SUSTAINABLE ATTRIBUTE VALUATION AND
MANUFACTURER’S SUSTAINABLE PRODUCT LAUNCH DECISION MAKING:
AN EGG INDUSTRY CASE
By
Thasanee Satimanon

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Agricultural, Food and Resource Economics - Doctor of Philosophy
2013

ABSTRACT
SUSTAINABLE ATTRIBUTE VALUATION AND
MANUFACTURER’S SUSTAINABLE PRODUCT LAUNCH DECISION MAKING:
AN EGG INDUSTRY CASE
By
Thasanee Satimanon
Sustainability is one of the most important issues facing all agents along the global food
supply chain. Consumers have an increased awareness about environmental problems, animal
welfare, and communities, which has led to greater demand for sustainable food and drink
products. This demand induces retailers and manufacturers to produce and market more
sustainable products. Some manufacturers are reluctant to implement sustainable practices, or
develop and market new sustainable food products due to a lack of information about price
premiums of sustainable attributes and a systematic decision model.
This dissertation aims to: analyze price premiums of sustainable attributes; construct
firms’ decision making model; determine under which conditions firms launch sustainable
products; and determine which conditions encourage a firm to be a leader who launches a
sustainable product before others. Hedonic analysis is used to find the price premiums of
sustainable products while game theory, comparative statics, and simulation are utilized to
construct the firms’ decision making regarding under which conditions firms launch sustainable
products, and which conditions encourage a firm to be a leader.
The U.S. egg industry was selected to estimate the price premiums of sustainable
attributes, and simulate and test the firms’ decision making model. According to the study the
welfare-managed attribute (free-range or cage-free attribute) is significantly positive. That is, a

welfare-managed egg has a price premium over a conventional egg equal to 3.57 cents while the
recyclable attribute (paper-pulp packaging attribute) is insignificant.
The follower’s decision making to launch a sustainable product is the main contribution
of the game theory study. That is, the follower decides to produce and market a “me-too”
sustainable product when consumers realize the difference between the conventional and
sustainable product and are willing to pay more for the sustainable product. If not, the costs of
both types have to be sufficiently close to induce the follower to launch the sustainable product.
The leader decides to launch a sustainable product when the relative maximum willingness to
pay for the sustainable product over the conventional product is greater than or equal to its
relative cost or when the maximum willingness to pay for the sustainable product is sufficiently
higher than the cost of the sustainable product.
The main factors that encourage a firm to be a leader are: 1) a cost advantage in
producing a sustainable product, 2) a cost disadvantage in producing a conventional product, and
3) a firms relative cost being lower than the consumers’ willingness to pay for a sustainable
product. The U.S. egg industry simulation results show that both leader and follower firms
should market the sustainable eggs to achieve higher profits. The simulation results agree with
what is observed in terms of the growth in the free-range and cage-free eggs market in the U.S.

For my parents, Poonsuk and Pojjanee Tonganupong, and my husband

iv

ACKNOWLEDGEMENTS

I would like to express my deepest gratitude to Dr. Dave Weatherspoon, my major
professor, for his valuable motivations, suggestions, guidance, and kindness throughout every
stage of completing this dissertation. My work would not have been possible without his strong
support and encouragement. I am deeply indebted to Dr. Thomas Jeitschko, Dr. John Hoehn, Dr.
Christopher Peterson, and Dr. Robert Shupp, members of my dissertation committee, for their
helpful suggestions and constructive comments that were essential in improving my dissertation.
I also acknowledge the Department of Agricultural, Food, and Resource Economics
(AFRE) for financial support during my academic years. I am also indebted to the Graduate
School, the College of Agriculture and Natural Resources, Office for International Students and
Scholars, Community Volunteers for International Programs, and the Henry and Annabel
Larzelere Fellowship, which extended its financial support for my dissertation. I would like to
thank the AFRE faculty and staff for their helpfulness, especially Dr. James Oehmke, Dr.
Thomas Readon, Dr. Eric Crawford, Dr. Scott Loveridge, Dr. Scott Swinton, Dr. Steven Miller,
Dr. Suzanne Thornsbury, Dr. David Schweikhardt, Dr. Robert Myers, Ms. Debbie Conway, Ms.
Ann Robinson, and Ms. Janet Munn. My cordial thanks also go to my AFRE friends for their
warm friendship and cheerful encouragement, especially Sonja Perakis, Athur Mabiso, Lara De
Villa, and Benchamaphorn Sombatthira.
Most importantly, my deepest appreciation goes to my family. I would never be who I am
today without the unconditional love, invaluable encouragement, and constant support from my
parents. Also, I would like to thank my parent-in-law for their love and encouragement. Last but
not least, I would like to say thank you to my husband for his love, supports, suggestions, and

v

encouragement throughout every steps to fulfill my degree. He inspired me to pursue a Ph.D. and
has never left me alone during the difficult times. Without him, this dissertation as well as my
life could not be completed. Thank you so much, P’Big. You complete me.
Nevertheless all deficiencies and ambiguities in this dissertation are solely my
responsibility.

vi

TABLE OF CONTENTS

LIST OF TABLES

ix

LIST OF FIGURES

x

CHAPTER 1: INTRODUCTION
REFERENCES

1
4

CHAPTER 2: HEDONIC ANALYSIS OF SUSTAINABLE FOOD PRODUCTS
2.1 Introduction
2.2 Objectives
2.3 Egg Industry
2.4 Methodology
2.5 Data and Variable Description
Credence Goods
Egg Packaging
2.6 Results
2.7 Conclusion and Management Implications
APPENDIX
APPENDIX A
REFERENCES

6
6
9
10
11
14
16
16
17
20
21
22
25

CHAPTER 3: FOOD MANUFACTURERS’ SUSTAINABLE PRODUCT LAUNCH
STRATEGY: GAME THEORY APPROACH
31
3.1 Introduction
31
3.2 Objective
35
3.3 Methodology
35
3.4 Model
38
3.5 Egg Industry Simulations
43
3.6 Results
45
3.6.1 Conditions under which firms launch a sustainable product
46
3.6.1.1 Follower’s decision making for launching a sustainable
product
47
3.6.1.2 Leader’s decision making for launching a sustainable
product
54
3.6.2 Conditions that encourage a firm to be a leader
76
3.6.2.1 Comparison of profits in stage 1
76
3.6.2.2 Comparison of profits in stage 2
78
3.6.2.3 Comparison of profits in stage 3
84
3.7 Conclusion
91
APPENDICES
93
APPENDIX B
94
APPENDIX C
102
vii

APPENDIX D
APPENDIX E
REFERENCES

110
118
122

CHAPTER 4: CONCLUSION
REFERENCES

127
130

viii

LIST OF TABLES

Table 2.1

Definitions of the Variables and their Descriptive Statistics

22

Table 2.2

Results for Hedonic Prices of Egg Attributes

24

Table 3.1

The price premium of free-range/cage-free eggs and the source of
information

43

Table 3.2

The egg industry simulations to test whether proposition 3b is a rare
case or not when varying the degree of substitution ( 1   3 ),
maximum willingness to pay for a sustainable product ( as ), and the
a 
relative maximum willingness to pay for a sustainable product  s 
a 
 c

Table 3.3

The egg industry simulation to test whether proposition 4b is a rare
case or not when varying the value of the relative maximum
a 
willingness to pay for a sustainable product  s  , and the maximum
a 
 c
willingness to pay for a sustainable product ( as )

Table 3.4

65

69

Egg industry simulation of proposition 5b when varying the relative
a 
maximum willingness to pay  s  , the relative degree of substitution
a 
 c

 s 
 2
c
 c  , and the degree of substitution in stage 2 (  2 )
 
 2

ix

82

LIST OF FIGURES

Figure 2.1

Prices of the base level and specialty eggs (cents per egg)

19

Figure 3.1

Stages and types of food product innovation games

39

Figure 3.2

Backward induction decision tree for manufacturer’s decision to launch
a sustainable product

46

Profit simulation between stages 2 and 3 for the follower firm varying
the degrees of substitution (  c   s   3 ) for three levels of relative
2
2
costs of a sustainable product

51

Profit simulation between stages 2 and 3 for the follower firm varying
the degrees of substitution in stage 2 (  c   s ) for four levels of degree
2
2
of substitution (0.2, 0.4, 0.6, and 0.8) in stage 3

52

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7.1

Simulation of the relative maximum willingness to pay for a sustainable
a 
product  s  varying degrees of substitution (  =  c   s   3 )
a 
2
2
 c
Leader’s profits in stages 1 and 3 when the relative cost of a sustainable
product of both firms are the same and equal to the relative maximum
c
c s, j a s 
 and
willingness to pay for a sustainable product  s,i 

 cc,i cc, j ac 


degrees of substitution in stages 1 and 3 are the same ( 1   3 ).
(Condition set (a) holds.)

