“Market Structure & Firms’ Behavior”

Part 1
Ajaz Hussain, Economics, University of Toronto (STG)
(ECO101) “Theory”: Market Structure by # of Firms

Monopoly Oligopoly Perfect Competition

One firm has “full” market power Each of the “few” firms has “some” market power No firm has market power

Example: (Perfect) “Competition”

“The market for crude oil is global, and Texas as a whole (let alone a single firm)
constitutes only 1.3 percent of world oil production; thus, the exercise of market power by
Texas oil producers is implausible (Texas and world production data are for 2007 and
were accessed from the US Energy Information Administration [from here] , respectively,
on September 27, 2015)”

Source: Soren T. Anderson, Ryan Kellogg, and Stephen W. Salant, "Hotelling under
Pressure," Journal of Political Economy 126, no. 3 (June 2018): 984-1026.

2
“Practice”: Two Metrics of Market Structure
Define “Market” Boundary and let 𝑇 = Total # of firms in market
“Tricky” – see next slide
Hardest

𝐹𝑖𝑟𝑚 𝑖 𝑟𝑒𝑣𝑒𝑛𝑢𝑒𝑠
Calculate each firm’s % share of market revenues (𝑠𝑖 = 100)
𝑀𝑎𝑟𝑘𝑒𝑡 𝑟𝑒𝑣𝑒𝑛𝑢𝑒
Is

É
Concentration Ratio Herfindahl-Hirschmann Index
𝐶𝑅𝑛 𝐻𝐻𝐼

outstop n'companies s lowest

W
Calculate combined market share of the Calculate total squared market shares of all 𝑇 firms:
𝐶𝑅𝑛 range? 𝐻𝐻𝐼 = σ𝑇𝑖=1 𝑠𝑖2
𝐻𝐻𝐼 range?
top 𝒏 firms (by market share): 𝐶𝑅𝑛 = σ𝑛𝑖=1 𝑠𝑖

not
1,500 < 𝐻𝐻𝐼 < 2,500
𝐶𝑅1 ≥ manage 𝐶𝑅4 < 40 𝐻𝐻𝐼 ≤ 1,500
If the III.ly
90

on
“Monopoly” “Competition” “Not Concentrated”
“Moderately
Concentrated” raisins

at top 40 ≤ 𝐶𝑅4 ≤ 60 𝐶𝑅4 > 60 at top 𝐻𝐻𝐼 ≥ 2,500 US data

4 comps “Loose oligopoly” “Tight oligopoly” “Concentrated” Canadian data
monopoly Balancing Everything -- Stats
3
Summary: Market Structure According to 𝐶𝑅𝑛 and 𝐻𝐻𝐼
AFORMULA SHEET
Loose Oligopoly
40 ≤ 𝐶𝑅4 ≤ 60
Competitive Monopoly
𝐶𝑅4 < 40 𝐶𝑅1 ≥ 90
Tight Oligopoly
𝐶𝑅4 > 60

“Un-concentrated” “Moderately Concentrated” “Concentrated”

𝐻𝐻𝐼 ≤ 1,500 1,500 < 𝐻𝐻𝐼 < 2,500 𝐻𝐻𝐼 ≥ 2,500

Discussed Week 23 and Week 24 Discussed Week 21 and Week 22

Discussed Week 20 Lecture
Lectures Lectures

4
Defining Market Boundary is Easier Said Than Done CRN ES
HH E's
Example #1: Proposed Sirius and XM Satellite Radio “Merger” (2008)
euger
Satellite Radio Market “All” Radio Market
FewerFirm
lessent competition

pi
efficiency

Example #2: Proposed AA and BA Merger on LHR-JFK Route (2010)

Bermuda II Treaty

JFK Gov onlyTAKESconsOFONERoute LHR

botEachairport wasbe labovers
1
Meansbigger
Market
5
HHI: Examples #1

Source: Beware Energy's Junk Debt Army, Bloomberg, 12/14/15

The chart .. plots each sector's HHI score and its weighted average yield. The bubble
size shows how much debt is outstanding.

