Demand and Slutsky Equation

Varian: Intermediate Microeconomics, 8e, Chapters 6 and 8

() 1 / 51
Demand

The first part of the lecture explains

• what does the consumer’s demand
function for a good depend on,
• what are normal and inferior goods,
ordinary and Giffen goods, and
substitutes and complements,
• what is the inverse demand
function.

() 2 / 51
Introduction

Demand functions — relate prices and income to

choices:
x1 = x1 (p1 , p2 , m)
x2 = x2 (p1 , p2 , m)

Comparative statics – comparing two situations

before and after a change without being
concerned about any adjustment process.

Comparative statics in consumer theory – how

demand changes when income and prices change.

() 3 / 51
Changes in Income
A rise in income shifts the budget line out.
Normal good – increase in income increases demand (Figure 6.1):

∆x1
> 0.
∆m
Inferior good – increase in income decreases demand (Figure 6.2):

∆x1
< 0.
∆m
Examples of inferior goods: nearly any kind low quality good like gruel,
bologna, shacks.

() 4 / 51
() 5 / 51
() 6 / 51
Changes in Income (cont’d)
As income changes, the optimal choice moves along the income offer
curve or income expansion path.

The relationship between the optimal choice and income, with prices fixed,
is called the Engel curve.

() 7 / 51
Examples: Perfect Substitutes
Suppose that p1 < p2 : Consumer is specializing in consuming good 1 =⇒
horizontal income offer curve.

Demand for good 1 is x1 = m/p1

Engel curve is a straight line: m = p1 x1 .

() 8 / 51
Examples: Perfect Complements

Demand for good 1 is x1 = m/(p1 + p2 ).

Engel curve is a straight line: m = (p1 + p2 )x1 .

() 9 / 51
Examples: Cobb-Douglas Preferences
For u(x1 , x2 ) = x1a x21−a , the demand for good 1 x1 = am/p1
and the demand for good 2 is x2 = (1 − a)m/p2 .
(1−a)p1
The income offer curve is a straight line: x2 = ap2 x1 .

Engel curve is a straight line: m = p1 x1 /a.

() 10 / 51
Examples: Homothetic Preferences
The consumer has homothetic preferences, if the demand for good goes
up by the same proportion as income. Or if (x1 , x2 )  (y1 , y2 ), then
(tx1 , tx2 )  (ty1 , ty2 ) for any positive value of t.

If the consumer has homothetic preferences, the income offer curves and
Engel curves are straight lines. Thus perfect substitutes, perfect
complements and Cobb-Douglas are homothetic preferences.

Homothetic preferences are not very realistic. The other options:

• luxury good – demand increases by a greater proportion than income.
• necessary good – demand increases by a lesser proportion than
income.

() 11 / 51
Examples: Quasilinear Preferences
The indifference curves shift in a parallel way. =⇒ It is tangent to the
budget line at a bundle (x1∗ , x2∗ ), then another indifference curve must be
tangent at (x1∗ , x2∗ + k) for any constant k.

Real-life example: choice between single good that is a small part of the
consumer’s budget (e.g. salt or toothpaste) and all other goods.

() 12 / 51
Changes in Price
Changes in price lead to a tilts or pivots of the budget line.
Ordinary good – decrease in price increases demand (see Figure 6.9):

∆x1
< 0.
∆p1
Giffen good – decrease in price decreases demand (see Figure 6.10):

∆x1
> 0.
∆p1
Intuituion behind Giffen good: Suppose you consume a little of meat and a
lot of potatoes. A reduction of price of potatoes gives you extra money for
meat. Then you might need less potatoes.

() 13 / 51
() 14 / 51
() 15 / 51
Changes in Prices (cont’d)
As price changes the optimal choice moves along the price offer curve.

The relationship between the optimal choice and a price, with income and
the other price fixed, is called the demand curve.

() 16 / 51
Examples: Perfect Substitutes
The demand function for good 1 is


 m/p1 when p1 < p2 ;
x1 = any number between 0 and m/p1 when p1 = p2 ;
0 when p1 > p2 .