53

57

3
1
Profit simulation between stages 1 and 3 for the leader (  s,i *  c,i * )
when condition set (a) holds and varying the maximum willingness to
pay for a sustainable product ( as ) for four levels of degrees of
substitution (0.2, 0.4, 0.6, and 0.8) in stage 3 (  3 )

x

60

Figure 3.7.2

3
1
Profit simulation between stages 1 and 3 for the leader (  s,i *  c,i * )
when condition set (b) holds and varying the maximum willingness to
pay for a sustainable product ( as ) for two sets of degrees of
substitution in stages 1 and 3,  1 = 0.9 and  3 = 0.8, and  1 = 0.4
and  3 = 0.3

Figure 3.8

60

Leader’s profits in stages 1 and 3 when a relative cost of a sustainable
product is less than or equal to the relative maximum willingness to
c
a 
pay for a sustainable product  s,i  s  , degrees of substitution
 c c, i a c 


in stages 1 and 3 are the same ( 1   3 ), and firms have the same
costs to produce the same types of products ( cc,i  cc, j and c s,i  c s, j )
62

Figure 3.9

3
1
Profit simulation between stages 1 and 3 for the leader (  s,i *  c,i * )
varying the maximum willingness to pay for a sustainable product ( as )
c
for variety sets of the relative costs of a sustainable product  s,i
 c c, i

and the relative maximum willingness to pay for a sustainable
a 
product  s 
a 
 c

Figure 3.10






67

2
1
Profit simulation between stages 1 and 2 for the leader (  s,i *  c,i * )

varying the difference between the maximum willingness to pay for
a sustainable product and the cost of a sustainable product ( a s  c s,i )
c
for a variety of relative cost  s,i
 c c, i

a 
to pay  s 
a 
 c


 and relative maximum willingness



71

a
c
Figure 3.11.1 The relationship between s or s and  for the worst case scenario
cc
ac

73

a
c
Figure 3.11.2 The relationship between s or s and  for the best case scenario
cc
ac

73

xi

Figure 3.12.1 The relationship between a s  c s, i or c s,i * A and  for the worst case
scenario

74

Figure 3.12.2 The relationship between a s  c s, i or c s,i * A and  for the best case
scenario

75

Figure 3.13.1 Profit simulation between two firms in stage 1 varying the relative cost
 cc,i 

 for three levels of degree of substitution in stage 1 (  ), 0.8,
1
 cc, j 


0.5, and 0.2

77

Figure 3.13.2 Profit simulation between two firms in stage 1 varying the degree of
substitution in stage 1 (  1 ) for the three levels of the relative cost

 cc,i

 cc, j

Figure 3.14


 , 0.5, 1, and 2



78

Profit simulation between two firms in stage 2 (  s,i 2 *  c, j 2 * )
varying the maximum willingness to pay ( as ) for variety of relative
 c s, i
costs 
 c c, i


Figure 3.15

Figure 3.16

Figure 3.17


 and the relative maximum willingness to pay



 as 
 
a 
 c

Summary of decision making arranged by propositions, standard
findings, and their restrictions
The relationship between the expected maximum willingness to pay
for the sustainable product ( a s ) and the probability that the leader
firm expects that consumers are willing to pay more for the sustainable
product (  )

83

86

103

Leader profit simulation between stages 1 and 2 (  s, i 2 *  c, i1 * )
varying the probability of leader’s expectation that consumers are
willing to pay more for the sustainable product (  ) for two levels of
 c s, i 

the cost over the maximum willingness to pay 
 a  , 0.3 and 0.5.
 s 

xii

106

Figure 3.18

Profit simulation between two firms in stage 2 (  s, i 2 *  c, j 2 * )
varying the probability that the leader expects that consumers are
willing to pay more for the sustainable product (  ) for two levels of
 c s, i 

the cost over the maximum willingness to pay 
 a  , 0.3 and 0.5
 s 

Figure 3.19

Figure 3.20

108

Summary of decision making arranged by propositions, standard
findings and their restrictions when uncertainty is introduced

109

Profit simulation between stages 2 and 3 for the follower varying
the degrees of substitution (  c   s   3 ) when the level of relative
2
2
 c s, j 
 equals 1.66 (the worst case
costs of a sustainable product 
 c c, j 


scenario)

121

xiii

CHAPTER 1: INTRODUCTION

During the past few years, there has been an important trend observed along the food
supply chain, a “sustainable food trend.” This dissertation is about the “sustainable food product”
trend. More than 13,000 sustainable food and drink products were launched in the market from
2005-2010 (Browne 2010). Several key factors have supported the expansion of launching
sustainable products. First, consumers are becoming more aware of environmental problems and
have increasingly demanded manufacturers to produce food products in a more environmentally
friendly way (Oberholtzer, Greene, and Lopez 2006). Eighty-four percent of the U.S. consumers
interviewed indicated that they sometimes or regularly purchase sustainable foods and drinks,
especially those with “buy local” or “recyclable packaging” claims (Browne 2010). Moreover,
consumers’ food safety scares in last year (i.e., the impacts of the Gulf Coast oil spill on seafood,
and the impacts of egg recall from salmonella poisoning) may have increased consumers’
demand for sustainable food and drink products. This implies more market opportunities for
sustainable products.
Second, retailers acknowledge that there is an increase in consumers’ demand for
sustainable products and they must improve their strategies to respond to those demands.
Retailers are using their market power to encourage manufacturers to supply more sustainable
products. For example, Walmart, the world largest retailer, is currently creating a “Sustainability
Index” (targeting 2014 completion) that can measure the environmental performance of
suppliers, and improve efficiency by reducing costs and waste (Browne 2010). This
sustainability program is expected to influence not only other retailers but also manufacturers.

1

Third, manufacturers have gradually repositioned their strategies to be more sustainable.
The Accenture and United Nation Global Compact (UNGC) interviewed 766 CEOs around the
world in 2010 and found that manufacturers invest in sustainability for three reasons: brand, trust
and reputation (Broder 2010). Hence, launching sustainable products can be a public relation
strategy of the firms to improve their image and goodwill. Manufacturers also invest in
sustainable practices to reduce production costs and increase their competitiveness. The
Accenture and UNGC study also showed that CEOs realize that incorporating sustainability
practices can be a source of cost efficiency and revenue growth even during an economic
downturn (Broder 2010).
Lastly, in addition to the agents along the food supply chain, third party certifiers and
government agencies are also concerned about sustainability. Several standards and regulations
were implemented to support environmental and sustainable policies. For example, dolphin safe
and Marine Stewardship Council environmental standard for sustainable fishing are two
sustainable seafood product certifications. Governments have also announced regulations to
control pollution emissions from factories. These regulations are perceived to be increasing
sustainable production, which includes animal welfare policies in Europe and in California
1

(Proposition 2 of 2008) .
Some food manufacturers are reluctant to implement sustainable practices, or develop
and market new sustainable food products, although sustainability is at the forefront of most food
manufacturer CEOs’ minds. Part of their reluctance is due to a lack of information about price
premiums for sustainable attributes and a systematic decision making model.

1

Proposition 2 intends to improve animal production practices, such as, allowing animals to run
around freely, lie down, stand up, and fully extend their limbs outside their cage. This
proposition will become operative on January 1, 2015 (2008).
2

The objectives of this dissertation are to: analyze price premiums of sustainable
attributes; construct food firms’ innovation decision making model; explore under which
conditions firms launch sustainable products; and determine which conditions encourage a firm
to be a leader by launching a sustainable product first. Hedonic analysis is used to find the price
premiums of sustainable products while game theory, comparative static, and simulation are
employed to construct the firms’ decision making regarding under which conditions firms launch
sustainable products, and which conditions encourage a firm to be a leader.
The U.S. egg industry was selected to estimate the price premiums of sustainable
attributes, and simulate and test the firms’ decision making model. There are several reasons why
the U.S. egg industry is appropriate to study. First, eggs are an important commodity, which is
globally one of the cheapest sources of protein, and consumers make frequent purchases. Second,
eggs are not complicated food products, and consumers can easily understand the marketing
message from manufacturers about their sustainable attributes. Third, the new legislation
concerning animal welfare in egg production has influenced many egg manufacturers to market
sustainable eggs. Lastly, the cost and price data for conventional and sustainable eggs are
available.
The dissertation consists of an introduction (chapter 1), two independent papers (chapters
2 and 3) and a conclusion (chapter 4). The introduction includes the motivations and the reasons
why the U.S. egg industry was chosen. The core of the dissertation is composed of two
individual papers entitled: hedonic analysis of sustainable food products, and food
manufacturers’ sustainable product launch strategy: game theory approach. The conclusion
discusses the overall findings from the dissertation.

3

REFERENCES

4

REFERENCES

Ballotpedia. 2008. “California Proposition 2, Standards for Confining Farm Animals (2008).”
http://ballotpedia.org/wiki/index.php/California_Proposition_2_%282008%29.
Broder, Christina. 2010. “Accenture and UN Global Compact Study Finds Sustainability Now
Number One Priority for CEOs”. Big4.com. http://www.big4.com/news/accenture-andun-global-compact-study-finds-sustainability-now-number-one-priority-for-ceos/.
Browne, David. 2010. “Sustainable Food and Drink - US - August 2010”. Mintel.
http://academic.mintel.com.proxy2.cl.msu.edu/display/528334/?highlight=true#.
Mintel. 2011. Green Marketing-US-April 2011.
http://academic.mintel.com.proxy2.cl.msu.edu/sinatra/oxygen_academic/search_resul
ts/show&/display/id=543141 (accessed August 22, 2011).
Oberholtzer, Lydia, Catherine Greene, and Enrique Lopez. 2006. “Organic Poultry and Eggs
Capture High Price Premiums and Growing Share of Specialty Markets”. LDP-M-15001. Outlook Report from the Economic Research Service. United States Department of
Agriculture. http://www.ers.usda.gov/media/866670/ldpm15001_002.pdf.

5

CHAPTER 2: HEDONIC ANALYSIS OF SUSTAINABLE FOOD PRODUCTS

2.1 Introduction
Sustainability is one of the most important issues facing the global food supply chain.
There are 9,450 new food and beverage products claimed to be ethically or environmentally
1

produced globally from February 2009 to January 2010 (Mintel 2010). This represents almost
10% of all new food and beverage products (Mintel 2010). This nebulous concept has the ability
to change international trade patterns, make firms invest millions of dollars to change suppliers
(i.e. McDonalds’ sustainable supply chain (2010a), and change product components to minimize
damage to brand name image).
There are four critical reasons why analyzes need to be conducted on food manufacturers
2

and retailers concerning sustainability. First, from the CIES survey of the largest food
supermarket retailers globally, corporate social responsibility emphasizing sustainability was the
top issue that CEOs were concerned about in 2008 (“Top of Mind Survey” 2008). In the previous
years, sustainability was ranked 5th in 2007 and 11th in 2006. One of the main reasons
supermarket chains are concerned is that NGOs and customers are putting more pressure on them
to source “Sustainably,” and are being graded by certain NGOs, i.e., Greenpeace (2009a). CEOs
realize the importance of sustainability to the competitiveness of their businesses, but they are
not certain of which investments to make in order to strengthen their brands.

