MBA 773, Weeks 3&4:

Consumer & Cost Theory

Professor Jesse Davis, Fall 2022
Agenda: Weeks 3 & 4
1. Consumer Preferences
Consumer Objective Function/Optimal Bundle
2. Impact of Uncertainty
Expected Utility + Risk Tolerance/Risk Premia
3. Firm Production
What costs matter? Fixed v. Marginal v. Average Costs
4. Short-Run v. Long-Run
How do we select “fixed” aspects of production?
5. Microeconomics in the News (Wks 3&4)

1
Consumer preferences
1. Completeness
– Preferences are well-defined: we can rank (bundles of) goods
– Real-life complexity → computational “impossibility”
2. Transitive
– Preferences don’t depend upon the comparison
– Framing can induce “preference reversals”
3. Monotone
– More is better
– Forced consumption or storage can make a “good” a “bad”
4. Diminishing Marginal Utility (risk-aversion)
– Each additional unit is worth less
– Gambling, risk-seeking behavior

2
Optimal Bundles
Consumers maximize their utility subject to a budget constraint.
• Suppose there are just two goods, x and y:
max 𝑢(𝑥, 𝑦)
𝑥,𝑦
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜: 𝑝𝑥 𝑥 + 𝑝𝑦 𝑦 ≤ 𝐼

• The “I” represents the money you’ve allocated to spend.

– Borrowing/saving?
– Labor/leisure tradeoff?
• What does it mean if prices aren’t linear (here each unit costs “p”)?
– E.g., quantity discounts

3
Example: optimal bundle
Suppose you buy only two goods: eating at home and eating at a
restaurant. You have the following information:
• Prices: ph = $10 and pr = $25
• Income: $1000
• Utility: 𝑈 ℎ, 𝑟 = 2ln(ℎ) + 4ln(𝑟) (“ln” is the natural log)
2 4
• Marginal Utility: 𝑀𝑈ℎ = , 𝑀𝑈𝑟 =
ℎ 𝑟

What is the optimal split between eating at home and dining out?

4
Example: optimal bundle
1. Determine how much you can eat at home, given your dining out:

10h+25r = 1000
h = 100 – 2.5r

2. Solve for the optimum by equalizing “bang for the buck”:

𝑀𝑈𝑟 4ൗ𝑟
=
𝑝𝑟 25
2
𝑀𝑈ℎ 2ൗℎ 100 − 2.5𝑟
= =
𝑝ℎ 10 10

5
Example: optimal bundle
2. Solve for the optimum by equalizing “bang for the buck”:
2
100 − 2.5𝑟 4ൗ
𝑴𝑼𝒉ൗ 𝑟 𝑴𝑼𝒓ൗ
𝒑𝒉 = 10
=
25
= 𝒑𝒓
50 40
=
100 − 2.5𝑟 𝑟
50𝑟 = 4000 − 100𝑟
4000
𝑟∗ = ≈ 26.67
150
ℎ = 100 − 2.5𝑟 = 100 − 2.5 4000ൗ150 ≈ 33.33 = ℎ∗

6
Example: optimal bundle
How does this change as the value of eating out increases?
• Prices: ph = $10 and pr = $25
• Income: $1000
• Utility: 𝑈 ℎ, 𝑟 = 2ln(ℎ) + 5ln(𝑟)
2 5
• Marginal Utility: 𝑀𝑈ℎ = , 𝑀𝑈𝑟 =
ℎ 𝑟

What is the new optimal split between eating at home and dining out?