() 17 / 51
Examples: Perfect Complements
Consumed in fixed proportions =⇒ price offer curve is a straight line.

If consumed in 1:1 proportion, the demand for good 1 is given by

m
x1 = .
p1 + p2

() 18 / 51
Examples: Discrete Goods
Reservation price rn – price where consumer is just indifferent between
consuming and not consuming unit n of good.

For the price r1 the utility from 0 and 1 units is the same:
u(0, m) = u(1, m − r1 ).

For the price r2 , u(1, m − r2 ) = u(2, m − 2r2 ).

() 19 / 51
Examples: Discrete Goods (cont’d)
Special case: quasilinear preferences u(x1 , x2 ) = v (x1 ) + x2 .

u(0, m) = u(1, m − r1 ) can be written as v (0) + m = v (1) + m − r1 .

u(1, m − r2 ) = u(2, m − 2r2 ) as v (1) + m − r2 = v (2) + m − 2r2 .

If v (0) = 0, then v (0) + m = v (1) + m − r1 ⇐⇒ r1 = v (1).

Similarly, v (1) + m − r2 = v (2) + m − 2r2 ⇐⇒ r2 = v (2) − v (1).

Reservation prices just measure marginal utilities.

() 20 / 51
Substitutes and Complements
Blue and red pencils are perfect subsitutes, what about pencils and pens?
Substitutes – increase in p2 increases demand for x1 :
∆x1
> 0.
∆p2

Left and right shoes are perfect complements, what about shoes and
socks?
Complements – increase in p2 decreases demand for x1 :
∆x1
< 0.
∆p2

() 21 / 51
The Inverse Demand Curve
Usually think of demand curve as measuring quantity as a function of price
– e.g. the Cobb-Douglas demand for good 1 is x1 = am/p1 .

We can also think of price as a function of quantity = the inverse demand

function. The Cobb-Douglas indirect demand function is p1 = am/x1 .

() 22 / 51
The Inverse Demand Curve: Economic Interpretation

If both goods are consumed in positive amounts,

p1
|MRS| = ⇐⇒ p1 = p2 |MRS|.
p2
Price of good 1 is proportional to the absolute value of MRS between
good 1 and 2.

Suppose p2 = 1, then p1 measures how much of good 2 would the

consumer trade for one additional unit of good 1.

If good 2 is money, p1 measures marginal willingness to pay:

How many dollars is the consumer willing to give up for one additional unit
of good 1.

() 23 / 51
Summary

• The consumer’s demand function for a good

depends on the prices of all goods and income.
• Normal good – demand increases when income
increases X inferior good
• Ordinary good – demand decreases when its
price increases X Giffen good
• Good 1 is substitute (complement) for good 2 if
the demand for good 1 increases (decreases)
when the price of good 2 increases.
• The height of the inverse demand function at
any given level of consumption measures the
marginal willingness to pay for an additional
unit of good.

() 24 / 51
Slutsky Equation

The second part of the lecture explains

• what are the substitution and income effects,
• what is the Slutsky equation,
• what does the Law of Demand say.

() 25 / 51
Slutsky Equation
We want a way to decompose the effect of a price change into “simpler”
pieces.
That’s what analysis is all about – break up into simple pieces to
determine behavior of whole.
With this analytical tool, we will be able to answer the following questions:
• Does a reduction in price always increase the demand for the good?
What if we have a Giffen good?
• Does an increase in wage induce people to work more? What if the
wage increases from $10 to $100 an hour? What if it increases from
$100 to $10.000 an hour?

() 26 / 51
Slutsky Equation (con’t)
Break up price change into a pivot and a shift.
These are hypothetical changes. We can examine each change in isolation
and look at sum of two changes.

() 27 / 51
The Pivoted Line
The pivoted line has the same slope (relative prices) like the final budget
line, but the purchasing power is adjusted so that the original consumption
bundle (x1 , x2 ) is just affordable.