1

Ethical categories include ethical-animal, ethical charity, and ethical-human categories.
Moreover, Environmental categories include environmentally-friendly package, and
environmentally-friendly product categories.
2
The CIES survey is a survey of the Consumer Goods Forum which is an independent global
parity-based Consumer Goods network. (www.ciesnet.com)
6

Second, companies are trying to improve their supply chain by reducing costs and carbon
use simultaneously. The reduction in carbon emissions not only reduces the costs to the firm, but
may also promote the firm’s image and goodwill. For instance, in April 2008, Tesco announced
the launch of “The Carbon Reduction Label,” which focuses on energy usage and adopting the
concept of “sustainability” policies to its retail center and its own private brand products (“Tesco
and Carbon Trust Join Forces to Drive Forward Carbon Labeling” 2008). Promoting energy
saving is popular, for example, many companies began to use wind energy, and reclaim cooking
oil and solar energy to substitute gas in their production processes (Weil 2008).
Third, consumers are becoming more aware of environmental problems and are interested
in consuming products that are considered to be sustainably produced. This has led to a growing
3

number of green consumers . In the United States, the growth of consumers who are always or
almost always green consumers increased from 12% in 2006 to 36% in 2007 (O’Donnell 2008).
This implies more market opportunities for sustainable products since consumers are willing to
pay for high quality products as well as products that help improve the environment.
Lastly, several standards and regulations were implemented to support environmental and
sustainable policies. The examples of voluntary standards related to the environment are the ISO
14000 series. Also, there are several certifications for sustainable seafood products, such as,
Marine Stewardship Council’s fishery certification program and seafood eco-label, and dolphin
safe label. Governments have announced regulations to control pollution emissions from
factories as well. These regulations are perceived to be increasing sustainable production, which
includes animal welfare policies in Europe and in California (Proposition 2 of 2008).

3

A green consumer is a person who is concerned about environmental or social issues
constantly when deciding on purchasing products (Peattie and Charter 1992).
7

Due to the pressure from retailers, consumers, legislation and competition, more
sustainable food products are being launched (i.e. dolphin safe tuna products, cereal with
recyclable packaging, and free-range and other environmentally-friendly eggs). This study
analyzes the value of sustainable attributes for fresh eggs by using hedonic price analysis and
survey data of fresh egg prices in five city areas along the Eastern coast of the United States.
Sustainability has been defined by various organizations and companies and has led to a
brand definition. Therefore, we will focus only on why two attributes in this study represent
sustainable attributes. The first attribute is a welfare-managed attribute. In our study, welfare4

managed eggs include free-range eggs and free-cage eggs . To understand why the welfaremanaged attribute represents a sustainable attribute, the concept of sustainable agriculture is
introduced. According to the USDA, sustainable agriculture is defined as an integrated system of
plant and animal production practices that has a site-specific application that will occur over the
long term (United States Department of Agriculture 2007). Also, there are many approaches to
define “animal welfare”. A well-known definition is that ‘welfare’ is the state of a being in
relation to its environment (Broom 1991; Blandford et al. 2002). The conventional process for
raised hens is a battery cage system which provides space of 67 to 86 inches per bird (2010b);
consequently, hens in battery cages do not have enough space for free movement. Welfaremanaged systems including free range/cage free systems can improve animal welfare by
allowing them to extend their limbs freely. Hence, the welfare-managed attribute represents one
of the sustainable attributes as stated in Bennett (1998) “Consumers who are concerned about

4

There is no legal definition for free-range and free-cage eggs in the U.S.; however, according
to the Egg Nutrition Center, free-range eggs are from hens that are either raised outdoors or can
access outside. Free-cage eggs are from hens that live in indoor floor facilities, but do not
necessarily have access to the outdoors.
8

animal welfare prefer and are willing to pay more for methods of animal husbandry that allow
hens to roam freely instead of being in cages”.
The second attribute representing a sustainable attribute is paper-pulp packaging since
sustainability also includes an environmental dimension of recycling. For example, Spartan
Stores and Wegmans changed their packaging of their store-brand eggs to be new recyclable and
biodegradable molded fiber packaging to replace Styrofoam cartons that are not biodegradable
(2009b; Wegmans 2010). Therefore, paper-pulp packaging, which is recyclable and/or made
from recycled material, is considered to be one of the sustainability attributes in this study.

2.2 Objectives
This study aims to determine price premiums for sustainable attributes of fresh eggs by
using hedonic analysis. The sustainable attributes defined here include the free-range/cage-free
attribute and recyclable attribute for packaging which is paper-pulp packaging. Furthermore, we
will focus on the interaction between organic and sustainable attributes; that is, whether
sustainable attributes of eggs have a higher value when eggs are organic.
This work is unique for several reasons. First, most of the literature focuses on analyzing
the value of organic attributes more than sustainable attributes. Examples of papers that analyzed
the price premiums of organic products are: Gil, Gracia, and Sanchez 2000; Canavari et al. 2002;
Soler, Gil, and Sánchez 2002; Ara 2003; Wang and Sun; Batte et al. 2007; Griffith and Nesheim
2008. Second, most of the literature concerning price premiums for sustainable attributes used
the contingent valuation approach (Loureiro, McCluskey, and Mittelhammer 2001; Loureiro and
Hine 2002; Loureiro, McCluskey, and Mittelhammer 2002). Third, the unique data set was
collected from five East coast U.S. cities and has not been analyzed for sustainable attributes for
9

fresh eggs and the economic implications thereof. Moreover, there is no literature on price
premiums for sustainable attributes of eggs in the U.S. Most egg literature studied specialty egg
characteristics and the overall U.S. egg industry (Patterson et al. 2001; Knudson 2004;
Oberholtzer, Greene, and Lopez 2006; Patterson et al. 2008). Lastly, recent studies suggest that
eco-labels, an example of a sustainable attribute, should be added to complement other valued
product attributes such as organic attribute in order to attract more consumer purchases (Johnston
et al. 2001; Arquitt and Cornwell 2007). Hence, this work also aims to test the hypothesis that
multi-attribute eggs such as sustainable attributes and organic eggs are more valued.

2.3 Egg Industry
The egg industry is a great industry to better understand consumer evaluations of
sustainable attributes of food products for several reasons. First, the fresh egg industry in the
U.S. is a huge and important industry which had a market size equal to $ 5.73 billion in 2010
(Patterson 2011). Second, quality survey data for egg prices and their attributes in key eastern
U.S. cities are available. Third, eggs are not complicated food products and consumers can easily
understand the marketing messages, and the sustainable attributes are easily included into our
model. Fourth, organic eggs and free-range/cage-free are easily understood and well known
5

attributes among egg consumers . Lastly, due to the vote for proposition 2 in California in 2008,

5

Organic regulations require outdoor access for birds (Oberholtzer, Greene, and Lopez 2006);
therefore, organic eggs are a subset of free-range/cage-free eggs. However, we define organic
and welfare-managed attributes separately because we are interested in the interaction between
these two attributes. Egg manufactures sometimes label their organic eggs as cage-free eggs;
while, others do not. Consumers might be confused whether organic eggs are welfare-managed
eggs or not. The study is based on consumers’ perception; therefore, we identify the attributes of
each observation based on information on the label.
10

the industry recognizes the importance of free range/cage free in the future to their market and
the potential for this movement to spread across America.
There are two main segments for the egg market which are fresh shell eggs, and egg
6

substitutes . In 2010, fresh eggs had a market share equal to 93.6%, while egg substitutes had a
market share of only 6.4% (Patterson 2011). Hence, this study focuses only on the fresh egg
market. There are two types of fresh eggs, which are regular eggs and specialty eggs. Examples
of specialty eggs are free-range/cage-free eggs, organic eggs, eggs fortified with Omega-3 fatty
acids, low-cholesterol eggs, and vegetarian-fed eggs.
7

Store brands dominate national brands and regional brands in the fresh egg market. In
2010, store brands had a market share equal to 66.8%, while Eggland’s Best, Cal-Maine Foods,
Rose Acre Farms Inc., and others had market shares equal to 9.9%, 2.6%, 2.0%, and 18.8%,
respectively (Patterson 2011).

2.4 Methodology
Lancaster (1966) stated that a good does not give utility directly to a consumer, but it
possesses characteristics or attributes which give utility to the consumer. Hedonic prices are
defined as the implicit prices of attributes embodied in each good (Rosen 1974). Economic
agents can determine hedonic prices of attributes by observing prices of differentiated products
and specific amounts of attributes related to them (Rosen 1974). Examples of attributes are
brand, packaging, color, taste, etc. If a good has a number of characteristics or attributes, z , equal
6

Breaker or breaker plant category is not in the scope of this study because our study focuses on
consumer goods. Breakers are industrial goods which are not available in supermarkets but are
used in restaurants, hospitals, schools, and other food service (United States Department of
Agriculture 2010).
7
Store brand is interchangeable with private label.
11

to k , z  ( z1, z2 ,..., zk ), the price for a good is determined by a set of attributes or vector z , that
is price(z)  f ( z1, z 2 ,..., z k ). Hedonic pricing analysis and contingent valuation are the two
main approaches used to calculate price premiums of unique attributes. The contingent valuation
requires consumer survey data to determine if the premium of each attribute has value.
Numerous papers have utilized this approach to address price premiums for food product
attributes (Wessells, Johnston, and Donath 1999; Gil, Gracia, and Sanchez 2000; Loureiro,
McCluskey, and Mittelhammer 2001; Canavari et al. 2002; Loureiro and Hine 2002; Loureiro,
McCluskey, and Mittelhammer 2002; Ara 2003; Cranfield and Magnusson 2003; Batte et al.
2007). The weakness of this approach is that it only reflects consumers’ intentions but not their
actual actions in terms of purchasing behavior. Moreover, it is possible that the survey might
create a bias in the sense that consumers might over-estimate their willingness to pay for
sustainable products, which leads to the problem of over-estimating the price premium for
sustainable attributes. Several papers analyzed or have referred to the biases of the contingent
valuation approach (Diamond and Hausman 1994; Blumenschein et al. 1998; Aadland and
Caplan 2003; Ajzen, Brown, and Carvajal 2004; Blumenschein et al. 2008).
8

Historically, hedonic analysis primarily has used scanner data or privately collected
secondary data. Several authors used hedonic analysis for measuring a price premium of
differentiated food product (wine, coffee, etc.) attributes (Nimon and Beghin 1999; Combris,
Lecocq, and Visser 2000; Donnet, Weatherspoon, and Hoehn 2008; Griffith and Nesheim 2008).
The data for this study was collected from retailers who are concerned about consumer demand