7
Example: optimal bundle
1. Determine how much you can eat at home, given your dining out:

10h+25r = 1000
h = 100 – 2.5r

2. Solve for the optimum by equalizing “bang for the buck”:

5ൗ
𝑀𝑈𝑟ൗ 𝑟
𝑝𝑟 =
25
2
2ൗ
𝑀𝑈ℎൗ
= ℎ = 100 − 2.5𝑟
𝑝ℎ 10 10

8
Example: optimal bundle
2. Solve for the optimum by equalizing “bang for the buck”:
2
100 − 2.5𝑟 5ൗ
𝑴𝑼𝒉ൗ 𝑟 𝑴𝑼𝒓ൗ
𝒑𝒉 = 10
=
25
= 𝒑𝒓
50 50
=
100 − 2.5𝑟 𝑟
50𝑟 = 5000 − 125𝑟

∗
5000
𝑟 = ≈ 28.6
175
ℎ = 100 − 2.5𝑟 = 100 − 2.5 5000ൗ175 ≈ 28.6 = ℎ∗

9
What goods don’t we purchase?
• For any two goods purchased, at the optimum, it must be that
𝑀𝑈𝑥 𝑀𝑈𝑦
=
𝑝𝑥 𝑝𝑦

Optimality condition holds only when we purchase both goods.

• Suppose that you spend all of our income on eating out but…
𝑀𝑈ℎ 𝑀𝑈𝑟
<
𝑝ℎ 𝑝𝑟
• In this case, it would be optimal to eat zero meals at home.
• What does this tell us about increasing demand for our product?

10
Activity: New Wardrobe
We want to understand how our preferences and budget affect our optimal
bundle and the value we receive from a good.

1. Suppose that you get 1000 units of utility from your first pair of pants.
– How much utility do you get from having two pairs? 3, 4 pairs?
– How much utility do you get from having one shirt? 2, 3, 4 shirts?
– How much utility do you get from having one pair of shoes? 2, 3, 4 pairs?

2. Suppose that pants, shirts, and shoes all cost $50.

– If you have $100 dollars to spend, what do you buy?
– What if you have $350?

11
Where do we get market demand?
• For each individual, we can plot
Price

demand for a good by seeing how

Cami’s Demand the optimal quantity changes as
the price changes.
p1 Market Demand • To find the market demand, we
just add the curves together
My Demand horizontally.
p2 • At p1, the market demands q1
– Cami demands zero at that price
• At p2, market demand is qc,2 + qj,2
– Both of us demand q > 0
q1 qc,2 qj,2 q2 Quantity

12
 Consumer surplus
Price

• You are considering how many

200 new shirts to buy for recruiting.
Your Demand
150 – If a shirt costs $200, you buy 1.
– If a shirt costs $150, you buy 2.
100 – If a shirt costs $100, you buy 3.
– If a shirt costs $50, you buy 4.
50

1 2 3 4 Shirts

13
 Consumer surplus
Price

• Suppose that the price of a new

shirt is $125. This price is
200 Your Demand – $75 less than what you were
willing to pay for the first shirt
150 – $25 less than what you were
willing to pay for the second.
100
• We say that you have earned a
50 surplus of $25 + $75 = $100.

1 2 3 4 Shirts

14
 Consumer surplus
• Consumer surplus measures how
Price

well-off consumers are given prices.

Consumer Surplus • Formally, consumer surplus is the

difference between the maximum
200 amount a consumer is willing to pay
Your Demand and the actual amount paid.
150 – Absent wealth effects, willingness to
p pay can be approximated by the
100 demand curve.

50 • Implication? Consumer surplus is

the area under the demand curve
1 2 3 4 Shirts and above the price.

15
Consumer surplus and elasticity

• Consumer surplus is higher when

the demand is relatively inelastic.
“Inelastic” demand
Price

• Example: demand for pain

medicine is more inelastic than
demand for a specific brand, e.g.,
“Elastic” demand Motrin IB.
p – The inability to buy Motrin is
“inconvenient”
– The inability to buy pain
medicine would be “tragic”
q Quantity

16
How do we account for uncertainty?
Suppose you could choose between two investments:

1. Pays $2 million (with probability ½); otherwise, worth $0

2. Pays $4 million (with probability ¼); otherwise, worth $0

Which do you prefer?

