Why do we have to pay attention
Consumption
to consumption function?
an introduction to the most prominent work on Crucial for Long-run analysis.
consumption, including:
Determine economic growth
John Maynard Keynes: consumption and current
income Crucial for Short-run analysis
Irving Fisher: intertemporal choice Determine AD → key element of economic
Franco Modigliani: the life-cycle hypothesis fluctuation.
Milton Friedman: the permanent income hypothesis
Robert Hall: the random-walk hypothesis
David Laibson: the pull of instant gratification
slide 0 slide 1
The Keynesian consumption
Keynes’s conjectures
function
1. 0 < MPC < 1 C
2. Average propensity to consume (APC )
C = C + cY
falls as income rises.
(APC = C/Y )
c c = MPC
3. Income is the main determinant of 1
= slope of the
consumption. consumption
C function
Y
slide 2 slide 3
The Keynesian consumption Early empirical successes:
function Results from early studies
As
As income
income rises,
rises, consumers
consumers save
save aa bigger
bigger Households with higher incomes:
C fraction
fraction of
of their
their income,
income, so APC falls.
so APC falls.
consume more, ⇒ MPC > 0
C = C + cY save more, ⇒ MPC < 1
save a larger fraction of their income,
⇒ APC ↓ as Y ↑
C C Very strong correlation between income and
APC = = + c
Y Y consumption:
slope = APC ⇒ income seemed to be the main
Y determinant of consumption
slide 4 slide 5
1
Problems for the
The Consumption Puzzle
Keynesian consumption function
Based on the Keynesian consumption function, C
Consumption function
economists predicted that C would grow more from long time series
data (constant APC )
slowly than Y over time.
This prediction did not come true:
As incomes grew, APC did not fall, Consumption function
and C grew at the same rate as income. from cross-sectional
Simon Kuznets showed that C/Y was household data
very stable in long time series data. (falling APC )
Y
slide 6 slide 7
Irving Fisher and Intertemporal
The basic two-period model
Choice
Period 1: the present
The basis for much subsequent work on
consumption. Period 2: the future
Assumes consumer is forward-looking and Notation
chooses consumption for the present and future Y1, Y2 = income in period 1, 2
to maximize lifetime satisfaction.
C1, C2 = consumption in period 1, 2
Consumer’s choices are subject to an S = Y1 − C1 = saving in period 1
intertemporal budget constraint,
a measure of the total resources available for (S < 0 if the consumer borrows in period 1)
present and future consumption.
slide 8 slide 9
Deriving the intertemporal The intertemporal budget
budget constraint constraint
Period 2 budget constraint: C2 Y2
C1 + = Y1 +
C 2 = Y 2 + (1 + r ) S 1+r 1+r
= Y 2 + (1 + r ) (Y1 − C 1 )
present value of present value of
Rearrange terms: lifetime consumption lifetime income
(1 + r ) C 1 + C 2 = Y 2 + (1 + r )Y1
Divide through by (1+r ) to get…
slide 10 slide 11
2
The intertemporal budget The intertemporal budget
constraint constraint
C2 C2 Y2 C2 C2 Y2
C1 + = Y1 + C1 + = Y1 +
1+r 1+r 1+r 1+r
The
The slope
slope of
of
(1 + r )Y1 +Y 2 the
the budget
budget
Consump =
Saving line
line equals
equals 1
income in
The
The budget
budget −(1+r
−(1+r )) (1+r )
both periods
constraint
constraint shows
shows
all
all combinations
combinations Y2 Y2
of
of CC11 and C22 that
and C that Borrowing
just
just exhaust
exhaust the
the
consumer’s
consumer’s C1 C1
Y1 Y1
resources.
resources.
Y1 +Y 2 (1 + r )
slide 12 slide 13
Consumer preferences Consumer preferences
Higher
Higher The
C2 C2 The slope
slope of
of
An indifference indifference
indifference Marginal rate of an
an indifference
indifference
curve shows curves
curves substitution (MRS ): curve
curve atat any
any
all combinations represent
represent the amount of C2 point
point equals
equals
of C1 and C2 higher
higher levels
levels the consumer the
the MRS
MRS
that make the of
of happiness.
happiness. would be willing to 1 at
at that
that point.
point.
consumer substitute for MRS
equally happy. IC2 one unit of C1.