8

Scanner data are “retail purchase information (such as price, brand, product size, amount
purchased) gathered at the point of purchase by an electronic device that reads a coded ticket on
the product through the use of an electronic reader over which the product passes.” (www.
Answer.com)
12

and maximize their profits by determining the optimal attributes, prices and quantities to offer
(Steiner 2004; Karipidis et al. 2005). The partial derivative of the hedonic price function with
respect to a particular attribute is an implicit or shadow price at equilibrium that reflects both, the
maximum price consumers are willing to pay for an additional attribute, and the minimum price
for which suppliers are willing to sell according to their costs (Sanjuán-López, Resano-Ezcaray,
and Camarena-Gómez 2009). Moreover, consumers decide whether they should accept the price
and purchase the eggs or not based on the retailers’ offered price. Therefore, the price and
attributes collected from retailers can be used to find the value of attributes by using hedonic
analysis without ignoring the consumer side.
There are two advantages of using hedonic price analysis over contingent valuation.
First, the hedonic price approach does not require joint consumption of goods within a group.
Therefore, we can estimate the inverse demand of specific goods individually rather than
modeling the whole system of demand and supply. Second, according to Butler (1982), since all
estimates of hedonic price models are to some extent misspecified, models that use a small
number of key variables generally suffice. Butler suggested that only those attributes that are
costly to produce and yield utility are to be considered in the regression equation. Therefore, we
need to use less attributes in our model so that we reduce the misspecification problem and
increase the degrees of freedom.
Assume that an egg has k attributes plus sustainable attributes, organic attribute, and a
sustainable and organic attribute. The egg price then depends on its attributes (Rosen 1974)
defined as follows:

price (x)

= f ( x1, x2 ,..., xk , sustainable attribute, organic attribute, sustainable and organic
attribute),
13

where price (x) represents the price of an egg, and vector x represents attributes of the egg.
Specifically, the model in our study is specified as the following:
priceperegg =

 0  1o   2 wm  3owm   4vd  5ne
  6 regional   7 national  8brown   9 AA
 10 plastic  11 paper  12l arg e  13extral arg e  14 jumbo
 15acme  16 giant  17 pathmark  18safeway  19shopper
  20shaws   21shoprite  22 sup erfresh  23stopandsho
p
  24wegmans   25weis
  26shelllabel  27eggage  28unitsize  ,

where  ’s represent the coefficients for the product attributes and  is the error term. The
definitions, minimums, maximums, and means of each variable are depicted in table 2.1. In this
model, the base variables for each category of dummy variable attributes are dropped in order to
prevent perfect multicollinearity.

2.5 Data and Variable Description
9

The data used in our analysis are survey data of fresh egg prices and their attributes . The
data have 207 usable observations and were collected from retailers in five east coast cities
(Baltimore, MD; Boston, MA; New York, NY; Philadelphia, PA; and Washington DC) in 2007.
The data come from retail supermarkets (ACME, Giant, Pathmark, Safeway, Shoppers Food
Warehouse, Shaw’s, ShopRite, Super Fresh, Stop and Shop, Walmart, Wegmans, and Weis) in
each of these cities.

9

We would like to thank Dr. Paul H. Patterson, from the Poultry Science Department at Penn
State University for providing us with the data.
14

From table 2.1 (see appendix A), the first group of attributes is a group of specialty
characteristics of the eggs which are regular, organic, vegetarian-fed, welfare-managed including
10

free-range and free-cage, nutritionally enhanced , and a stacked attribute, organic and welfaremanaged. The second group is categorized by brand. To preserve the degrees of freedom, we
separate egg brands into three groups which are national, regional and store brands. The third
group is categorized by colors which are white and brown. The fourth attribute is grouped by
grades of eggs (grades A and AA), which reflect the quality and the freshness of the eggs, i.e.,
the firmness of the yolk, and the air cell in the egg. The fifth group is defined by packaging
materials which are Styrofoam, paper pulp, and clear plastic. The sixth group is determined by
egg sizes which are medium, large, extra large, and jumbo. The seventh group of attributes is
determined by the retailers where consumers purchase eggs (ACME, Giant, Pathmark, Safeway,
Shoppers Food Warehouse, Shaw’s, ShopRite, Super Fresh, Stop and Shop, Walmart, Wegmans,
and Weis). The eighth attribute is defined based on whether there is a label on the egg shell or
not. The next variable is the age of the egg that is defined as the number of days from when an
11

egg is laid until it is purchased at the store . The last attribute is an egg unit which is the
number of eggs per package.

10

From our survey data, nutritionally-enhanced eggs are high-omega 3, high-vitamins, and lowcholesterol.
11
Egg cartons with the USDA grade shield on them are regulated to display the "pack date"
which is defined as the day that the eggs were washed, graded, and placed in the carton (F. S.
and I. S. United States Department of Agriculture 2007). We get the information about the age of
the egg by using the pack date and assuming that eggs are packed the same day as they are laid.
15

Credence Goods
The attributes can be categorized into three categories which are search, experience, and
credence attributes (Caswell and Mojduszka 1996; Bureau, Marette, and Schiavina 1998;
Loureiro, McCluskey, and Mittelhammer 2002; De Pelsmacker, Driesen, and Rayp 2005).
Search attributes are those that consumers can observe immediately before purchase, i.e. color,
size, and price. Experience attributes, such as taste, are attributes that consumers discover only
after consumption. Credence attributes are attributes of which consumers can detect the quality
neither before nor after buying the product. The ethical attribute, such as cage-free, is an example
of a credence attribute. This leads to the problem of asymmetric information in the cage-free egg
market.
Asymmetric information is addressed by manufacturers labeling their products; however,
the credibility of manufactures is critical to getting price premiums and higher profits. Third
Party Certification proof with high public trust can increase ethical label credibility (Loureiro,
McCluskey, and Mittelhammer 2002; De Pelsmacker, Driesen, and Rayp 2005); however, there
is no well-known certification for cage-free eggs in the U.S. market. Consequently, reputation of
egg manufactures is the only signal for the cage-free attribute, and U.S. consumers might be still
confused and reluctant to trust cage-free labels, which could lead to low cage-free eggs
purchasing.

Egg Packaging
There are three types of material for egg packaging which are paper pulp, clear plastic,
and Styrofoam. Paper-pulp packaging is claimed to be recyclable and made from recycled paper.
Clear plastic and Styrofoam are technically plastics and recyclable. Clear plastic packaging for
16

eggs is made from polyolefins and defined as code 1 (Polyethylene terephthalate: PET)
recyclable symbol. Styrofoam packaging is made from polystyrene (PS) and defined code 6 for
its recyclable symbol (Marsh and Bugusu 2007).
Even though all materials for egg packaging are recyclable, paper and paperboard have
the highest recycle rate. In 2007, 54.5% of paper and paperboard was recovered for recycling;
while, plastics including Styrofoam had a recycle rate equal to 6.8% (United States
Environmental Protection Agency 2008). Moreover, some egg manufactures marketed their eggs
by changing material for their packaging from Styrofoam to paper pulp and claimed that their
new packaging was more environmentally-friendly. For example, Spartan Store and Wegmans
changed their egg packaging from Styrofoam packaging to paper-pulp packaging and claimed
that their packaging is more sustainable or more environmentally-friendly (2009b; Wegmans
2010). As a consequence, this study used paper-pulp packaging as its sustainable packaging
attribute.

2.6 Results
Table 2.2 (see appendix A) presents hedonic prices of egg attributes from the estimation.
The R-squared for the model shows that all egg attributes explain about 81.2% of the variation in
the prices of eggs. The attributes that significantly affect the price of eggs are specialty
characteristics, brands, grades, sizes, retailers (places where consumers buy eggs), and unit sizes.
Signs of significant variables are as expected and the same as previous literature (Ness and
Gerhardy 1994; Fearne and Lavelle 1996; Karipidis et al. 2005; Goddard et al. 2007) except the
sign for the stacked variable attribute organic and welfare-managed (owm).

17

Most specialty characteristic coefficients which are organic attribute (o), welfaremanaged attribute (wm), and nutritionally-enhanced attribute (ne) have positive values and are
significant. Organic, welfare-managed, and nutritionally-enhanced eggs have price premiums
over regular eggs equal to 16.50, 3.57, and 2.30 cents per egg, as shown in figure 2.1
respectively. This means that these attributes create value-added for the shell egg category. The
coefficient for the vegetarian-fed attribute is a negative value, but not significant. Hence, it is
ambiguous to conclude the value of the vegetarian-fed attribute.
The coefficient for the stacked variable attribute, organic and welfare-managed, equals 8.81 cents per egg and is significant. Therefore, an organic and welfare-managed egg has a
premium over a regular egg equal to 11.26 cents which is less than the premium for an organic
egg (11.26 = 16.50 + 3.57 – 8.81 cents representing the premium for organic, welfare-managed,
and organic and welfare-managed attributes). The authors did not expect the negative sign for the
stacked variable. We expected that welfare-managed eggs would get higher premiums when they
are also marketed (labeled) as organic because consumers can easily associate the perceived
animals health benefits and be willing to pay a premium for it. There are three hypotheses to
explain this result. First, consumers might be confused about the definition of eggs with these
attributes and hence not be willing to pay more for the stacked attributes. Second, it might be
possibly related to retailers’ strategies (Greenblum 2009)

12

to promote theirs store brands as

sustainable brands; hence, offer promotions for the organic and welfare-managed products.
Lastly, farmers might be able to share some production costs for the organic, free-range and/or
free-cage methods; hence, the prices reflect supply and demand side effects. The prices of
regular eggs and specialty eggs are compared in figure 2.1.

