is the
future) . Let Yl and Y2 to be Income in period 1 and 2 respectively. Let Cl and C2 to be consumption in period 1
and period 2 respectively, and S is savings. The consumers' budget constraint in the first period is
Yl : Cl+ S or S : Yl - Cl

If S > 0 the consumer is saving, and if S < 0 the consumer is borrowing. Saving (borrowing) yields an interest
rate (r) (costs) . In the next period, the budget constraint is therefore:
C2: Y2 + (l+r)S

C2: Y2 + (l +rl(Yl -Cl)

The term (1 + r)S is the rate of re turn of savings.

We can rearrange terms to write :

(l+r)Cl + C2 = Y2 + (l+r)Yl

If we divide the above expression by (l+rl,

we get : Cl+ C2/(l+r)=Yl + Y2/ l+r

pg, 1
r________

Fig. 8.1: lntertemporal Budgel Line

The above equation shows that the present value of lifetime consumption (on the left hand side of the
equation) equ als the present value of lifetime income (on the right-hand side of the same equation). Present
value means the value in terms of the consumption goods in period 1. In other words, 1/(1+r) is the relative
price of future consumption in terms of current consumption : one unit of consumption today is equivalent to
l+r units of consumption tomorrow. Figure 8.1 shows the intertemporal budget constraint.

The budget constraint shows all combinations of Cl and C2 that just exhaust the consumer's resources. At
point E, consumption equals income in both periods. At points which are between F and Ethe consumer has
positive savings, while at points which are between F and G the consumer has negative savings (that is, s/he
borrows).

Indifference Curves

Properties of Indifference Curves

• If Cl and C2 are goods (i.e., objects for which more is preferred to less), indifference curves are downward
sloping .

• An indifference curve located northeast of another one yields higher utility.

• One (and only one} indifference curve is associated with each point in the positive quadrant; they densely
populate it. • Indifference curves do not cross one another.

• The negative of the slope of the indifference curve is known as the intertemporal marginal rate of

substitution of C2 for Cl .
• The indifference curves we focus on are convex. If you are consuming a lot in _period 1 and almost nothing in
period , you are not willing to give up lots of period-2 consumption for an add1t1onal unit of penod-1
2
consum ption.
But if you are consu ming very little in period 1 and a lot in period 2, you are willing to give up a lot of period-2
consumption for an additional unit of period-1 consumption . This property of preferences is known as
diminishing marginal rate of substitution of C2 for Cl.

_( ~ 2 -
 ___
IN_:"ER~~~PORAL CHOIC_
E_ J
Indifference Curves represent u = U(C,, C2 )
c,

c,

The budget co nstrai nt is com bined w ith t he notion of indi fference cu rves, which show combinat ions of prese nt
and futu re co nsumption level s tha t leave the consum er equally happy. Moreover, higher indifference cu rves
rep resent higher levels of happin ess. As a utility maximiser, the consumer aims to be at the high est possi bl e
indiffe rence curve given t he consumer's budget constraint.

c'l

O'-------"ce------;,._--
8 C 1

Fig. 8:'3: Optimisation iJ! Consumption

the opti mal poin t is F, where the slope of the budget line just touches the highest possible indifference curve
(IC2). Only at this point the consumer maximizes his/her utility out of consumption given a budget constraint.
At point F, t he slope of the indifference curve, which is the marginal rate of substitution (MRS) between
present and futu re consumption, equals the slope of the intertemporal budget constraint (l+r). At point F,
th erefore, MRS= 1+r.

Poi nt F ca n change if there are changes in income (the budget line shifts) or changes in the rate of interest (the
slope of the bu dget line cha nges). lnt ertemporal consumption theory indicates that consumption decisions
depend on ly on the present va lue of lifetime income (as opposed to disposable income today) . Since the
consumer can borrow or len d between periods, the level of interest rate plays an important role in
consum ption decisions.

