Professional Documents
Culture Documents
Income & Subsitution Effects PDF
Income & Subsitution Effects PDF
I2
py Successive optima when I
I1 I2 > I1 changes will have same MRS
py
because the slope of the budget
constraint remains the same.
I1 I2 x
px px
x* x**
Engel Curves show how x* or y* changes
as I changes, Ceteris Paribus.
y
* y*
y I3 > I2 > I1
y * f(I)
y*
*
y
y*
y*
* 0
y I1 I2 I3 x I1I 2 I3 I or E
px px px
Line tangent to
f(I) at point B.
* Line tangent to
x Line tangent to
y f(I)
*
y*
f(I) at point A.
f(I) at point A.
x* f(I)
A
B A
0 0
I or E I or E
If a line tangent to f(I) at point A has a negative If a line tangent to f(I) at point A has a positive
intercept, expenditures on the good (pxx) intercept, expenditures on the good (pyy)
decrease as a percentage of total expenditures increase as a percentage of total expenditures
(E) as I increases around point A. (E) as I increases around point A.
If a line tangent to f(I) at point B has a zero intercept, expenditures on the good (pxx) remain
constant as a percentage of total expenditures (E) as I increases around point B.
Normal and Inferior Goods
y*
y* Mostly we talk about 0.
Normal good 0 I
I
y*=f(I) Luxury, e.g., eating out, boats
y*
Curves may
bend up for
luxury goods
(fI I > 0) and Necessity, e.g., underwear
down for
necessities
(fI I < 0).
I
Both luxuries and necessities are Normal Goods.
x *
Inferior good 0
I
e.g., macaroni & cheese y
x
x* I1 > I0
X**
X=f(I) I1
0 X** x* I0
I0 I1 I x
income effect. The substitution effect is
B the change in x* in going from A to C,
A while the income effect is the change in
C U2 x* in going from C to B.
U1 To find C, use the original indifference
0 I I x curve and find the point of tangency
p x with a fictitious budget constraint that
S In p x
has the new price ratio.
Tot x *
What kind of good is x with regard to income? It is a normal good because 0.
I
• Summary of substitution and income effects – The
movement from A to B is composed of two effects:
– Substitution effect - Caused by change in px/py| U = U1.
• Because px is lower, the price ratio is smaller and the new
tangency point must be at a smaller MRS (smaller –dy /dx).
• Can measure the substitution effect by holding real income
constant (hold U constant), remaining on the same Indifference
Curve, but using the new price ratio to find point C. The change
in x* in going from A to C measures the substitution effect.
• For a given price change (px), the magnitude of the substitution
effect depends on the availability of substitute goods.
y y No substitution effect
Large substitution effect
U = U1
U = U1 U = U0
U = U0
x x
– Income Effect – Caused by an increase in real income
represented by an increase in utility, or movement to a higher
indifference curve. A price decrease brings about an increase in
real income; purchasing power. The income effect is the change
in x* in going from C to B. The magnitude of the income effect
depends on the portion of income spent on x.
• The sum of the Income and Substitution Effects is the total effect of a
price change (total change in x*).
• Could show a similar analysis for a price increase (text p. 127).
• In most situations, the two effects are complementary, in that they move in
the same direction and reinforce each other as in the case of Normal
Goods.
x * x *
0 0
p x U* I
The increase in nominal income
For a decrease in px , the substitution
required to reach the higher
effect causes x* to increase holding
indifference curve (higher real income
real income constant (have to reduce
resulting from the lower price), causes
nominal income to keep real income
x* to increase further.
constant).
Income and Substitution Effects for
Inferior Goods
x *
• Increasing income causes a decline in purchases of x, 0.
I
The income effect is perverse for an inferior good.
A C Gives Substitution
B
A
effect x
*
0.
p x
U2 U U1
C
x *
U1 CB Gives Income effect 0.
I
x
In this case, a decrease in px causes an increase in x*, but only
Tot In
because the substitution effect is large enough to offset the
S perverse income effect.
The substitution effect is always (for typical
Indifference Curves) opposite the price movement!
• The income effect generally cannot be classified. The
income effect is opposite the price movement for a
normal good and in the same direction as the price
movement for an inferior good.
In a two-good world:
Demand functions x* x(px , py , I) Traditional or
Marshallian
y* y(px , py , I) demand functions
px
Could hold py, preferences, x* x(px p y , I)
and I (nominal) constant and
vary px to get typical demand
x(px p y , I)
curve in px and x space.
x
One person, so indifference curves can’t cross.
y
px px px
Per unit Shows that successive price
of time
declines result in larger
optimal quantities of x (except
for Giffen goods).