12

Ms. Greenblum is a senior director of Nutrition Education, Egg Nutrition Center.
18

Figure 2.1 Prices of the base level and specialty eggs

13

(cents per egg)

National brand eggs and regional brand eggs have price premiums equal to 5.33 cents and
3.95 cents compared to store brand eggs. Prices of grade AA eggs are significantly higher than
prices for grade A eggs. Its price premium equals to 3.28 cents. All coefficients of sizes are
significant. That is the larger size egg has a higher price premium. The coefficient for unit size is
negative and significant. Therefore, the price per egg is lower when consumers buy eggs in
bigger packages. Eggs from almost all retailers

13

14

have significantly higher prices than the price

Assuming that other attributes are the same, the base level egg for each category is defined as
an egg with the following attributes: regular, store brand, white color, grade A, Styrofoam
packaging, medium size, no shell label, egg age of 14 days, a dozen egg unit size, and Walmart
is the base store. Specialty eggs have the same attributes as the base level eggs except they are
not regular eggs.
14
From table 2.2, these retailers are ACME, Giant, Pathmark, Safeway, Shoppers Food
Warehouse, Shaw’s, ShopRite, Super Fresh, Stop and Shop, Wegmans, and Weis.
19

of eggs from Walmart. Lastly, the coefficients of the rest of the variables which are various types
of packaging, brown color, shell label and egg age are all insignificant.

2.7 Conclusion and Management Implications
We tested two attributes that we consider sustainable, welfare-managed, and paper-pulp
packaging; only one was found to positively and significantly influence price. Welfare-managed
eggs receive the price premium equals to 3.57 cents per egg as compared to regular egg. The
attribute that has the greatest impact on price was the organic attribute which increase the price
per egg by 16.50 cents. Interestingly, when organic and welfare-managed were combined the
price premium was only 11.26 cents per egg. This implies that consumers are not willing to pay
for both labeled attributes simultaneously, which has major implication for egg manufactures and
retailers. In the short run, egg manufactures should maximize profit by offering and labeling
either organic or welfare-managed eggs, and hence continue to segment the market until
consumers perceive these attributes as being different.
Our results are ambiguous for the paper-pulp packaging attribute. Some egg
manufacturers have claimed that their Styrofoam packaging and/or clear-plastic packaging are
recyclable. A survey of consumers’ perception about recyclable packaging might be helpful to
answer this question; however, it is beyond the scope of this study. The best strategy for
manufactures and retailers may be to market sustainable packaging for each specific region of
the U.S. since each state has different laws and opportunities to recycle plastic and Styrofoam
products.

20

APPENDIX

21

APPENDIX A

Table 2.1 Definitions of the Variables and their Descriptive Statistics
Variables
Dependent
variable
Price per egg
Specialty
Characteristics
r, o, wm, owm,
vd, and ne

Brands
store, regional,
and national

Colors
white, and brown

Grades
A, and AA

Types of
Packaging
foam, plastic,
and paper

Definition

Min
(cent)

Max
(cent)

0.06633

0.59667

0.23698

DV* which is 1 for
regular (r), organic
(o), welfare-managed
(wm), organic and
welfare-managed
(owm), vegetarian-fed
(vd), and nutritionallyenhanced (ne) eggs,
respectively and 0
otherwise

0

1

0.4198,
0.1481,
0.2305,
0.1111,
0.3868,
and
0.2593

Regular (r)

DV which is 1 for that
type of brand, and 0
otherwise

0

1

0.4139,
0.2664,
and
0.3197

Store brand
(store)

DV which is 1 for a
white (brown) egg,
and 0 for a brown
(white) egg

0

1

0.4321,
and
0.5679

White color
(white)

DV which is 1 for an
egg is grade A (AA),
and 0 if an egg is
grade AA (A)

0

1

0.9508,
and
0.0492

Grade A
(A)

DV which is 1 for an
egg package made
from Styrofoam,
plastic, and paperpulp, respectively and
0 otherwise

0

1

0.2025,
0.4298,
and
0.3678

Styrofoam
(foam)

A price per egg

22

Mean
(cent)

Base
Variables

Table 2.1 (cont’d)
Variables
Size
Medium, large,
extra large, and
jumbo

Retailers
Acme, Giant,
Pathmark,
Safeway,
Shopper, Shaws,
Shoprite,
Superfresh, Stop
and Shop,
Walmart,
Wegmans, and
Weis

Others
Shell label

Definition

Min
(cent)

Max
(cent)

Mean
(cent)

Base
Variables

DV which is 1 for a
medium, large, extralarge, and jumbo egg,
respectively, and 0
otherwise

0

1

0.0459,
0.7156,
0.1651,
and
0.0734

Medium size
(medium)

DV which is 1 for an
egg sold by that
retailers and 0
otherwise

0

1

0.0451,
0.0697,
0.1393,
0.1189,
0.0533,
0.0902,
0.0984,
0.0902,
0.0820,
0.0984,
0.0656,
and
0.0492

Walmart
(walmart)

0

1

0.1681

No shell label
(no shell
label)

1

41

14.02

6

60

12.45

DV which is 1 for an
egg with shell label,
and 0 otherwise
Egg age
a number of days
counted from when an
egg is laid until it is
bought at the store
Unit size
a number of eggs per
unit
*Note: DV represents a dummy variable.

23

Table 2.2 Results for Hedonic Prices of Egg Attributes
Attributes

Coefficient (S.E.)
Unit: Dollars
Price per egg

Dependent variable
Specialty Characteristics
o
0.165***
(-0.023)
wm
0.0357***
(-0.0122)
owm
-0.0881***
(-0.0267)
vd
-0.0065
(-0.0089)
ne
0.0230***
(-0.00877)
Brands
regional
0.0395***
(-0.0125)
national
0.0533***
(-0.00939)
Colors
brown
0.00186
(-0.00834)
Grades
AA
0.0328**
(-0.0127)
Types of packaging
plastic
0.0106
(-0.0114)
pulp
-0.00306
(-0.00936)
Sizes of eggs
large
0.0456***
(-0.0119)
extra large
0.0575***
(-0.0119)
jumbo
0.0715***
(-0.0135)
Retailers
Shaws
0.0601***
(-0.0116)
Stop and shop
0.0540***
(-0.0167)
Giant
0.0597***
(-0.0117)
Safeway
0.116***
(-0.0149)
Wegmans
0.00609
(-0.0145)
Weis
0.0393**
(-0.0189)
Shopper
-0.0125
(-0.0124)
Pathmark
0.0775***
(-0.0133)
Shoprite
0.0634***
(-0.0128)
Superfresh
0.0502***
(-0.013)
Acme
0.0645***
(-0.0146)
Others
shell label
0.00425
(-0.0121)
egg age
0.000214
(-0.000371)
unit size
-0.00149**
(-0.000728)
0.0897***
(-0.0174)
Constant
207
Observations
0.812
R-squared
Note: * significant at 10%; ** significant at 5%; *** significant at 1%
24

REFERENCES

25

REFERENCES

Aadland, David, and Arthur J. Caplan. 2003. “Cheap Talk Revisited: New Evidence from
CVM.” In 2003 Annual Meeting, July 27-30, Montreal, Canada. American Agricultural
Economics Association (New Name 2008: Agricultural and Applied Economics
Association).
Ajzen, Icek, Thomas C. Brown, and Franklin Carvajal. 2004. “Explaining the Discrepancy
Between Intentions and Actions: The Case of Hypothetical Bias in Contingent
Valuation.” Personality and Social Psychology Bulletin 30 (9): 1108–1121.
Ara, Shihomi. 2003. “Consumer Willingness to Pay for Multiple Attributes of Organic Rice: A
Case Study in the Philippines.” In 25th International Conference of Agricultural
Economists, Durban, South Africa.
Arquitt, Steven P., and T. Bettina Cornwell. 2007. “Micro-macro Linking Using System
Dynamics Modeling: An Examination of Eco-labeling Effects for Farmed Shrimp.”
Journal of Macromarketing 27 (3): 243–255.
Batte, Marvin T., Neal H. Hooker, Timothy C. Haab, and Jeremy Beaverson. 2007. “Putting
Their Money Where Their Mouths Are: Consumer Willingness to Pay for Multiingredient, Processed Organic Food Products.” Food Policy 32 (2): 145–159.
Bennett, Richard. 1998. “Measuring Public Support for Animal Welfare Legislation: A Case
Study of Cage Egg Production.” Animal Welfare 7 (1): 1–10.
Blandford, David, Jean-Christophe Bureau, Linda Fulponi, and Spencer Henson. 2002.
“Potential Implications of Animal Welfare Concerns and Public Policies in Industrialized
Countries for International Trade.” In Global Food Trade and Consumer Demand for
Quality, 77–99. Springer.
Blumenschein, Karen, Glenn C. Blomquist, Magnus Johannesson, Nancy Horn, and Patricia
Freeman. 2008. “Eliciting Willingness to Pay Without Bias: Evidence from a Field
Experiment*.” The Economic Journal 118 (525): 114–137.
Blumenschein, Karen, Magnus Johannesson, Glenn C. Blomquist, Bengt Liljas, and Richard M.
O’Conor. 1998. “Experimental Results on Expressed Certainty and Hypothetical Bias in
Contingent Valuation.” Southern Economic Journal: 169–177.
Broom, Donald M. 1991. “Animal Welfare: Concepts and Measurement.” Journal of Animal
Science 69 (10): 4167–4175.
Bureau, Jean-Christophe, Stephan Marette, and Alessandra Schiavina. 1998. “Non-tariff Trade
Barriers and Consumers’ Information: The Case of the EU-US Trade Dispute over Beef.”
European Review of Agricultural Economics 25 (4): 437–462.