Effect of changes in Income on lntertemporal Consumption

pg 3
 . ·ther Yl or Y2 shifts the budget
· rease in income ei · f first &
to increase in income, an me choose a better combination o .
consumption responds _ h budget constraint allows consum~r t~ curve. consumer responds to shift
constraint rightward . H1g er er can reach to higher indifference
tion so consum . d
second period consum P . . nsumption in both peno s.
. t by choosing more co
in budget constrain . d then both are norma I goo d s.
consume more in both the per10 s .
If income increases and consumer want~ tof_ t period or second period the consumer spreads it over
• ·ncome occurs in irs .
Whether the increase 1n 1 . . lied consumption smoothening.
consumption in both periods. This behaviour is ca .
e that a erson's current consumption depends largely on his current .
Keynes consumption theory assum s p . . b sed on the resources the consumer expects over his
income. But Fisher's model says that consumption is a
lifetime.
Changes in Real Interest Rate & its affect on Consumption
A . ease in real interest rate rotates the consumer's budget constraint around the point (Yl, Y2), and
c~a1~;:s the consumption in both periods. so consu~er_moves from point E to E" . Economists decompose the
impact of increase in real interest rate on consumption into 2 effects

a) Income Effect
b) Substitution Effect
The income effect is the change in consumption due to increase in interest rate which causes income of
consumer to increase. Increase in interest rate makes consumer better off by increasing his return on
savings. If consumption in both periods is normal good, the consumer wants to spread this improvement
in welfare in both periods by increasing consumption .

Substitution Effect is the change in consumption that results from the change in relative prices of
consumption in 2 periods. Consumption in period two becomes less expensive and present consumption
becomes more expensive due to rise in interest rate and opportunity cost. Since interest rate earned on
saving is higher, the consumer now sacrificesmore first period consumption to obtain an extra unit of
second period consumption, substitution effect makes consumer to choose more consumption in period
two and less consumption in period one.

If substitution effect is more than income effect, then increase in interest rate makes present
consu mpti on less attractive and savings increases. But present period consumption Cl decreases.

pg. 4
 But if income effect is more than substitution effect then increase in interest rate lowers savings and
increases present period consumption. Depending upon the relative size of income & substitution effect
an increase in interest rate either increase savings or decreases savings.

lt\g. -t..s: Income·ar1d Suos_iitu~Jir.i '~ ~-:

ot• Change iii .the''Rat~\bt11tr~1
11

1:

Constraints on borrowings 11

t1
I.
Fisher's model assumes that the consumer can save or borrow. The ability to borrow allows current
consumption to exceed current income. When the consumer borrows, he can consume more than his income
in period 1. But for many people such borrowing is impossible.

The inability to borrow prevents current consumption from exceeding current income. A constraint on
borrowing can be represented as

c1::;n

The inequality states that consumption in period 1 must be less than or equal to income in period one. This
constraint on the consumer is called borrowing constraint or liquidity constraint.

The consumer's optimal choice also satisfies the intertemporal budget constraint & borrowing constraint. The
shaded area represents the combination offirst & second period consumption that satisfy both constraints.

I
I

pg. 5 !
[
 Budget'
/ con~rairil

' ;0',t::,,.LLt.U ..L.L.,~~ - - - -""""'.~ ~

Yi
.,..,gtJi~( ~ors~rllptJo.n~Possibilities 'in the
,~i\~·,.of a·t}orrowirig Constraint

,.

., -;>~:t
. ' . Ii ~--·

C
.I r,-
0 "-"!--~-~---..;;.,:..~,..
t>,. ;c, ·

In first diagram consumer wishes to consume less in period one than he earns. Cl<Yl

The borrowing constraint is not binding and does not affect consumption. In this consumer's optimal choice is
at point B.

But in the second diagram, the borrowing constraint is binding and consumption in both periods depends on
the present value of lifetime income.

Cl=Yl and C2=Y2

For those consumers who would like to borrow but cannot, consumption depends only on current income. If
the optimal choice is at point Ethen this is best available choice for consumer. But if he wants to consume at
point E' then he cannot choose that point because borrowing is not allowed and constraint is binding.

MAINSTREAM CONSUMPTION THEORIES

Kuznets's .empirical findings provided the initial stimulus for theoretical research on aggregate consumption
patterns. The first theories to appear were the life-cycle hypothesis and the permanent-income hypothesis.
These two theories started by employing Fisher's intertemporal choice as the microeconomic foundation of

pg. 6
 ECONOMICS MANTRA

aggregate consumption, thus rejecting Keynes's psychological-based approach to consum ption. Current
,, : : .: :· - orthodoxy with the core assumptions of utility maximizing and forwa rd-looking agents is essentially based on
these two theories.