17
How do we account for uncertainty?
Both investments have the same expected payoff:
1. 𝑥ҧ = 50%*($2 million) + 50%*(0) = $1 million
2. 𝑥ҧ = 25%*($4 million) + 75%*(0) = $1 million
But they differ in their riskiness, captured here by the variance:
1. 𝜎 2 = 50%*(2 – 1)2 + 50%*(0 – 1)2 = 1
2. 𝜎 2 = 25%*(4 – 1)2 + 75%*(0 – 1)2 = 3

While there is no “right” answer, most people prefer investment #1 because

it’s less uncertain, i.e., less “risky”.

This is representative of risk-aversion.

18
 Expected utility
Utility

Risk-aversion is equivalent to
consumer’s possessing a declining
u($4m) marginal utility of wealth.

u($2m) • Utility is always increasing in

wealth…
• … but under the assumption that
marginal utility is declining, the
slope is decreasing
• Note that:
u($2m) – u($0) > u($4m) – u($2m)
$2m $4m Wealth

19
 Risk Tolerance
Utility

Risk-aversion is equivalent to a
consumer possessing a declining
marginal utility of wealth.
u($2m) – u($0) > u($4m) – u($2m)
u($4m)
Risk-seeking is equivalent to a
u($2m) consumer possessing an increasing
marginal utility of wealth.
u($4m)
u($2m) – u($0) < u($4m) – u($2m)
u($2m)
$2m $4m Wealth

20
Expected Utility
With risk-aversion, the expected payoff ≠ expected utility:
– Expected payoff = σ𝑖 𝑝𝑖 𝑥𝑖 (p: probability, x: payoff)
– Expected utility = σ𝑖 𝑝𝑖 𝑢 𝑥𝑖 (p: probability, u(x): utility from payoff)

Consumers makes choices to maximize expected utility

– As uncertainty increases, the expected utility decreases.
– As risk-aversion increases, the expected utility decreases.
– As the payoff decreases, the expected utility decreases.

Risk-aversion implies that the optimal course of action (good

purchased or investment) accounts for both payoffs and uncertainty.

21
Expected Utility
Suppose you flip a coin: heads, you win $100,000, tails, you win $0.
With risk-aversion, expected payoff > price you’d be willing to pay:
– Expected payoff = σ𝑖 𝑝𝑖 𝑥𝑖 (p: probability, x: payoff)

• 50%*100,000 + 50%*$0 = $50,000

– Expected utility = σ𝑖 𝑝𝑖 𝑢 𝑥𝑖 (p: probability, u(x): utility from payoff)

• 50%*u(100,000)+50%*u($0) < u($50,000)

• This implies you would not give up $50,000 for this bet!

22
Risk Premium
We can measure the impact of uncertainty using the risk premium.

Expected Utility = u(Expected Payoff – Risk Premium)

σ𝑖 𝑝𝑖 𝑢 𝑥𝑖 =𝑢 σ𝑖 𝑝𝑖 𝑥𝑖 − 𝑅𝑖𝑠𝑘 𝑃𝑟𝑒𝑚𝑖𝑢𝑚

• Essentially, the risk premium measures how much of the expected payoff you
would be willing to give up to make the investment risk-free.
– Example: If you are indifferent between an investment that produces $10k risk-free and
a risky investment with expected payoff of $15k, the risk premium is $5k.
• Alternatively, the risk premium is the required compensation (reduction in price)
for holding a risky asset.
– Example: You would never be willing to pay more than the present value of $10k (risk-
free) for the risky investment with expected payoff of $15k described above.

23
Risk Premium
We can measure the impact of uncertainty on your willingness to pay
using the risk premium.

Risk Premium You Demand = Expected Payoff – Price You’d Pay

In reality, however

Risk Premium You Earn = Expected Payoff – Price You Pay

(1) What does this imply about risky asset holdings across individuals?
(2) What does this imply about expected returns on investments?

24
Risk-aversion: Implications

How do consumers reduce the impact of uncertainty?

– Diversification: By allocating funds across multiple investments whose
outcomes are not perfectly correlated, some risk can be eliminated.
– Insurance: When we buy insurance, we are making an explicit payment
to eliminate risk.