26

Butler, Richard V. 1982. “The Specification of Hedonic Indexes for Urban Housing.” Land
Economics 58 (1): 96–108.
Canavari, Maurizio, Guido Maria Bazzani, Roberta Spadoni, and Domenico Regazzi. 2002.
“Food Safety and Organic Fruit Demand in Italy: a Survey.” British Food Journal 104
(3/4/5): 220–232.
Caswell, Julie A., and Eliza M. Mojduszka. 1996. “Using Informational Labeling to Influence
the Market for Quality in Food Products.” American Journal of Agricultural Economics
78 (5): 1248–1253.
Combris, Pierre, Sébastien Lecocq, and Michael Visser. 2000. “Estimation of a Hedonic Price
Equation for Burgundy Wine.” Applied Economics 32 (8): 961–967.
Cranfield, John AL, and Erik Magnusson. 2003. “Canadian Consumer’s Willingness-to-pay for
Pesticide Free Food Products: An Ordered Probit Analysis.” International Food and
Agribusiness Management Review 6 (4): 13–30.
Diamond, Peter A., and Jerry A. Hausman. 1994. “Contingent Valuation: Is Some Number
Better Than No Number?” The Journal of Economic Perspectives 8 (4): 45–64.
Donnet, M. Laura, Dave D. Weatherspoon, and John P. Hoehn. 2008. “Price Determinants in
Top‐quality E‐auctioned Specialty Coffees.” Agricultural Economics 38 (3): 267–276.
Fearne, Andrew, and David Lavelle. 1996. “Perceptions of Food ‘quality’ and the Power of
Marketing Communication: Results of Consumer Research on a Branded-egg Concept.”
Journal of Product & Brand Management 5 (2): 29–42.
Gil, Josè M., Azucena Gracia, and Mercedes Sanchez. 2000. “Market Segmentation and
Willingness to Pay for Organic Products in Spain.” The International Food and
Agribusiness Management Review 3 (2): 207–226.
Goddard, Ellen Willoughby, Peter Boxall, John Paul Emunu, Curtis Boyd, Andre Asselin, and
Amanda Neall. 2007. Consumer Attitudes, Willingness to Pay and Revealed Preferences
for Different Egg Production Attributes: Analysis of Canadian Egg Consumers.
Department of Rural Economy, University of Alberta.
Greenblum, Marcia D. 2009. “Retailers’ Strategies About Sustainable Eggs.”
Greenpeace. 2009a. “More Than Half of Leading U.S. Supermarkets Show Progress in Latest
Greenpeace Sustainable Seafood Scorecard.” http://www.greenpeace.org/usa/en/mediacenter/news-releases/more-than-half-of-leading-u-s/.
Griffith, Rachel, and Lars Patrick Nesheim. 2008. “Household Willingness to Pay for Organic
Products”. Centre for Microdata Methods and Practice, Institute for Fiscal Studies.
Johnston, Robert J., Cathy R. Wessells, Holger Donath, and Frank Asche. 2001. “Measuring
Consumer Preferences for Ecolabeled Seafood: An International Comparison.” Journal of
Agricultural and Resource Economics: 20–39.
27

Karipidis, Philippos, Efthimia Tsakiridou, Nikolaos Tabakis, and Konstantinos Mattas. 2005.
“Hedonic Analysis of Retail Egg Prices.” Journal of Food Distribution Research 36 (3):
68–73.
Knudson, William A. 2004. “The Market for Organic and Fortified Eggs”. Working Paper no. 20104. The Strategic Marketing Institute, MSU Product Center for Agriculture and Natural
Resources. http://www.productcenter.msu.edu/documents/Working/2-0104.pdf.
Lancaster, Kelvin J. 1966. “A New Approach to Consumer Theory.” The Journal of Political
Economy 74 (2): 132–157.
Loureiro, Maria L., and Susan Hine. 2002. “Discovering Niche Markets: A Comparison of
Consumer Willingness to Pay for Local (Colorado Grown), Organic, and GMO-free
Products.” Journal of Agricultural and Applied Economics 34 (3): 477–488.
Loureiro, Maria L., Jill J. McCluskey, and Ron C. Mittelhammer. 2001. “Assessing Consumer
Preferences for Organic, Eco-labeled, and Regular Apples.” Journal of Agricultural and
Resource Economics: 404–416.
———. 2002. “Will Consumers Pay a Premium for Eco‐labeled Apples?” Journal of Consumer
Affairs 36 (2): 203–219.
Marsh, Kenneth, and Betty Bugusu. 2007. “Food Packaging—roles, Materials, and
Environmental Issues.” Journal of Food Science 72 (3): R39–R55.
McDonald. 2010a. McDonald’s Corporate Responsibility: Values in Practice Video.
http://www.aboutmcdonalds.com/mcd/csr/video/sustainable_supply.10.html.
Mintel. 2010. “New Food and Beverage Products.” http://www.gnpd.com/sinatra/home/.
Ness, Mitchell R., and Hubert Gerhardy. 1994. “Consumer Preferences for Quality and
Freshness Attributes of Eggs.” British Food Journal 96 (3): 26–34.
Nimon, Wesley, and John Beghin. 1999. “Are Eco-labels Valuable? Evidence from the Apparel
Industry.” American Journal of Agricultural Economics 81 (4): 801–811.
O’Donnell, Fiona. 2008. “Green Living - US - February 2008”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/display/295904/#.
Oberholtzer, Lydia, Catherine Greene, and Enrique Lopez. 2006. “Organic Poultry and Eggs
Capture High Price Premiums and Growing Share of Specialty Markets”. LDP-M-15001. Outlook Report from the Economic Research Service. United States Department of
Agriculture. http://www.ers.usda.gov/media/866670/ldpm15001_002.pdf.
Patterson, Bill. 2011. “Eggs - US - May 2011”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/display/543044/?highlight=true#.

28

Patterson, P. H., K. W. Koelkebeck, D. D. Bell, J. B. Carey, K. E. Anderson, and M. J. Darre.
2001. “Egg Marketing in National Supermarkets: Specialty Eggs--part 2.” Poultry
Science 80 (4): 390–395.
Patterson, P.H., H.S. Thesmar, M.J. Reynnells, M.J. Darre, and T.L. Cravener. 2008. “Specialty
Eggs in North America.”
Peattie, Ken, and Martin Charter. 1992. “Green Marketing.” The Marketing Book 5: 726–755.
De Pelsmacker, Patrick, Liesbeth Driesen, and Glenn Rayp. 2005. “Do Consumers Care About
Ethics? Willingness to Pay for Fair‐Trade Coffee.” Journal of Consumer Affairs 39 (2):
363–385.
Progressivegrocer. 2009b. “Spartan Store Brand Eggs Now in Green Packaging.” January 23.
http://www.progressivegrocer.com/progressivegrocer/content_display/features/corporatesocial-responsibility/e3i8e1f30d51da4f8e86906c7d215fef836.
Rosen, Sherwin. 1974. “Hedonic Prices and Implicit Markets: Product Differentiation in Pure
Competition.” The Journal of Political Economy 82 (1): 34–55.
Sanjuán-López, Ana Isabel, Helena Resano-Ezcaray, and Dena María Camarena-Gómez. 2009.
“Developing Marketing Strategies for Jiloca Saffron: a Price Hedonic Model.” Spanish
Journal of Agricultural Research 7 (2): 305–314.
Soler, Francisco, José M. Gil, and Mercedes Sánchez. 2002. “Consumers’ Acceptability of
Organic Food in Spain: Results from an Experimental Auction Market.” British Food
Journal 104 (8): 670–687.
Steiner, Bodo E. 2004. “Australian Wines in the British Wine Market: a Hedonic Price
Analysis.” Agribusiness 20 (3): 287–307.
“Tesco and Carbon Trust Join Forces to Drive Forward Carbon Labeling.” 2008. CarbonTrust.
http://www.carbontrust.co.uk/news/news/archive/2008/Pages/29_04_08_Carbon_Label_
Launch.aspx.
“Top of Mind Survey.” 2008. http://www.ciesnet.com.
United Egg Producers. 2010b. “Housing, Space, Feed, and Water.”
http://www.uepcertified.com/program/guidelines/categories/housing-space-feed-water.
United States Department of Agriculture, Food Safety and Inspection Service. 2007. “Food
Labeling.” http://www.fsis.usda.gov/factsheets/food_product_dating/index.asp.
———. 2010. “Fact Sheets: Egg Products Preparation – Shell Eggs from Farm to Table.”
http://www.fsis.usda.gov/factsheets/focus_on_shell_eggs/index.asp.
United States Department of Agriculture, National Agricultural Library. 2007. “Sustainable
Agriculture: Definitions and Terms.”
http://www.nal.usda.gov/afsic/pubs/terms/srb9902.shtml#toc2.
29

United States Environmental Protection Agency. 2008. “Municipal Solid Waste in The United
States 2007 Facts and Figures”. EPA530-R-08-010. United States Environmental
Protection Agency. http://www.epa.gov/osw/nonhaz/municipal/pubs/msw07-rpt.pdf.
Wang, Qingbin, and Junjie Sun. “Consumer Preference and Demand for Organic Food: Evidence
from a Vermont Survey.” In Association: American Agricultural Economics
Association> 2003 Annual meeting, July 27-30, Montreal, Canada.
Wegmans. 2010. “The New Egg Cartons in Our Stores: How the Idea Was Hatched.” April 19.
http://www.wegmans.com/webapp/wcs/stores/servlet/PressReleaseDetailView?langId=1&storeId=10052&catalogId=1&productId=690897.
Weil, Marty. 2008. “Consider Alternative and Renewable Energy Sources.” January 16.
http://www.foodprocessing.com/articles/2008/008.html?page=1.
Wessells, Cathy R., Robert J. Johnston, and Holger Donath. 1999. “Assessing Consumer
Preferences for Ecolabeled Seafood: The Influence of Species, Certifier, and Household
Attributes.” American Journal of Agricultural Economics 81 (5): 1084–1089.