IC1 IC1
C1 C1
slide 14 slide 15
Optimization How C responds to changes in Y
C2 C2 An
An increase
increase
The optimal (C1,C2) Results:
At
At the
the optimal
optimal point,
point, in Y11 or
in Y or YY22
is where the Provided they are
MRS == 1+r
MRS 1+r shifts
shifts the
the
budget line both normal goods,
budget
budget line
line
just touches C1 and C2 both
outward.
outward.
the highest increase,
indifference curve. O …regardless of
whether the
income increase
occurs in period 1
C1 or period 2. C1
slide 16 slide 17
3
Keynes vs. Fisher How C responds to changes in r
C2
Keynes: As depicted here, An
An increase
increase in in rr
Current consumption depends only on pivots
pivots the
the budget
budget
C1 falls and C2 rises.
line
line around
around thethe
current income. However, it could point
point (Y
(Y11,Y
,Y22).).
Fisher: turn out differently… B
Current consumption depends only on
the present value of lifetime income. A
The timing of income is irrelevant Y2
because the consumer can borrow or lend
C1
between periods. Y1
slide 18 slide 19
How C responds to changes in r Constraints on borrowing
income effect: If consumer is a saver, In Fisher’s theory, the timing of income is irrelevant:
the rise in r makes him better off, which tends to Consumer can borrow and lend across periods.
increase consumption in both periods. Example: If consumer learns that her future income
will increase, she can spread the extra consumption
substitution effect: The rise in r increases
over both periods by borrowing in the current period.
the opportunity cost of current consumption,
which tends to reduce C1 and increase C2. However, if consumer faces borrowing constraints
(aka “liquidity constraints”), then she may not be
Both effects ⇒ ↑C2. able to increase current consumption
Whether C1 rises or falls depends on the relative …and her consumption may behave as in the
size of the income & substitution effects. Keynesian theory even though she is rational &
forward-looking.
slide 20 slide 21
Constraints on borrowing Constraints on borrowing
C2 C2
The borrowing
The budget
constraint takes
line with no
the form: The budget
borrowing
line with a
constraints C1 ≤ Y 1
borrowing
Y2 Y2 constraint
Y1 C1 Y1 C1
slide 22 slide 23
4
Consumer optimization when the Consumer optimization when the
borrowing constraint is not binding borrowing constraint is binding
C2 C2
The borrowing The optimal
constraint is not choice is at
point D.
binding if the
consumer’s But since the
optimal C1 consumer
is less than Y1. cannot borrow, E
the best he can D
do is point E.
Y1 C1 Y1 C1
slide 24 slide 25
The Life-Cycle Hypothesis The Life-Cycle Hypothesis
due to Franco Modigliani (1950s) The basic model:
W = initial wealth
Fisher’s model says that consumption depends
Y = annual income until retirement
on lifetime income, and people try to achieve
(assumed constant)
smooth consumption.
R = number of years until retirement
The LCH says that income varies systematically T = lifetime in years
over the phases of the consumer’s “life cycle,”
and saving allows the consumer to achieve Assumptions:
smooth consumption. zero real interest rate (for simplicity)
consumption-smoothing is optimal
slide 26 slide 27
Implications of the Life-Cycle
The Life-Cycle Hypothesis
Hypothesis
Lifetime resources = W + RY The LCH can solve the consumption puzzle:
To achieve smooth consumption, The life-cycle consumption function implies
consumer divides her resources equally over time: APC = C/Y = α(W/Y ) + β
C = (W + RY )/T , or
Across households, income varies more than
C = αW + βY wealth, so high-income households should have
where a lower APC than low-income households.
α = (1/T ) is the marginal propensity to Over time, aggregate wealth and income grow
consume out of wealth together, causing APC to remain stable.
β = (R/T ) is the marginal propensity to consume
out of income
slide 28 slide 29
5
Implications of the Life-Cycle
Hypothesis
$
Unlike Life-Cycle Hypothesis which
The
The LCH
LCH emphasizes that income follows a regular
Wealth
implies
implies that
that pattern over a person’s life,
saving
saving varies
varies
systematically
systematically Income
Permanent Income Hypothesis emphasizes
over
over aa that people experience random and temporary
person’s
person’s Saving changes in income from year to year.
lifetime.
lifetime. Consumption Dissaving
Retirement End
begins of life
slide 30 slide 31
The Permanent Income Hypothesis The Permanent Income Hypothesis
due to Milton Friedman (1957) Consumers use saving & borrowing to smooth
Y = YP + YT consumption in response to transitory changes
in income.
where
Y = current income The PIH consumption function:
Y P = permanent income C = αYP
average income, which people expect to
persist into the future where α is the fraction of permanent income
that people consume per year.