0 x* I x
I
I x/unit of time
px px x * px
px
Shows how x* changes as px
px
px
changes when nominal income
(I) and prices of all other goods
x x(p x p y , I)
px
x *
0
(py) are constant and when the
individual’s preference system
p x
constant. This is a Marshallian
x(px p y , I) demand curve (uncompensated
0
x* x *' x x/unit of time demand curve).
•Shifts in the demand curve are caused by shifts in
the individual’s preferences or utility function
(shifts in the indifference curves implying changes
in the MRS), the prices of other goods, and
nominal income.
Increase I, the demand curve for x shifts up for normal good (down
for inferior good).
Increase py, the demand curve for x shifts up for substitute good
(down for a complement good).
Increase MUx relative to MUy, the demand curve for x shifts up.
At any x, MRSxy is larger. The absolute slope of the indifference
curve is larger (steeper indifference curve at any level of x).
• Changes in px cause movements along the curve, i.e.,
“changes in the Quantity Demanded” rather than shifts
in demand.
Compensated Demand Curves (Hicksian
Demand Curve) – We could eliminate the income effects of changes
in px and show the effects on x*, holding utility or real income constant.
y
p x p x p x p x
Income compensated demand curve
(Hicksian) shows only the substitution
U1
effects of changes in px, while py,
preferences and utility (real income)
px
x/u.t.
are held constant.
px
xc (px py , U1)
'
p x
Means real income constant!
p ''
x
p '''
x Compensated demand curve always has a
negative slope because the substitution effect
is always negative if MRS is diminishing.
(x c /p x 0 always if MRSxy is diminishing)
0 x/u.t.
Compensated demand curve is steeper
(flatter) than the Marshallian demand
curve for a normal (inferior) good.
px For px > px*, the income effect causes less (more)
consumption for a normal good (an inferior good),
so x* decreases more (less) for x than for xc.
px
Beginning optimal point
px *
px x Marshallian
xc Hicksian
x * x c (p x , p y , U) “Hicksian” or “Compensated”
and the ordinary demand function.
nominal income
x x(p x , p y , I)
* “Marshallian”, “Ordinary”, or
“Uncompensated”
•These two demand functions are equal at the
“beginning point” (U max point) where they cross.
xc (px , py , U) x(px , py , I)
Next, remember the Expenditure Function
E* E(p x , p y , U)
Substitute the Expenditure Function into x above and get the
following (because I = E at U max).
x (px , py , U) x(px , py , E(px , py , U))
c
xc
x x E x x c x E
px px E px px px E px
We will explore this relationship on next page.
(-) except for Always (-) (-) for normal goods
Giffen good (+) for inferior goods
Marshallian Hicksian
(total effect) (sub effect) Income effect
x x c
x E
p x p x E p x
This is not the xc
Minus Usually (+); Always (+ or 0)
slope of the px
Marshallian
changes for normal (+), because a decrease
demand curve; it is x sign. for inferior (-) (increase) in px
the inverse of the implies a decrease
slope. The more (increase) in E to
negative ∂x/∂px, px*
maintain same utility
the flatter the
level as before the
demand curve. If x
were on the
px price change. Also,
vertical axis, this the Expenditure
would be the slope x Function is non-
of the demand x x c x E decreasing in prices.
curve.
p x p x E p x
Summary of Equation on Previous Page
• x p x is the gross change in x* in response to a change
in px (total effect). It is the inverse of the slope of the
Marshallian demand curve.
• The first term on the right-hand side is the change in x* in
response to a change in px holding utility (real income)
constant. It is the substitution effect! It is always
negative because MRS is diminishing. It is the inverse of
the slope of the compensated (Hicksian) demand curve.
x E
*
• The second term E p x shows the response of x
to a change in px through the effect of px on income (I=E).
It is the income effect!
• For a normal good, the substitution and income effects
have negative signs; they reinforce each other.
The Slutsky Equation
x c x
Sub. effect = U
(real income) constant
p x p x
x E x E x
Inc. effect = x
E px I px I
E=I
TE x
x(e x,px 1).
p x
So the change in total expenditures on the good is determined
by the price elasticity of demand as follows:
TEx
If elastic (ex,px < -1), then p 0 (ΔTE is opposite the Δp ).
x
x
TEx
If unit elastic (ex,px = -1), then p 0 (ΔTE = 0).
x
TEx
If inelastic ex,px > -1, then 0 (ΔTE is same direction as Δpx).
px
Compensated Demand Elasticities
If the compensate d demand function is given by
x c (p x , p y , U), then the compensate d own price
x p x
c
elasticity is : e c
x, p x c
p x x
and the compensate d cross - price elasticity is :
x p y
c
e x,p y
c
c.
p y x
The relationsh ip between these compensate d
price elasticiti es and Marshalli an price elasticiti es
can be shown by putting the Slutsky Equation in
elasticity form.