30

CHAPTER 3: FOOD MANUFACTURERS’
SUSTAINABLE PRODUCT LAUNCH STRATEGY: GAME THEORY APPROACH

3.1 Introduction
There are four key reasons why food manufacturers are interested in sustainable
strategies, investing in sustainable practices, and/or launching a sustainable product in the
market. First, launching sustainable products can be a public relations strategy of the firms to
improve their brand image and exposure. There were 13,000 new sustainable food and beverages
launched between 2005-10 (O’Donnell 2008). Consumers are more likely to realize which
brands are green when brands are familiar and have good reputations in general, especially if
brands have green marketing campaigns (O’Donnell 2008). The Accenture and United Nation
Global Compact (Broder 2010) interviewed 766 CEOs around the world in 2010 and found that
manufacturers invest in sustainability for three reasons: brand, trust and reputation (Broder,
2010). Second, manufacturers also invest in sustainable practices to reduce production costs and
increase their competitiveness. The Accenture and UNGC also showed that CEOs realize that
incorporating sustainability practices can be a source of cost efficiency and revenue growth even
during an economic downturn (Broder 2010). Third, global retailers are using their market power
to strongly encourage manufacturers to produce sustainable products. For example, Walmart, the
world largest retailer, has more than 100,000 global suppliers and more than 8,000 stores, is in
the process of creating a “Sustainability Index” (2014 completion date). It will measure the
environmental performance of suppliers to inform its customers about a product’s “life cycle”.
The intention is to improve efficiency by reducing costs and waste (Broder 2010). With
Walmart’s market power, its sustainability program is expected to influence not only other

31

retailers but also manufactures in the near future. Lastly, consumers are demanding that
manufacturers be more environmentally friendly (Oberholtzer, Greene, and Lopez 2006) and
want to know where their food comes from and how it is produced. Eighty-four percent of U.S.
consumers interviewed indicated that they sometimes or regularly purchase sustainable food and
drink, especially those with local and recyclable packaging claims (O’Donnell 2008).
Although sustainability is at the forefront of most food manufacturer and retailing CEOs’
minds, most food manufacturers are reluctant to implement sustainable practices, and develop
and market new sustainable food products. In part, their reluctance appears to be driven by two
primary issues: 1) a lack of technology to produce sustainable products or implement sustainable
practices, and 2) a lack of a systematic decision model that includes extensive sustainable related
variables, especially demand side variables.
This study constructs a model representing food manufacturers’ decision making to
launch a sustainable product. The model that will be developed in this paper is a culmination of
product launch strategies, game theory, and the agribusiness literature. Several works study, both
theoretically and empirically, the innovation strategies and launch strategies of firms in duopoly
markets (Yoon and Lilien 1985; Acs and Audretsch 1987; Dockner and Jørgensen 1988;
Debruyne et al. 2002; Bröring 2007). In particular, numerous industrial organization papers
investigated new product launch strategies using a signaling game approach (Robertson,
Eliashberg, and Rymon 1995), and reaction strategies approach (Debruyne et al. 2002). In the
agribusiness field, several papers used game theory to construct agricultural product launch
strategies (Russo, Cardillo, and Perito 2003; Hitsch 2006; Bröring 2007).

32

The model developed in this paper is later applied to the U.S. egg industry. There are four
main reasons why the egg industry is a good industry to use as an example to understand
manufacturers’ decision making. First, eggs are an important commodity; i.e. it is globally one of
the cheapest sources of protein (Zelman 2009), and has lots of vitamins and minerals (United
States Department of Agriculture 2010b; Jegtvig 2013). Therefore, consumers frequently
purchase eggs. Second, eggs are not a complicated food product, and consumers can easily
understand the marketing message from manufacturers about sustainable attributes. Third,
animal welfare and the new legislation concerning egg production has encouraged many egg
manufacturers to begin marketing sustainable eggs (including free-range and cage-free eggs).
Lastly, the cost and price premiums data for sustainable eggs as well as conventional eggs are
easily accessible making simulation modeling feasible.

The egg industry in the U.S. is an important industry with a market size equal to $ 5.73
billion in 2010 (Patterson 2011). There are two main segments in the egg market, fresh eggs and
egg substitutes. In 2010, fresh eggs had a market share equal to 93.5% or $ 5.36 billion
(Patterson 2011); therefore, the study focuses only on the fresh egg market. Fresh egg categories
include regular eggs and specialty eggs. Examples of specialty eggs are free-range/cage-free
eggs, organic eggs, eggs fortified with Omega-3, low-cholesterol eggs, and vegetarian-fed eggs.
At this time, there is no legal definition for free-range or cage-free eggs in the U.S.
However, according to the Egg Nutrition Center, free-range eggs are from hens that are either
raised outdoors or can access the outside while cage-free eggs are laid by hens that live uncaged
indoors, but do not necessarily have access to the outdoors. Consumers who have animal welfare
concerns prefer and have a higher willingness to pay for a method of animal husbandry that

33

allows hens to roam freely instead of being in cages

16

(Bennett 1998). Sustainable attributes can

include recyclable packaging, free-range, and cage-free attributes. (Satimanon and Weatherspoon
2010) estimated values of sustainable egg attributes by using a hedonic price approach and found
that the recyclable-claim attribute was insignificant. Therefore, in this study, sustainable eggs
include only free-range and cage-free eggs.

The uniqueness of the approach used in this study lies in the ability to analyze the
combination of four factors. First, this study captures the concern about the difference between
17

consumers’ maximum willingness to pay

for the sustainable and the conventional products that

are the constant terms in the inverse demand functions. Second, the model captures the degrees
of substitution between products, which include both the degree of substitution between different
product types (conventional and sustainable products), and the degree of substitution between
manufacturers’ brands. Third, the study incorporates these dimensions within a food supply
chain context, and simulations are generated using egg industry data. Lastly, the model can be
extended to incorporate demand uncertainty in the sense that a firm does not know whether
consumers are willing to pay a premium for a new sustainable food product.

16

Hens are generally raised in a cage system. There are about 5% of eggs in the U.S. (and 10%
around the world) from non-cage housing systems (United Egg Producers 2011).
17
In short, it will be called maximum willingness to pay instead of consumer’s maximum
willingness to pay from now on.
34

3.2 Objective
The objective of this study is to model the manufacturers’ decision making process for
launching a sustainable product to determine: 1) the conditions under which firms should launch
the sustainable product; and 2) the conditions under which the leader firm’s profit is higher than
the follower firm’s profit, that is, the conditions that encourage a firm to be a leader firm. Data
from the egg industry will be used for simulations in order to parameterize and test the model
and to better understand the sustainable egg market.

3.3 Methodology
Several methods are commonly used to determine under which conditions firms will
market a new product and receive higher profits. These methods include empirical analysis, real
options, and game theory when assessing product launch decision making. Several empirical
studies analyzed factors and determinants that resulted in a successful launch of products. For
examples, Benedetto (1999) used mail survey data to analyze factors that encourage the success
of launching a product and suggested that the quality of marketing strategies are the most
important factors. Henard and Szymanski (2001) employed a meta-analysis of the new product
performance literature to find the drivers of new product success. They showed that product
advantage, marketing potential, meeting customer needs, pre-development task proficiencies,
and dedicated resources are important indicators for successful product launches. Debruyne et al.
(2002) used a mail survey to study competitor’s reactions to new launches. They found that
competitors often react by changing prices; moreover, the characteristics of the new products
determine patterns of competitive reactions. Narver, Slater, and MacLachlan (2004) used
empirical data from a variety of businesses, and found that a proactive market orientation (the

35

attempt to understand and to satisfy customers’ latent needs) is important for a firm’s new
product success. These studies relied on quality firm level data that is not available for the egg
industry; hence, these approaches were not used in this study.
Real options, which provides a mechanism in which the farmer’s or firm’s decision to
adopt or launch a new product is determined solely based on the future stream of profits with
respect to uncertainty. This approach is usually used at the product or industry level, given a
level of uncertainty and fixed-investment costs. At the farm level, the adoption of new product
studies focused on high fixed investment costs (Tzouramani and Mattas 2004; Tauer 2006;
Stokes, Rajagopalan, and Stefanou 2008) and assumed continuous time and infinite period
assumptions to project whether the farmers will adopt certain crops or not. The unit of analysis
of those studies was at the farmer level; this paper’s unit of analysis is at the manufacturer level.
The different levels result in different assumptions for output prices; the farmer level having
fixed prices because they are operating under perfect competition; and the manufacturer level
which is assumed to be imperfect competition. Russo, Cardillo, and Perito (2003) combined real
options and game theory to find product launch conditions in the Italian fruit-drink industry.
They focused on uncertainty and the forms of innovation (monopoly of innovation, competitive
innovation, and imitated innovation) and found that the option value of the research and
development investment, the risk of failure, and competitors’ strategies have significant impacts
on product-innovation strategy. Real options studies mainly assume a level of uncertainty,
which relates to high fixed costs and irreversible investment. However, in this study, egg
manufacturers have low fixed costs and no irreversible investments. The manufacturers can use
the same machine whether they want to produce a sustainable egg or conventional egg.
Therefore, real options theory is not selected as a methodology to use in this study. In addition,

36

real options theory also has a tractability issue as the model becomes more complex. The
benefits of using game theory include: the price of the new and old product can be endogenously
determined by the actions of the firms in the market; and the modeling in game theory is more
tractable given the same level of complexity. Therefore, game theory is a more appropriate
method of modeling this industry.
Most of the product launch literature using game theory focuses on high fixed investment
cost industries such as Lint and Pennings (2001), Murto and Keppo (2002), and Lee and
O’Connor (2003). Fulton and Giannakas (2004) used game theory to study genetically modified
(GM) technology adoption to produce and market GM food products. They found that the launch
conditions depend on the preferences of consumers and producers, and life science companies’
objectives. That study used a two-stage sequential game that did not model the stage after the
adoption when other producers can also adopt and market GM products. That is, they did not
consider a follower firm’s decision making to adopt the new technology and launch the GM
product to the market. Fulton and Giannakas’ model informs the development of the model in
this paper which considers under which condition a follower firm should launch a sustainable
product.

The past studies that employed game theory did not focus on the change in profits when
competitors decide to supply the same type of product later. The model has to have a different
structure to allow homogenous preferences of consumers; variability in willingness to pay; and
allow consumers to consume both types of goods simultaneously. In addition, the model needs to
allow for stages when a follower firm also decides to launch a sustainable product as mentioned
before.

37

18

The model must allow for two players , a leader and a follower, in a regional market
and focus on the importance of a relative maximum willingness to pay for a sustainable product,
a relative marginal cost of a sustainable product, and a degree of substitution between two brands
and two types of products. Lastly, the model should be designed for simulation use.