Y T = transitory income
temporary deviations from average income
slide 32 slide 33
The Permanent Income Hypothesis PIH vs. LCH
The PIH can solve the consumption puzzle: Both: people try to smooth their consumption
The PIH implies in the face of changing current income.
APC = C/Y = α Y P/Y LCH: current income changes systematically
If high-income households have higher transitory as people move through their life cycle.
income than low-income households,
APC is lower in high-income households. PIH: current income is subject to random,
transitory fluctuations.
Over the long run, income variation is due mainly
(if not solely) to variation in permanent income, Both can explain the consumption puzzle.
which implies a stable APC.
slide 34 slide 35
6
The Random-Walk Hypothesis The Random-Walk Hypothesis
If PIH is correct and consumers have rational
due to Robert Hall (1978) expectations, then consumption should follow a
based on Fisher’s model & PIH, random walk: changes in consumption should
in which forward-looking consumers base be unpredictable.
consumption on expected future income A change in income or wealth that was
anticipated has already been factored into
Hall adds the assumption of
expected permanent income,
rational expectations,
so it will not change consumption.
that people use all available information
to forecast future variables like income. Only unanticipated changes in income or wealth
that alter expected permanent income
will change consumption.
slide 36 slide 37
The Psychology of Instant
Implication of the R-W Hypothesis
Gratification
Theories from Fisher to Hall assume that
consumers are rational and act to maximize
IfIfconsumers
consumersobey obeythe
thePIH
PIH lifetime utility.
and
andhavehaverational
rationalexpectations,
expectations, Recent studies by David Laibson and others
then
thenpolicy
policychanges
changes consider the psychology of consumers.
will affect consumption
will affect consumption
only
only ififthey
they are
areunanticipated.
unanticipated.
slide 38 slide 39
The Psychology of Instant Two questions and time
Gratification inconsistency
Consumers consider themselves to be imperfect 1. Would you prefer (A) a candy today, or
decision-makers. (B) two candies tomorrow?
In one survey, 76% said they were not saving 2. Would you prefer (A) a candy in 100 days, or
enough for retirement. (B) two candies in 101 days?
Laibson: The “pull of instant gratification” In studies, most people answered (A) to 1 and (B) to 2.
explains why people don’t save as much as a A person confronted with question 2 may choose (B).
perfectly rational lifetime utility maximizer would But in 100 days, when confronted with question 1,
save. the pull of instant gratification may induce her to
change her answer to (A).
slide 40 slide 41
7
Summing up Chapter Summary
Keynes: consumption depends primarily on 1. Keynesian consumption theory
current income. Keynes’ conjectures
Recent work: consumption also depends on MPC is between 0 and 1
APC falls as income rises
expected future income
current income is the main determinant of
wealth current consumption
interest rates Empirical studies
Economists disagree over the relative importance in household data & short time series:
of these factors, borrowing constraints, confirmation of Keynes’ conjectures
and psychological factors. in long-time series data:
APC does not fall as income rises
slide 42 slide 43
Chapter Summary Chapter Summary
2. Fisher’s theory of intertemporal choice 3. Modigliani’s life-cycle hypothesis
Consumer chooses current & future Income varies systematically over a lifetime.
consumption to maximize lifetime satisfaction of Consumers use saving & borrowing to smooth
subject to an intertemporal budget constraint. consumption.
Current consumption depends on lifetime Consumption depends on income & wealth.
income, not current income, provided consumer
can borrow & save.
slide 44 slide 45
Chapter Summary Chapter Summary
4. Friedman’s permanent-income hypothesis 5. Hall’s random-walk hypothesis
Consumption depends mainly on permanent Combines PIH with rational expectations.
income. Main result: changes in consumption are
Consumers use saving & borrowing to smooth unpredictable, occur only in response to
consumption in the face of transitory fluctuations unanticipated changes in expected permanent
in income. income.
slide 46 slide 47
8
Chapter Summary
6. Laibson and the pull of instant gratification
Uses psychology to understand consumer
behavior.
The desire for instant gratification causes
people to save less than they rationally know
they should.
slide 48