How much we have to adjust the income to keep (x1 , x2 ) just affordable?

() 28 / 51
The Pivoted Line (con’t)
Let m0 be the income that will make (x1 , x2 ) just affordable and p10 the
final price of good 1. Then
m0 = p10 x1 + p2 x2 .
m = p1 x1 + p2 x2 .

Substracting the second equation from the first, we have

m0 − m = x1 [p10 − p1 ].

Using ∆m = m0 − m and ∆p1 = p10 − p1 , we get

∆m = x1 ∆p1 .

Example: 20 candy bars, price increase from 50 to 60 cents a piece: 20

candy bars are just affordable if the income increases by

()
∆m = 20 × 0.1 = $2.00. 29 / 51
Slutsky Substitution Effect

Substitution effect is change in demand due to pivot.

Slutsky substitution effect measures how demand changes when we change

prices, keeping purchasing power fixed:
∆x1s = x1 (p10 , m0 ) − x1 (p1 , m).

Sometimes called the change in compensated demand: The consumer is

compensated for the effect of a price change by ∆m.

() 30 / 51
() 31 / 51
Sign of the Substitution Effect

Substitution effect is negative – “negative” means quantity moves

opposite the direction of price.

The bundles on the pivoted budget line to the left of X (with less of
good 1 than x1 ) were affordable at the old prices.

If the consumer chooses the best bundles he can afford, then X must be
preferred to all the bundles on the pivoted line inside the original budget
set (revealed preference).

Similarly we could prove that if p1 increases, x1 goes down.

() 32 / 51
Example: Calculating the Substitution Effect

Suppose m = $120, p1 = $3 per quart and the demand for milk is

m
x1 = 10 + .
10p1
The demand for milk is 10 + 120/(10 × 3) = 14 quarts per week.

Now the price of milk falls to p10 = $2 per quart. The demand for milk is
10 + 120/(10 × 2) = 16 quarts per week.

In order to calculate the substitution effect, we need to calculate the

income necessary to keep the purchasing power constant
m0 = m + ∆m = m + x1 ∆p1 = 120 + 14 × (2 − 3) = 106.

The substitution effect is

∆x1s = x1 (p10 , m0 ) − x1 (p1 , m) = x1 (2, 106) − x1 (3, 120) = 1.3.

() 33 / 51
Income Effect

Income effect is change in demand due to shift.

This measures the change in demand for good 1 when we change the
income from m0 to m and keep the prices constant at (p10 , p2 ):
∆x1n = x1 (p10 , m) − x1 (p10 , m0 ).

Income effect is negative for normal goods (a rise in price reduces the
income which reduces the demand) and positive for inferior good.

() 34 / 51
() 35 / 51
The Total Change in Demand

Total change in demand is substitution effect plus the income effect. This
equation is called Slutsky indentity:

∆x1 = ∆x1s + ∆x1n .

x1 (p10 , m) − x1 (p1 , m) = x1 (p10 , m0 ) − x1 (p1 , m) + x1 (p10 , m) − x1 (p10 , m0 ).

It is an identity – the first and fourth terms cancel out.

If good is normal, the substitution effect and the income effect are
negative. The total effect is negative.

If good is inferior, the substitution effect is negative and the income effect
is positive. The total effect is ambiguous (see Figure 8.3).

() 36 / 51
() 37 / 51
The Rates of Change
If we want to express Slutsky identity in terms rates of change, it is
convenient to define negative income effect:
∆x1m = x1 (p10 , m0 ) − x1 (p10 , m) = −∆x1n .

Then the Slutsky identity is

∆x1 = ∆x1s − ∆x1m .

Dividing the identity by ∆p1 , we get

∆x1 ∆x1s ∆x1m

= − .
∆p1 ∆p1 ∆p1
∆m
From ∆m = x1 ∆p1 , we get ∆p1 = x1 . Subsituting into (38), we get

∆x1 ∆x1s ∆x1m

= − x1 .
∆p1 ∆p1 ∆m

() 38 / 51
The Law of Demand
The Law of Demand – If the demand for a good increases when income
increases, then the demand for that good must decrease when its price
increases.
Follows directly from the Slutsky equation:
• If demand increases when income increases, it is a normal good.
• If we nave a normal good, then the total effect of a price change is
negative.