Life-cycle hypoth esis

The life-cycle hypothesis of consumption function was developed mainly by Franco Modigliani and Richard
Brumberg in 1954 (Modigliani and Bru mberg, 1954). Its underlying co nceptual basis is that individuals
maximize their utility of consumption over their life cycle, and not over th eir disposable income over, say, a
year. In this hypothesis individuals try to smooth their consumption during worki ng life and retirement life.

At Othere is beginning of the person's working li fe and T2 marks th e point of his/her retirement. In the early
years of a person's working life and also after retirem ent, income is lower than consumption expenditures. For
the time span Tl to T2 the reverse is true. As the person proceeds through life, his/her. productivity increases
and as a res ult his/ her income increases up to the point of retirement T2. From this point onwards,
consumption decl ines but not as much as income. The hypothesis suggests that individuals consider the profile
of the stream of income during their whole productive life and pace their consumption expenditures
accordingly. At the early stages of their working life and between retirement and death they have negative
savings. These are balanced by the positive savings occurring at the time period Tl to T2.

The total resources available to the individual over his/her entire life span are the sum of the individual's net
worth at the end of the proceeding period, plus his/her income during the current period from non-property
sources, plus the total of the discounted values of the non-property incomes expected in the future time
periods. Ass uming for simplicity that real interest rate is zero, the individual divides his/her resources equally
over time to smooth out consumption.

C=(W+ RY)/T

where Wis initial wealth, Y is annual income until retirement (assumed constant), R is the number of years
until retirement, and Tis lifetime in years. The same relation can be

written as: C=aw+ ~Y

with a= 1/T and = R/T,

where a is the marginal propensity to consume out of wealth, and~ is the marginal propensity to consume out
of income.

Implications of Hypothesis

The life-cycle hypothesis can also account for the discrepancies in empirical consumption data th t
d' d f a we
,s_cttusse be ore. The reasoning is the following. The life-cycle consumption function divided by income can be
wn en as

C/Y = a(W/Y) + ~-

Across households, income varies more than wealth, and this implies that high-income households should
have a lower AP_Cthan lo~-income households. However, in the long run, aggregate wealth and income grow
together and this results m a stable APC.

.Apart
ct· ·ctfro m assuming
• . perfectly rationa 1consumers, the life-cycle
. consumption theory also assumes that
m 1v1
w·II h ua 1share md1fferent t o the f orm m
• wh.IC h resources accrue. For example, a given increase in resources
. 1 ave t e same. effect on consum Pf ion whet her that .increase takes the form of an increase in current
mcome, expected mcome, or net worth .

Consumers are assumed to have erf . - ..

correctly (n f d p ect com putational ab1l1t1es and to be able to estimate future income
0 un amental uncertainty).

pg . 7
 ECONOMICS MANTRA

one of the main criticisms in this regard is that switching of assets is not a costless transaction. Further, this
theory ignores imperfections in capital markets, which impose a limit on the amount an individual can borrow
(liquidity constraint) . Liquidity constraints affect the ability of households to transfer resources across time
periods. finally, another line of criticism has to do with the notion of expected income, which is part of the
individual's total resources.

Effect of Increase in Wealth on consumption function

Increase in wealth shifts the consumption function upwards. But in the short run wealth is constant and in long
run when wealth increases the consumption function continues to shift upward, this upward shift prevents th e
APC from falling as income increases.

Consumption, Income & Wealth over the life Cycle

If th e consumer smoo th s consumption over the lifetime which is indicated by horizontal line. Consumer will
~arn & save during th e working life and accumulate wealth during working life. Then dissave during retirement
hfe and run down her wealth during retirement.
 -1flftllf'1if'1l''lltlffilf11i'ilft?1M:1&1B1wz
$

I
:1

.R e t irement
begins

Permanent Income Hypothesis (Milton Friedman)

Apart from the life-cycle theory, the other attempt to criticize Keynes's approach to consumption was made by
Milton Friedman with his permanent-income hypothesis (Friedman, 1957), where permanent income is an
individual's income over his/her lifetime. In his attempt to define a consumption function, Friedman (1957)
rejects Keynes's use of current income as the determinant of consumption expenditure, based on the idea
consumers are forward-looking meaning future concerns affect current consumption decisions.

Forward-looking consumers is a common point between Friedman's theory and the lifecycle theory. However,
according to Friedman current income is subject to random, transitory fluctuations while according to life cycle
theory, current income changes systematically as people move through their life cycle.