You'll notice a couple of sectors stand out somewhat. In theory, diversity offers investors
some protection. That may hold true to some degree in the basic industries sector, which
encompasses everything from construction to chemicals alongside metals and mining
companies. Yet metals feature prominently in that sector, with the top five issuers by
amount being ArcelorMittal, Alcoa, Peabody Energy, Teck Resources and Fortescue
Metals Group. All are exposed to the bear market in commodities tied largely to fears of
a slowdown in Chinese demand. Similarly, energy issuers may be legion, but they are all
Source: “Resource Monopolies” tied to the oil price, which sank to a new multi-year low on Monday.
6
HHI: Examples #2

Source: Here’s How They Play Monopoly in America, and Who Wins, Bloomberg, April 5, 2017 Source: Economist Magazine, Corporate Concentration

The Dirty Little Secret of Credit Card Rewards Programs

Source: New York Times, March 4, 2023

Distributional Effects of Payment Card Pricing and Merchant

Cost Pass-through in the United States and Canada
Source: Bank of Canada, March 2020

Source: CBC March 9, 2023

7
HHI: Example #3 (Market Structure and Finance)

“ .. Julio Rotemberg, then a newly minted economist from Princeton,

posited that “firms, acting in the interest of their shareholders,” might
“tend to act collusively when their shareholders have diversified
portfolios.

The idea, which Rotemberg explored in a working paper, was that if

investors own a slice of every firm, they will make more money if firms
compete less and collectively raise prices, at the expense of consumers.
Knowing this, the firms’ managers will de-emphasize competition and
behave more cooperatively with one another.”

Source: Are Index Funds Evil? A growing chorus of experts argue that
they’re strangling the economy—and must be stopped
The Atlantic, Sept. 2017
8
 rewind IUD

“Market Structure & Firms’ Behavior”

Part 2
Ajaz Hussain, Economics, University of Toronto (STG)
Examples: Horizontal Mergers
Sunny Days Ahead – The WestJet-
Sunwing Merger is Cleared for Takeoff
Source: McMillan March 10, 2023

Deadline for Rogers,

Shaw merger
extended to March 31
Companies still need
approval from
Canadian industry
minister Francois-
Philippe Champagne
Source: National Post,
Feb 17, 2023

Source: Rogers’s Shaw Bid Has ‘Serious’ Competition

Issues, Canada Says, Bloomberg, March 30, 2021
10
 Final Exam
(Deloitte) The state of the deal: 2022 M&A trends

11
(US) Horizontal M&A Guidelines
𝐻𝐻𝐼
He
𝐻𝐻𝐼 ≤ 1,500 1,500 < 𝐻𝐻𝐼 < 2,500 𝐻𝐻𝐼 ≥ 2,500
“Unconcentrated” “Moderately Concentrated” “Concentrated”

Small Change in Concentration: Mergers involving an increase in the HHI of less than 100 points are unlikely to have
adverse competitive effects and ordinarily require no further analysis.

Unconcentrated Markets: Mergers resulting in unconcentrated markets are unlikely to have adverse competitive
effects and ordinarily require no further analysis.

Moderately Concentrated Markets: Mergers resulting in moderately concentrated markets that involve an increase in
the HHI of more than 100 points potentially raise significant competitive concerns and often warrant scrutiny.

Highly Concentrated Markets: Mergers resulting in highly concentrated markets that involve an increase in the HHI of
between 100 points and 200 points potentially raise significant competitive concerns and often warrant scrutiny.
Mergers resulting in highly concentrated markets that involve an increase in the HHI of more than 200 points will be
presumed to be likely to enhance market power. The presumption may be rebutted by persuasive evidence showing
that the merger is unlikely to enhance market power.
12
Calculating Change in 𝐻𝐻𝐼 from a Proposed Merger
𝐻𝐻𝐼𝑝𝑟𝑒−𝑚𝑒𝑟𝑔𝑒𝑟 = 𝑠12 + 𝑠22 + … + 𝑠𝑖2 + … + 𝑠𝑗2 + … + 𝑠𝑇2
2quagggreabive

Suppose companies 𝑖 and 𝑗 propose a “merger” deal:

2
𝐻𝐻𝐼𝑝𝑟𝑜𝑝𝑜𝑠𝑒𝑑−𝑚𝑒𝑟𝑔𝑒𝑟 = 𝑠12 + 𝑠22 + … + 𝑠𝑖 + 𝑠𝑗 + … + 𝑠𝑇2

If the proposed merger is “approved” and companies 𝑖 and 𝑗 become one company:

Δ 𝐻𝐻𝐼 =
N a sits sits
p
z sis
If g
attJan hove

13
“Firm with 𝐶𝑅1 ≥ 90: “Pricing Mechanisms”
Part 1
Ajaz Hussain, Economics, University of Toronto (STG)
Companies with Market Power: Some Applications
Pricing
“Organization” (In ECO204 2022-2023, we
focus on some aspects of
pricing)