How can a goods producer (i.e., a firm) reduce the impact on sales?
– Warranties or product guarantees
– Return Policies
– Reputation

25
Economic costs
• Proper economic decision-making requires knowledge of all
relevant economic costs.

• Two main types of economic costs:

– Explicit costs are payments for factors of production, e.g., wages
paid or cost of materials
– Opportunity costs represent foregone earnings/income, e.g.,
value of your labor or price at which you could sell inputs
• This is the value of the next best alternative.

26
Ignore sunk costs!
• We typically have multiple opportunities to decide whether to
push forward with a decision
– As a result, when we consider whether we want to continue an
action, we may have already expended resources

• A sunk cost is an expenditure that cannot be recovered.

• Example: You constructed a new retail store. What’s sunk?
– Your sunk costs can vary over time

Sunk costs should not influence your decisions.

27
Accounting costs are not the same
• Accounting costs are backward-looking and focus exclusively
on explicit costs.
– As a result, while they will incorporate …
• Wages paid to employees
• Rent on an office space
• Price paid to acquire raw materials
– … they do not include
• Entrepreneur’s foregone wages
• Price you would receive from subletting the office space
• Current market price of raw materials

28
What is the optimal production plan?
• For any two inputs utilized, at the optimum, it must be that:
𝑀𝑃𝐿 𝑀𝑃𝐾
=
𝑤 𝑟
– At the optimum, marginal productivity per dollar spent is the same

• All economic costs (explicit + opportunity) are captured by “w” and “r”.

• Optimality condition holds only when we utilize both inputs. If,

𝑀𝑃𝐿 𝑀𝑃𝐾
<
𝑤 𝑟
then no labor is utilized in production.

29
How do prices impact input choice?
It will generally be the case that as the price of an input
increases, we will use less of it (and we will not use more).
This is just the law of demand!
– If the cost of capital goes up, firms demand less of it (𝜕𝐾
𝜕𝑟
≤ 0)
– If the cost of labor goes up, firms demand less of it (𝜕𝑤
𝜕𝐿
≤ 0)

How do price changes impact our utilization of other inputs?

– Substitutes: If demand for input one increases when the price of input two
𝜕𝐾
increases, e.g., if 𝜕𝑤 > 0 or if 𝜕𝐿
𝜕𝑟
>0
– Complements: If demand for input one decreases when the price of input two
𝜕𝐾
increases, e.g., if 𝜕𝑤 < 0 or if 𝜕𝐿
𝜕𝑟
<0

30
Activity: Building a Cost Curve
We want to understand how to create a cost curve.
You can use either capital (K) or labor (L) to produce your good.
– Output: f(K,L) = 10K + L implies MPK = 10 & MPL = 1
– Costs: r = $8,000 (cost of capital), w = $500 (cost of labor)

1. Do you use capital or labor to produce the good?

2. What is the total cost to produce 1,000 units?
3. What is the total cost to produce q units?
4. At what wage would you utilize either capital or labor (i.e. you are
indifferent)?

31
Total Costs
A firm can also decompose its costs along production levels.
1. Fixed costs: invariant to the quantity produced
• Examples: Rent, legal expenditures, advertising
• Can be sunk or non-sunk
– If they are sunk, they are incurred even if nothing is produced
– If they are non-sunk, they are incurred only with production
– In this class fixed costs are non-sunk
• In the short-run, certain capital costs may be fixed
2. Variable costs: will change with the quantity produced
• Examples: Materials, labor, shipping
• In the long-run, both labor and capital are variable

32
Total Cost Curve
Cost

• Total Costs = Fixed Costs +

Variable Costs
• What determines the total cost
Total cost required to produce each level
of output?
– The optimal production plan!
Variable
– The total cost is the minimum
cost cost required to produce a
given number of units
Fixed cost
Output

33
 Total Cost Curve
• Average cost is TC(q)/q
Cost

– Per-unit cost of current

production
Total cost – Slope of the line from the
origin to the point (q,C(q))
• Marginal cost is ∂TC(q)/∂q
– Per-unit cost of additional
TC(q) production
– Slope of the total cost curve at
quantity q
– Implies that MC = ∂VC(q)/∂q
• Fixed cost does not vary with
the level of output
q Output