Consumers use their savings (or borrow) to smooth consumption between good and bad years. These imply
that current income differs from permanent income: Vt== VP+ YT , where Vis current income at time t, VP is
permanent income projected at time t and YT is transitory (or unexpected changes in) income.

The transitory component has an expected value of zero reflecting the notion that over time transitory gains
are offset by future transitory losses and vice versa . Thus, in the long run observed levels of income (V) are

equal to permanent income (VP) .

An important part of Friedman's theory was his assumption that permanent income is an average of income
over the last several years. This implies that if there is a sudden rise in current income, there would be only a
small increase in permanent income, contrary to Keynes's theory. Income would have to increase for several
years continuously before people would expect permanent income to increase. In other words, consumers
correct their previous estimates of permanent income by the amount of deviation of current income from
previous period estimated permanent income (adaptive expectations).

pg. 9
)
 ___ IIIIW_ __

• ( ) · d' ·d d into permanent consumption, CP , and transitory

r . C is •t IVI e·ncome t ransitory consumption
In the same way as income, consumption . 1s
• regar d e d as temporary .
1
consumption, CT. Thus Ct =CP + C · Like tranSI ory ' . f ermanent income while
Friedman assumes permanent consumption is a constant proportion {da) o d~' I nned" cons~mption
· b · t rpreted as planne an unp a
permanent and transitory consumption may e in e
respectively. · hc nforeseen increment in income does not
Based on Friedman's assumption that yr is uncorrelated wit • any u th .f . . h
result in unplanned consumption. Friedman justifies
· this
· premise by pointing
t· planout at evenhis/her
but adjust .' income is ot er
asset
than expected, the consumer would tend to stick to his/her consump ion ,

holdings.
. . . . . yp 'th O < a < 1 where a is the traction of
Given the above, Friedman's consumption function 1s<: = a , WI •

permanent income that people consume per year.

The APC will be: APC = CY= acP /Y

Observed short-run behaviour is explained through the value of transitory income for different income groups.

For high-income groups, transitory gains exceed transitory losses such that transitory income is on average
positive over time. For low-income groups, transitory losses exceed transitory gains, while for middle -income
groups the value of transitory income is equal to zero over time such that observed and permanent income

take the same value.

over the long run, income variation is due mainly if not solely to variation in permanent income, which implies

a stableAPC
In general, Friedman's permanent income hypothesis offered another way of explaining the conflicting results
of early empirical studies on consumption. It distinguished between a short-run and a long-run consumption
function . The long-run function was essentially a proportional relation while the short-run function wa s a non-

proportional one.
The criticisms of the life-cycle theory involving the notions of liquidity constraints and of the observability and
measurability of permanent income also apply to Friedman's theory.

Random Walk Hypothesis

conte~porary mainstream theories of consumption functions are essentially extensions of the life-cycle and of
the permanent income theories. They are also based on Fisher's intertemporal choice model. The new el ement
is the assumption of rational expectations: people use all available information to forecast future variables

likeincome.
One indicative example of such theories is Robert Hall's (1978) "random walk theory of consumption". The
rational expectations assumption means that consumption should follow a random walk: changes in
consumption should be unpredictable. It also means that a change in income or wealth that was anticipated
has already been factored into expected permanent income, so that it will not change consumption .

Ther efore, only unanticipated changes in income or wealth that change expected permanent income will
change consumption. In the rational expectations framework, agents an_ticipate the future and therefore make
all the required adjustments in the current period. The equation for future consumption is

C,+1 = C,+Q,+1
In. this equation, Qt+l is a rational expectations error that cannot be predicted with any informat·10n known at
time t . All _future information is reflected in current consumption, Ct. The random-wa lk characteristic of
consumption is seen by writingC,+ 1 -C, = Q,+ 1
Con sumption is a random walk, as changes over time are unforeseeable.

pg. J O
 ~--- -----------.:9111
Clea r ly, th e po li cy imp lications of these models are that pol icy changes will affect consumption only if th ey are
n ot antici pated . These arguments greatly diminish the Keyn esian case for government intervention, given that
stabili zati on policies can not be applied in any systematic way.
If co n sumers obey per m an ent income hypothesis & have rationa l expect ations, then only unexpected policy
ch a nges influence co nsu mption . These poli cy changes take effect when they change expectations.