Firm with Market Power

𝐶𝑅1 ≥ 90
𝐻𝐻𝐼 > 2,500

“Competitive Strategy”
Marketing, Branding,
Exogenous and Endogenous
Advertising
Barriers to Entry

15
“Price Neglect” p TUC TFC Pla E TUCCE TFC
McKinsey Consulting: Price changes have the biggest, fastest impact on profits
+1% price = profits up 8.6% LARGEST IMPACT
-1% variable cost = profits up 5.9%
+1% volume = profits up 2.8%
-1% fixed cost = profits up 1.7%
Survey of professional pricing society members:
40% match online prices to those offline
30% set all prices by matching rivals

“Amazon typically relies on algorithms that scrape competitors’ prices before automatically matching or
narrowly undercutting them on its website”
At Whole Foods, Amazon Takes Rare Lead in Cutting Prices, Wall Street Journal, August 28, 2017
28% have no Internet pricing strategy
22% set all prices only to recover costs and tack on profit (Cost-Plus) Not OPTIMAL
18% do customer research to determine value to buyer
16
Firm with Market Power: Some Static Pricing Mechanisms
Pricing Scheme in Each Period Independent of Other Periods (i.e. no inter-temporal demand/production linkages)

Period 1 Period 2 Period 3 Period 4 …

Ra3 Qo
Firm with Market Power: Pricing Mechanisms

Fixed access tee

Uniform Market Price “Price Discrimination”

P'g
price
same p per out var waseca

Two-Part Tariff
Customized Pricing
(2nd Degree Price Discrimination with an access fee
(1 Degree Price Discrimination)
st
and a usage price)

Product-Line Pricing (aka Segment Pricing) Mixed Bundling

(3rd Degree Price Discrimination with a uniform price (Offer customers choice of buying products a la carte
in each segment) with and without risk of arbitrage or as a bundle [at uniform prices])”

17
Data for Pricing Decisions
Basic Data Relevance Comment
Price/demand data B2C Helps construct price-response function
Bid-price data B2B Helps construct bid-response function
Pricing guidance and business rules B2C, B2B Specifies constraints in pricing decision
Market segmentation and intelligence B2C, B2B Supports price customization (price discrimination)
Advanced Data Relevance Comment
Product and location hierarchies B2C Supports tier pricing
Special event data B2C Supports timing of special events
Inventory data B2C Helps estimate effect of inventory on sales
Out-of-stock data B2C Helps “un-constrain” demand
Promotion data B2C Supports timing of special promotion
Competitive data B2C, B2B Links own performance with rivals’ prices
Click data B2C Supports segmentation and price customization

Source: Bodea and Ferguson, Pricing Segmentation and Analytics, Harvard Business Publishing
18
“Firm with 𝐶𝑅1 ≥ 90: “Pricing Mechanisms”
Part 2
Ajaz Hussain, Economics, University of Toronto (STG)
Firm with Market Power: Some Static Pricing Mechanisms
Pricing Scheme in Each Period Independent of Other Periods (i.e. no inter-temporal demand/production linkages)

Period 1 Period 2 Period 3 Period 4 …

Firm with Market Power: Pricing Mechanisms

Uniform Market Price “Price Discrimination”

Two-Part Tariff
Customized Pricing
(2nd Degree Price Discrimination with an access fee
(1 Degree Price Discrimination)
st
and a usage price)

Product-Line Pricing (aka Segment Pricing) Mixed Bundling

(3rd Degree Price Discrimination with a uniform price (Offer customers choice of buying products a la carte
in each segment) with and without risk of arbitrage or as a bundle [at uniform prices])”

20
𝐶𝑅1 ≥ 90 Firm Charging a Uniform Price

𝐶𝑅1 ≥ 90 company sells a good/service at a uniform price

Assume 100% demand/cost certainty

p in
dieters Cesbulkus sins
que
“Profit Maximizer” “Revenue Maximizer”

IIIT pucca TFC

to
Need 𝑃 𝑞 : “Price Need Cost function: Need 𝑃 𝑞 : “Price
Response Function” (i.e. Response Function” (i.e.
𝐶(𝑞) = 𝑇𝐹𝐶 + 𝑇𝑉𝐶(𝑞)
demand function/curve) demand function/curve)

How do we obtain the price response and cost functions?