() 39 / 51
Examples: Perfect Complements
Substitution effect is zero: The entire change in demand is due to income
effect.

() 40 / 51
Examples: Perfect Substitutes
Income effect is zero: The entire change in demand is due to substitution
effect.

() 41 / 51
Examples: Quasilinear Preferences
Quasilinear preferences imply that shift in income causes no change in
demand: The entire change in demand is due to substitution effect.

() 42 / 51
Application: Rabating a Tax
In 1974, OPEC instituted an oil embargo against the US and was able to
stop oil shipments for several weeks.

Many plans to reduce US dependency on foreign oil.

One of the plans:

• put a tax on gasoline, in order to reduce the consumption of oil
• return the revenues, so that consumers are not hit by the tax.

Critics: Consumers will use the rebated money to purchase more gasoline.

What does economic analysis say about the plan?

() 43 / 51
Application: Rabating a Tax (cont’d)
Original budget constraint: px + y = m

After tax budget constraint: (p + t)x 0 + y 0 = m + tx 0 ⇐⇒ px 0 + y 0 = m

So (x 0 , y 0 ) was affordable originally and rejected in favor of (x, y ) =⇒

consumer must be worse off.

The consumer will choose less gasoline and more of other goods (see
Figure 8.7).

() 44 / 51
() 45 / 51
Application: Voluntary Real Time Pricing
Electricity producers have extreme capacity problem: easy to produce up
to capacity, impossible to produce more.
Increasing capacities is expensive =⇒ reducing the use of electricity
during peak demand is attractive.
Peak demand due to weather is easy to forecast (in hot weather in Georgia
– 30 % of usage due to air conditioning).
How to set up pricing system so that people have incentives to reduce
consumption in peak demand.

() 46 / 51
Application: Voluntary Real Time Pricing (cont’d)
One solution: Real Time Pricing (RTP)

RTP: industrial users equipped with special meters – the price of

electricity can vary from minute to minute depending on the total demand
of electricity.

Georgia Power Company – largest RTP program in the world: In 1999

reduced demand by 750 MW on high-price day by inducing some large
customers to cut their demand by as much as 60 %.

() 47 / 51
() 48 / 51
The Hicks Substitution Effect
Hicks substitution effect a change in demand for a good due to a change
in the relative prices while keeping the utility of the consumer constant.

() 49 / 51
The Sign of Hicks Substitution Effect
Negative – proof by revealed preferences:

(x1 , x2 ) is demanded bundle at prices (p1 , p2 ) and (y1 , y2 ) is demanded

bundle at prices (q1 , q2 ).

Since (x1 , x2 ) ∼ (y1 , y2 ), it is not true that

p1 x1 + p2 x2 > p1 x1 + p2 x2 and q1 y1 + q2 y2 > q1 y1 + q2 y2 .

Hence, these inequalities are true:

p1 x1 + p2 x2 ≤ p1 x1 + p2 x2 and q1 y1 + q2 y2 ≤ q1 y1 + q2 y2 .

Adding and rearrangig the inequalities gives

(q1 − p1 )(y1 − x1 ) + (q2 − p2 )(y2 − x2 ) ≤ 0.

Using p2 = q2 (change only in p1 ), we get (q1 − p1 )(y1 − x1 ) ≤ 0.

The change in price and quantity demanded must have the opposite sign.
() 50 / 51
Summary

• There are two effects of a change in price:

substitution and income effect.
• The substitution effect is the change in
demand due to the change in relative prices
when purchasing power is held constant.
• The income effect is the change in demand due
to the change in purchasing power.
• Slutsky equation says that the total change in
demand is the sum of the substitution and the
income effect.
• The Law of Demand says that normal goods
must have downward-sloping demand.

() 51 / 51