34
 Connecting marginal costs & productivity
• Marginal cost falls when the
Cost/unit

next unit costs less to produce

than the prior
– If productivity is increasing, then
fewer inputs are used to produce
the next unit

• Marginal cost rises when the

Marginal Cost next unit costs more to
produce than the prior
– If productivity is decreasing, then
more inputs are used to produce
the next unit
q* Output

35
 Connecting average & marginal costs

Three scenarios to consider:

Cost/unit

1. If MC(q) < AC(q), then average

Average cost cost is falling
2. If MC(q) > AC(q), then average
cost is rising
3. If MC(q) = AC(q), then average
Marginal Cost cost is unchanged

In this example, average cost is

minimized at q*.
q* Output

36
 Connecting average & fixed costs
• Marginal cost could also always
be increasing
Cost/unit

– If more inputs are always

Average cost required on the margin, then
marginal cost will be upward-
sloping
• In spite of this, average cost
may have a decreasing region
Marginal Cost – As more units are produced,
the average fixed cost falls
– This can lower the average
total cost, even as marginal
q* Output cost increases

37
What is the short-run? The long-run?
• Long-run and short-run is meant to distinguish between firm’s
level of flexibility and do not correspond to a set length of time
• In the long-run, we assume firms can choose each input
optimally: there are no constraints to how it is produced
• In the short-run, we assume that certain inputs are fixed in the
production process: given these constraints, the firm chooses
optimally how to produce the good.
– We will generally assume that in the short-run, capital is invariant.
– Of course, other factors may also be pre-determined, including labor.

38
 Short-run v. long-run total cost
Cost

• In the long-run, the firm can

choose each input optimally
Short-run – The total cost must be the
total cost minimum possible cost for each
quantity
• In the short-run, can only
TC(q*) choose some inputs optimally
Long-run – The firm is constrained (some
total cost inputs are set): short-run total
cost must exceed the long-run
– Note that there is generally one
exception: the point at which the
fixed level of inputs is optimal (q*)
q* Output

39
Short-run v. long-run average cost
Short-run
Cost\unit

• In the short-run, can only choose

average cost some inputs optimally
• Suppose there are only three
possible choices for the quantity
L of capital utilized (which
R correspond to the three curves).
• In the long-run, the firm chooses:
1. Curve L if q < q1
2. Curve R if q > q2
M 3. Curve M if q1 < q < q2
• The long-run average cost curve is
just the optimally-chosen short-
run average cost curve
q1 q2 Output

40
Activity: Where would you locate?
We want to understand how fixed and marginal costs affect production in
both the short-run and the long-run.

You are choosing between three locations to manufacture your product.

– Location #1: 10,000 sq. ft, marginal cost/unit is $1.
– Location #2: 50,000 sq. ft, marginal cost/unit is $0.50.
– Location #3: 100,000 sq. ft, marginal cost/unit is $0.25.
Suppose your lease costs $1/sq. ft each month. Objective: minimize costs.
1. Which do you choose if you produce 50,000 units/month? 100,000/month?
2. For what quantities would you choose Location #1? Location #3?
3. (Challenge): If there is a 50% chance that you produce 50,000/month and 50%
chance you produce 100,000/month, which do you choose?

41
Microeconomics in the News
• “Investors Should Go Where the Float Is”, 8/19/22

• “Jobs for the City of Tomorrow”, 6/7/21

• “Some People Can’t Wait to Dress for Work Again”, 4/22/21

• “Peloton Rides Covid-19 Wave, Adding Products, Cutting Bike Price”, 9/8/20

• “Demand for Houses Boosts Home Construction”, 9/1/20

• “Junk Bond Investors Turn Picky”, 8/20/19

Berman and Johnson (2016), “The Unintended Consequences of Changes in

Beverage Options and the Removal of Bottled Water on a University Campus”,
American Journal of Public Health

42