21
Whither and Which Price-Response Function?
From consumer theory, we know there are an infinite # of demand functions (𝑞 = 𝑓(𝑃))
and thus there are an infinite # of demand curves (price-response functions) (𝑃 = 𝑓 𝑞 )

Linear Concave Convex Semi-log

Infinite # of
Price-Response Price-Response Price-Response Price-Response other Price-
Function Function Function Function Response
Function
𝑃 = 𝑎 − 𝑏𝑞 𝑃 = 𝑎 − 𝑏෠ 𝑞 ሶ 2
𝑃 = 𝑎 − 𝑏𝑞 𝑃 = 𝑎𝑞 −𝛼

෠ 𝑏,ሶ 𝛼 > 0 parameters

𝑎, 𝑏, 𝑏,

Exercise: What is the “market size” for linear, concave, convex demand curves?
22
 Week 21: Demand Models Excel File

Please review Constrained Optimization Lectures

23
Some Price-Response and Respective Revenue Functions

CONVEX

Linear
concave

SEMI LOCt

q
Linear Demand Curve 𝑃 = 𝑎 − 𝑏𝑞 = 100 − 𝑞
Concave Demand Curve 𝑃 = 𝑎 − 𝑏෠ 𝑞 = 100 − 10 𝑞
4 Classic

MEI
mon Convex Demand Curve ሶ 2 = 100 − 0.01𝑞 2
𝑃 = 𝑎 − 𝑏𝑞
Into
thes
Semi-Log Demand Curve 𝑃 = 𝑎𝑞 −𝛼 = 100𝑞 −0.1

24
Examples: “Linear” Demand Models
Source: Gamsi, Laffont, Vuong, “Econometric Analysis of Collusive Behavior in a Soft−Drink Market”,
Journal of Economics and Management Strategy

𝑞𝑐 = 26.17 − 3.98𝑃𝑐 + 2.25𝑃𝑝 + 2.60𝐴𝑐 − 0.62𝐴𝑝 + 9.58𝑆 + 0.99𝑌

O N ers 0 If
𝑞𝑃 = 17.48 + 1.40𝑃𝑐 − 5.48𝑃𝑝 − 4.81𝐴𝑐 + 2.83𝐴𝑝 + 11.64𝑆 + 1.92𝑌

∎ 𝑞 = quarterly quantity of syrup sold ∎ 𝑃 = real price of syrup ∎ 𝑌 = real income

1 if spring/summer
∎ 𝐴 = square root of quarterly real advertising expenses ∎ 𝑆 = ቊ
0 if winter/fall
Source: HBS Case The Prestige Telephone Company

March 2003 Commercial Demand Curve

𝑃𝑐 = 1,466.67 − 4.83𝑞𝑐
25
Examples: Semi-Log Demand Models
Source: Bodea and Ferguson, Pricing Source: Demand Function of Beef per capita (Ajaz’s
Segmentation and Analytics, Harvard calculations using USDA data)
Business Publishing)
𝛽 𝛽
𝑞𝑏𝑒𝑒𝑓 = 𝛽0 × 𝑃𝑐ℎ𝑖𝑥
1
× 𝑃𝑏𝑒𝑒𝑓
2
× 𝑌𝛽3
𝑞 = 𝛽0 𝑃𝛽1
I
ln 𝑞 = ln 𝛽0 + 𝛽1 ln 𝑃
ln 𝑞𝑏𝑒𝑒𝑓 = ln 𝛽0 + 𝛽1 ln 𝑃𝑐ℎ𝑖𝑥 + 𝛽2 ln 𝑃𝑏𝑒𝑒𝑓 + 𝛽3 ln 𝑌
o

Bo
É “Log-Log” Regression:
“Log-Log” Regression:

ln 𝑞𝑏𝑒𝑒𝑓 = 8.86 + 0.7 ln 𝑃𝑐ℎ𝑖𝑥 − 0.83 ln 𝑃𝑏𝑒𝑒𝑓 − 0.33 ln 𝑌

ln 𝑞 = 15.98 − 3.44 ln 𝑃
0.7 − 0.83
𝑞 = 8,710,154 𝑃−3.44 𝑞𝑏𝑒𝑒𝑓 = 𝑒 8.86 × 𝑃𝑐ℎ𝑖𝑥 × 𝑃𝑏𝑒𝑒𝑓 × 𝑌 − 0.33