Relationships Among Elasticities
Slutsky Equation in Elasticity Form
x x x Original Slutsky
U const x
p x p x I
Multiply by px/x.
x p x p x x x 1
U const p x x
p x x x p x I x
Multiply the second term on the right-hand side by I/I.
x p x p x x p x x x I
U const
p x x x p x I I x
ecx,py
e x,px “Substitution Elasticity”.
The elasticity of the s x e x,I
compensated demand curve.
e x,px e c
x, p x s x e x,I
(-) except (-) for normal or (+)
(-) always
Giffen for inferior good.
f f
Degree of homogeneity
p x1 I
(m=0).
Homogeneity Condition Conclusion
The sum of own-price,
Because demand functions are cross-price, and income
homogeneous of degree zero, we can use elasticities equals zero for
Euler’s theorem on the demand function any specific good. This
for x (uncompensated) to get the following reaffirms that demand is
homogeneous of degree
for a two-good world: zero (equal percentage
x x x
px py I 0 changes in prices and
p x p y I
income leave the quantity
Dividing through by x gives:
demanded unaffected). If
x p x x p y x I you know or can estimate
0 two of the three terms,
p x x p y x I x
you can calculate the
e x, p x e e
x, p y 0
x, I
third. If you had more
than two goods, you
would have several cross-
(-) except (+) for gross (+) for normal
substitutes; (-) good; (-) for price elasticities, some of
Giffen
for gross inferior good which could be < 0.
complements
Engel Aggregation
Assuming a typical consumer (diminishing MRS) and
two goods, the budget constraint is:
p x x * p y y* I at optimallity.
For utility maximization, an increase in I must be
accompanied by an increase in total expenditures
because I = E.
x x(p x , p y , I)
*
x px I py
b ex,px b , ex,I c , ex,py d .
p x x x x
Elasticities are typically evaluated at the means of x, px, py, and I.
ex,px is not constant along the demand curve. x
p x
is
constant, but px/x is not constant!
px
e x, p x 1
= e x, p x 1 (between –1 and 0)
px
Inelastic
x small
x
=
Constant Elasticity Demand Functions
x ap I p b c d
x y
Cobb-Douglas type
x p x b 1 c d px
e x,px abp x I p y b c d b
p x x ap x I p y
Or, ln x ln a b ln p x c ln I d ln p y
px
So ex,p b everywhere on the curve. The
x
same property holds for other elasticities
(income and cross-price elasticities).
x
Consumer Surplus
Use individual’s x curve --- Consumer surplus is
px
“choke price” the area under the x
pChx curve, above px0. It
market Area p0x ApChx CS is the amount of
price
extra expenditures an
p0x A individual would be
willing to make
above what he/she
x
has to make to get
0 x each unit of the good.
x10
• The concept of consumer surplus is used to
evaluate the effects of price changes on
consumers. If the price goes from px0 to px1,
the consumer will suffer a net loss of
consumer surplus equal to the area Px0 ABPx1.
px 1 1 0
(p p )x (px px )(x* x** ) net loss
1
x
0
x
**
2
For linear in consumer
demand p1
x
B surplus caused by
function 0 A the price of x
px
x increasing from
0 ** * x px0 to px1.
x x
• We will use the compensated demand curve
and the expenditure function to illustrate the
measurement of consumer surplus.
E1 E(p , py , U0 )
1
x
Expenditures at px1 , given U0 and py.
p1x p0x
p1x The change in consumer surplus is
slope E0–E1, which is the negative of the
y py
slope
p0x
change in expenditures required to
maintain the same level of utility
when the price changes. For an
py
increase in px, E0–E1<0, meaning
U0 real income has fallen and nominal
E1 E0 income has to increase to maintain
x U0. For a decrease in px, E0–E1>0,
E 0 E1 0 meaning real income has increased
and nominal income has to decrease
to maintain U0.
• We can differentiate the expenditure function
to get the compensated demand function.
E As shown earlier by
xc xc (px , py , U0 ) envelope theorem,
px
E p x is the change in consumer surplus
“Shephard’s Lemma.”
for a very small change in px.
The instantaneous change in E resulting from a very
small change in px is equal to xc (quantity demanded
on the compensated demand curve).
Because the change from px0 to px1 covers some
distance, must integrate x c (px , p y , U0 ) to get
the change in consumer surplus.
p1x
ΔCS
c
x (p x , p y , U 0 )dp x
p 0x
1
B
p x
0 CS A
p x
xc
0 x
x1 x0
• The compensated demand function gives the
“true” estimate of the change in consumer surplus
resulting from a price change, but it is difficult to
estimate, and once estimated, the estimate of
consumer surplus is usually quite similar to the
estimate obtained from the Marshallian demand
curve, so using the Marshallian demand curve is
more practical in the real world.