0.83 0.33
a
Tutorial: 𝑃𝑟𝑖𝑐𝑒 𝐸, 𝐶𝑟𝑜𝑠𝑠 𝑃𝑟𝑖𝑐𝑒 𝐸, 𝐼𝑛𝑐𝑜𝑚𝑒 𝐸?
Tutorial: Notice that 𝑃𝑟𝑖𝑐𝑒 𝐸 = −3.44
In
26
A Beautiful & Practical Result for Profit Maximizing Companies

If the actual demand-curve is not known, then – under some

conditions – the “optimal profit maximizing price of the unknown
demand curve” is best approximated by the “optimal profit
maximizing price of any linear demand curve”

See next slide for “conditions”

Source: A Simple Rule for Pricing with Limited Knowledge of

Demand. MIT Sloan Research Paper No. 5145-15
27
Optimal Price of Linear Demand Curve ≈ Optimal Price of Unknown Demand Curve

A Simple Rule for Pricing with Limited Knowledge of Demand. MIT Sloan Research Paper No. 5145-15
𝑃
Assumption 1: Company knows max 𝑊𝑇𝑃

Then of all the possible approximate demand

curves, the best model is any linear demand curve
𝑃Π∗
Suppose “actual” (unknown) price-response curve
[this particular curve could be for a new drug]

𝑀𝐶 = 𝐴𝑉𝐶

Linear Linear Assumption 2: Company has constant 𝑀𝐶 = 𝐴𝑉𝐶

Demand Demand
𝑀𝑅

𝑞Π∗

Assumption 3: Firm can produce any amount and can dispose of unsold units at zero cost
ANY DEMAND CURVE 28
 p
𝐶𝑅1 ≥ 90 Firm Charging Uniform Prices

𝐶𝑅1 ≥ 90 company sells a good/service at a uniform price

Assume 100% demand/cost certainty

“Profit Maximizer” “Revenue Maximizer”

Need 𝑃 𝑞 : “Price
Assume linear price- Assume linear cost
Response Function” (i.e.
response function function
demand function/curve)

Recall linear cost function:

𝐶 = 𝑇𝐹𝐶 + 𝑇𝑉𝐶 𝑞 = 𝑇𝐹𝐶 + constant 𝑞 where 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 = 𝐴𝑉𝐶 = 𝑀𝐶
29
Example 1: Linear Price-Response and 𝑇𝑉𝐶 𝑞 Functions
Coke & Pepsi
Inputs Bottling & Distribution Retail Channels
Concentrate Syrup

Coke and Pepsi Concentrate Syrup Demand models. All pecuniary variables are in constant dollars

𝑞𝑐 = 26.17 − 3.98𝑃𝑐 + 2.25𝑃𝑝 + 2.60𝐴𝑐 − 0.62𝐴𝑝 + 9.58𝑆 + 0.99𝑌

𝑞𝑃 = 17.48 + 1.40𝑃𝑐 − 5.48𝑃𝑝 − 4.81𝐴𝑐 + 2.83𝐴𝑝 + 11.64𝑆 + 1.92𝑌

∎ 𝑞 = quarterly quantity of syrup sold ∎ 𝑃 = real price of syrup ∎ 𝑌 = real income

1 if spring/summer
∎ 𝐴 = square root of quarterly real advertising expenses ∎ 𝑆 = ቊ
0 if winter/fall

∎ 𝐴𝑉𝐶𝑐 = 𝑀𝐶𝑐 = $5 ∎ 𝐴𝑉𝐶𝑃 = 𝑀𝐶𝑝 = $4

Source: Gamsi, Laffont, Vuong, “Econometric Analysis of Collusive Behavior in a Soft−Drink Market”,
Journal of Economics and Management Strategy
30
Example 2: Linear Price-Response and 𝑇𝑉𝐶 𝑞 Functions
Recall from The Prestige Telephone Company case
Commercial Data Service Demand Curves Jan 2003 – March 2003 and the “Breakeven Month”
𝑃𝑐 ($/hr)

$1,466.67 March Commercial Demand Curve

𝑃𝑐 = 1,466.67 − 4.83𝑞𝑐

000 “Commercial data services”:

$800 J F M ?

𝑀𝐶𝑐 = 𝐴𝑉𝐶𝑐 = 28

123 135 138 Breakeven 𝑞𝑐

𝑞𝑐 ≈ 177 (hrs/mo)
Graph not to scale
31
“Firm with 𝐶𝑅1 ≥ 90: “Pricing Mechanisms”
Part 3
Ajaz Hussain, Economics, University of Toronto (STG)
The “General” Uniform Price PMP
max Π = 𝑅 − 𝑇𝑉𝐶 − 𝑇𝐹𝐶 𝑠. 𝑡. 𝑞𝑚𝑖𝑛 ≤ 𝑞 ≤ 𝑐 = capacity
𝑞 𝑐>0
𝑞𝑚𝑖𝑛 = 0
max Π = 𝑅 − 𝑇𝑉𝐶 − 𝑇𝐹𝐶 𝑠. 𝑡. 𝑞 − 𝑐 ≤ 0 , −𝑞 + 𝑞𝑚𝑖𝑛 ≤ 0
𝑞

max L = 𝑅 − 𝑇𝑉𝐶 − 𝑇𝐹𝐶 − 𝜆1 𝑞 − 𝑐 − 𝜆2 −𝑞 + 𝑞𝑚𝑖𝑛

𝑞,𝜆1 ,𝜆2

𝑅 = 𝑃𝑞 = 𝑃 𝑞 × 𝑞 max L = 𝑅 − 𝑇𝑉𝐶 − 𝑇𝐹𝐶 − 𝜆1 𝑞 − 𝑐 + 𝜆2 𝑞

𝑞,𝜆1 ,𝜆2

𝜕L 𝑑𝑅 𝑑𝑇𝑉𝐶 𝑑𝑇𝐹𝐶
= − − − 𝜆1 + 𝜆2 = 0 With constant returns  FOC:
𝜕𝑞 𝑑𝑞
ด 𝑑𝑞 𝑑𝑞 𝑀𝑅(𝑞) = 𝑀𝐶 + 𝜆1 − 𝜆2
𝑀𝑅(𝑞) 𝑀𝐶(𝑞) 0

𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0
PMP FOC and KT conditions Next, split this into 2 cases
𝑀𝑅(𝑞) = 𝑀𝐶 + 𝜆1 − 𝜆2
𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0 𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0 First, split this into 2 cases
𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0
33
The “General” Uniform Price PMP
PMP FOC and KT conditions
𝑀𝑅(𝑞) = 𝑀𝐶 + 𝜆1 − 𝜆2
𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0
𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0
𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0

Suppose 𝑞 = 0 Suppose 𝜆2 = 0
If so, 𝜆2 ≥ 0 If so, 𝑞 ≥ 0

𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0 𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0

Suppose 𝑞 = 𝑐 Suppose 𝜆1 = 0 Suppose 𝑞 = 𝑐 Suppose 𝜆1 = 0

If so 𝜆1 ≥ 0 If s𝑜, 𝑞 ≤ 𝑐 If so 𝜆1 ≥ 0 If s𝑜, 𝑞 ≤ 𝑐

Case A Case B Case C Case D

𝑞=0 𝑞=0 If so, 𝑞 ≥ 0 If so, 𝑞 ≥ 0
𝑞=𝑐 If so, 𝑞 ≤ 𝑐 𝑞=𝑐 If so, 𝑞 ≤ 𝑐
If so, 𝜆1 ≥ 0 𝜆1 = 0 If so, 𝜆1 ≥ 0 𝜆1 = 0
If so, 𝜆2 ≥ 0 If so, 𝜆2 ≥ 0 𝜆2 = 0 𝜆2 = 0
34
The “General” Uniform Price PMP
PMP FOC and KT conditions
𝑀𝑅(𝑞) = 𝑀𝐶 + 𝜆1 − 𝜆2
𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0
𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0

Case B Case C Case D

𝑞=0 If so, 𝑞 ≥ 0 If so, 𝑞 ≥ 0
If so, then 𝑞 ≤ 𝑐 𝑞=𝑐 If so, 𝑞 ≤ 𝑐
𝜆1 = 0 If so, 𝜆1 ≥ 0 𝜆1 = 0
If so, 𝜆2 ≥ 0 𝜆2 = 0 𝜆2 = 0

𝑀𝑅 0 = 𝑀𝐶 + 0 − 𝜆2 𝑀𝑅 𝑐 = 𝑀𝐶 + 𝜆1 − 0 𝑀𝑅 𝑞 = 𝑀𝐶 + 0 − 0

⇒ 𝜆2 = −𝑀𝑅 0 + 𝑀𝐶 ≥ 0 ⇒ 𝜆1 = 𝑀𝑅 𝑐 − 𝑀𝐶 ≥ 0 ⇒ 𝑀𝑅 𝑞 = 𝑀𝐶

⇒ 𝑀𝑅 0 ≤ 𝑀𝐶 ⇒ 𝑀𝑅 𝑐 ≥ 𝑀𝐶
Make sure you Interpret and truly
Make sure you Interpret and truly Make sure you Interpret and truly understand this statement (this is
understand this statement understand this statement what you saw in ECO 101)

35
The “General” Uniform Price PMP: Lagrange Multipliers
PMP FOC and KT conditions max Π = 𝑅 − 𝑇𝑉𝐶 − 𝑇𝐹𝐶 𝑠. 𝑡. 𝑞𝑚𝑖𝑛 ≤ 𝑞 ≤ 𝑐
𝑀𝑅(𝑞) = 𝑀𝐶 + 𝜆1 − 𝜆2 𝑞
𝜆1 ≥ 0, 𝑞 ≤ 𝑐, 𝜆1 𝑞 − 𝑐 = 0
𝜆2 ≥ 0, 𝑞 ≥ 0, 𝜆2 𝑞 = 0 L = Π − 𝜆1 𝑞 − 𝑐 − 𝜆2 −𝑞 + 𝑞𝑚𝑖𝑛

At the optimal solution (why?):

∗ ∗
L ∗ = Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 − 𝜆1∗ 𝑞∗ − 𝑐 − 𝜆∗2 −𝑞∗ + 𝑞𝑚𝑖𝑛 = Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑
0 0

∗ ∗
L = Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑

By the Envelope Theorem (see “explainers”):

∗ What should you do
𝑑L 𝑑L ∗ 𝑑Π 𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 with capacity if 𝜆1∗ >
= 𝜆1 ⇒ = 𝜆1∗ ⇒ = 𝜆1∗
𝑑𝑐 𝑑𝑐 𝑑𝑐 , <, = 0?
∗
𝑑L 𝑑L ∗ ∗
𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 What does it mean for
= −𝜆2 ⇒ = −𝜆2 ⇒ = −𝜆∗2 𝜆∗2 >, <, = 0?
𝑑𝑞𝑚𝑖𝑛 𝑑𝑞𝑚𝑖𝑛 𝑑𝑞𝑚𝑖𝑛
36
The “General” Uniform Price PMP: Summary
Case B: when 𝑀𝑅 0 ≤ 𝑀𝐶 Case C: when 𝑀𝑅 𝑐 ≥ 𝑀𝐶 Case D: 𝑀𝑅 𝒒 = 𝑀𝐶(𝑞)
𝑞=0 𝑞≥0 𝑞≥0
𝑞≤𝑐 𝑞=𝑐 𝑞≤𝑐
∗ ∗ ∗
𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑
𝜆1 = =0 𝜆1 = = 𝑀𝑅(𝑐) − 𝑀𝐶 ≥ 0 𝜆1 = =0
𝑑𝑐 𝑑𝑐 𝑑𝑐
∗ ∗ ∗
𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑 𝑑Π𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑒𝑑
𝜆2 = − = 𝑀𝐶 − 𝑀𝑅(0) ≥ 0 𝜆2 = − =0 𝜆2 = − =0
𝑑𝑞𝑚𝑖𝑛 𝑑𝑞𝑚𝑖𝑛 𝑑𝑞𝑚𝑖𝑛
𝑀𝑅 0 ≤ 𝑀𝐶 𝑀𝑅 𝑐 ≥ 𝑀𝐶 𝑀𝑅 𝑞 = 𝑀𝐶
Produce 0 Produce 100% capacity Produce between 0 and 100% capacity

$ $ $
𝑀𝐶(0) 𝑀𝑅(0)
𝑀𝑅(𝑐) 𝑀𝑅 𝑞 = 𝑀𝐶
𝑀𝐶
B 𝑀𝐶(𝑐)
𝑐 𝑞
𝑀𝐶 𝑀𝐶
Demand Demand
Demand C D
0 0 𝑐 𝑞 0 𝑐 𝑞
𝑀𝑅 𝑀𝑅
𝑀𝑅
Real life example? Should company expand capacity? Should company reduce capacity?

37
Blank:

38
Blank:

39
Blank:

40
Blank:

41
Blank:

42
Blank:

43
Blank:

44