3.4 Model
In this analysis, there are two firms, each firm produces only one type of product, either a
conventional product or a sustainable product in each stage for simplicity. There are four stages
in the analysis as shown in Figure 3.1. Stage 1 is a status quo stage in which both leader (firm i )
and follower (firm j ) produce a conventional product ( c ). Both firms set price as a strategy
simultaneously for their branded conventional product (Bertrand game). This stage will continue
as a repeated game until the leader decides to launch a sustainable product. Stage 2 happens
when the leader firm has the know-how to produce a sustainable product ( s ) and decides to
launch it to make a higher profit. In this stage, both firms use price as the choice variable and set
their prices simultaneously (Bertrand game). This stage is repeated until the follower also
decides to launch a sustainable product. When this occurs, stage 3 is entered and both leader and
follower firms produce sustainable branded products. The leader sets the price first and the
follower sets the price of its product later (Stackelberg price leader game) since the leader first
produced the sustainable product in stage 2. This stage occurs once and the market moves to
stage 4, which is similar to stage 1, except both firms are producing sustainable products. This
series of stages are intended to represent the innovation cycle of a new product launch. Stage 2 to

18

Even though there are two firms in this model, we can apply the model in a general case when
there is a leader and many followers.
38

stage 4 may not exist in reality. They exist only when the leader and/or follower’s launch
conditions are satisfied.

Stage 1

Stage 2

Bertrand game
(Simultaneous
game)
Firm i and firm j
produce
conventional
products.

Stage 3

Bertrand game
(Simultaneous
game)
The leader firm i
produces a
sustainable product
while the follower
firm j produces a
conventional
product.

Stackelberg Price
Leader Game
(Sequential game)
Firm i and firm j
produce
sustainable
products.

Stage 4
Bertrand game
(Simultaneous
game)
Firm i and firm j
produce
sustainable
products.

Figure 3.1 Stages and types of food product innovation games.

This study is based on a simple model for two vertically differentiated products

19

(Dixit

1979; Singh and Vives 1984; Shy 1995) because branded products are similar but they are not
homogeneous. The structure of inverse demand functions for the vertically differentiated
products of firm i in stage 1 is represented by equation (3.1) where pc, i is the price of the
conventional product of firm i ; ac represents the maximum willingness to pay for a
conventional product, which has a value greater than zero; and qc, i is the quantity demand for
the conventional product of firm i .
pc,i  ac  qc,i   1qc, j

19

(3.1)

Vertically differentiated products are defined as products that are different in quality.
39

20

The negative sign of qc, i shows an inverse relationship between price and quantity

(law of

demand). Also in equation (3.1), qc, j is the quantity of the conventional product of firm j . The
negative sign of qc, j shows a negative relationship between price and quantity of the substitute
good. The coefficient  1 is the degree of substitution between the conventional products of firm
i and firm j in stage 1, and has a value between zero and one (Singh and Vives 1984; Poddar

and Saha 2010). If  1 is equal to zero, firm i is a monopoly, that is, the quantity of the
“substitute” good from firm j has no effect on the price of the conventional good from firm i .
On the other hand, if  1 equals one, q c, j is a perfect substitute product of q c , i . This means that
the higher value of  , the higher of the degree of substitution or the products are more similar.
The inverse demand functions for each stage are as follows:
Stage 1: Bertrand Game
pc, j  ac  qc, j   1qc,i .

(3.2)

Stage 2: Bertrand Game

s
ps,i  as  qs,i   2 qc, j , and

(3.3)

c
pc, j  ac  qc, j   2 qs,i .

(3.4)

Stage 3: Stackelberg Price Leader Game
p s,i  a s  q s,i   3 q s, j , and
p s, j  a s  q s, j   3 q s, i .

20

(3.5)
(3.6)

The coefficient of q c , i is normalized to one for simplicity.
40

The structure of the inverse demand functions of firm i and j in every stage is similar
except the degree of substitution in stage 2 is asymmetrical because the types of products of firm
21

i and j are different. The inverse demand functions in stages 1 and 3 are similar . In both

stages, firms produce the same type of products; thus the constant terms in equations (3.1) and
(3.2), ac , are the same; as well as, the constant terms in equations (3.5) and (3.6), as , are the
same. Moreover, the degrees of substitution in stages 1 and 3 are symmetric. In stage 2, the
constant term in the inverse demand functions for firm i and j and the degrees of substitution in
equations (3.3) and (3.4) are different since they produce two different types of branded
products.
The model assumes that 0   s   c   3   1  1.  is always greater than zero
2
2
because the two products are substitute products, and is less than one because the own-price
effect dominates the cross-price effect (Shy 1995).  1 , which is the degree of substitution in
stage 1, represents the brand difference of firm i and j .  s and  c are the degrees of
2
2
substitution in stage 2 from the inverse demand functions for the sustainable product and the
conventional product respectively while technically representing the brand and product
differences inherent in stage 2. We assume here that  s and  c represent only the product
2
2
differences because the brand differences have a small effect relative to product differences, and
we would like to keep the model as simple as possible. Given this,  c is greater than  s
2
2
because a consumer that prefers a sustainable product has a lower degree of substitution for a

21

The inverse demand functions in stage 4 are similar to equations (3.1) and (3.2) except all
subscript c ’s are replaced with s ’s.
41

conventional product while a consumer who prefers a conventional product has a higher degree
of product substitution for a sustainable product. For example, when conventional eggs are on
sale, a consumer who intends to buy free-range eggs is likely to have more difficulty switching
to the discounted conventional eggs. However, if free-range eggs are on sale and have a price
close to conventional eggs, a consumer who normally purchases conventional eggs is likely to
switch to the discounted free-range eggs relatively easily.  3 is the degree of substitution in
stage 3 representing the brand difference and the brand loyalty for the leader firm in a new
market.  3 is lower than  1 because  3 captures both brand difference and first-mover
advantage (in the sense that consumers have brand loyalty to the leader’s brand and launching
the sustainable product first supports the leader’s goodwill and reputation).
The model also assumes that the maximum willingness to pay for the conventional
product ( ac ) is less than the maximum willingness to pay for the sustainable product ( as ). This
implies that consumers on average have a greater willingness to pay for a sustainable product
than a conventional one. In this model, the marginal cost of production of the sustainable product
( c s ) is greater than the marginal cost of the conventional product

22

( cc ). We assume that the

marginal cost is less than the maximum willingness to pay, i.e., ac  cc,i , ac  cc, j , a s  c s, i ,
and a s  c s, j . The industry literature shows that the production cost of a sustainable egg is
higher than the cost of producing a conventional egg by 20 – 66% (“Impacts of Banning Cage
Egg Production in the United States A Report Prepared for United Egg Producers” 2009). The

22

Sustainable practices such as using wind energy and/or solar energy to substitute gas energy
can reduce production cost of firms; however, launching sustainable products to capture higher
price premium usually creates an additional cost. For example, Sumner et al. (2008) showed that
pullets, feed, housing, and labor costs of non-cage system are higher than those costs of cage
system.
42

derivation for the equilibrium profits in stage 1(  c,i1 * and  c, j1 * ), stage 2 (  s,i 2 * and

 c, j 2 * ), and stage 3 (  s,i 3 * and  s, j 3 * ) can be found in appendix B. Appendix C presents
the expansion of the model to include uncertainty on the maximum willingness to pay for a
sustainable product. That is, manufacturers expect that the maximum willingness to pay for a
sustainable product is equal to as with a probability  , and ac with a probability 1-  .
Moreover, the proof of all propositions and standard findings are shown in appendix D.

3.5 Egg Industry Simulations
There are many studies about the retail price premium of sustainable eggs. It ranges from
47.72% to 105.15% over conventional eggs as shown in Table 3.1.

Table 3.1 The price premium of free-range/cage-free eggs and the source of information.
Source
Satimanon and Weatherspoon, 2010
Chang, Lusk and Norwood , 2010
Mintel, 2008
United Egg Producer, --- (based on the
USDA weekly retail shell egg)

Price premium of a free-range or
cage-free egg (%)
47.72
57.00
60.00
105.15

“Impacts of Banning Cage Egg Production in the United States A Report Prepared for United
Egg Producers” (2009) collected data about the additional cost for sustainable egg production at
the farm level from several sources. The additional cost or the marginal cost of the sustainable
egg ranged from 20% to 66% higher than the marginal cost of the conventional egg production.
In addition, Sumner et al. (2008) showed that the difference of production costs between
conventional cage production systems ($ 0.75 per dozen) and non-cage production system
43

($ 1.05 per dozen) were equal to about 41%. However, there is no data for egg production costs
at the manufacturer level. Therefore, we assume that all of the costs at the manufacturer level for
the sustainable and the conventional eggs are the same except the cost for the raw material (eggs
from farmers). Moreover, $ 0.75 is the cost for the conventional egg ( cc ) or it is assumed that
the egg price sold by farmers is the same as the farmers’ cost. This is for simplicity since there is
no interest in estimating the farmers’ margin in this study. The average price for a dozen of grade
A large eggs in 2010 (United States Department of Labor 2011) was $1.66 and was used in the
simulations.
The parameters from the egg industry are used for illustrative purposes to create
simulations and graphs for each proposition/standard finding. Moreover, parameters from the
egg industry and the solution from the profit equations in appendix B are used to simulate and
explain the launching of free-range or cage-free eggs. Mathematica is employed for the
simulations and checking of all propositions and standard findings. All of the simulations in each
proposition show the relationship between the differences in profits and parameters. There are
two scenarios for the simulations; the best case and worst case scenarios. In the best case
scenario, we use the highest price premium value (105.15%) and the lowest additional marginal
cost (20%) for the sustainable egg manufacturer while in the worst case scenario, we use the
lowest price premium value (47.72%) and the highest additional marginal cost (66%).

44

23 c s, i

To comply with the variables in the model for the best (worst) case scenario ,



is equal to 1.2 (1.66), and the proportion of prices in stage 2 



cc, j

in stage 2

p s, i 2 * 

 is equal to 2.05 (1.48).
p c, j 2 * 


The steps and the example of simulation are shown in appendix E.

3.6 Results
The results are divided into two parts to fulfill two objectives: 1) conditions under which
firms launch a sustainable product, and 2) conditions that encourage a firm to be a leader. The
conditions are found under some restrictions; therefore, comparative statics is used to explain the
situation where the restrictions do not hold. Exploration of the conditions under which firms
decide to launch a sustainable product is conducted; then, consideration for the conditions that
encourage a firm to be a leader. The egg industry simulations are presented in each proposition
as well as the standard findings in order to help the reader understand the intuition of the results
and the egg industry modeling.

23

When cc, j equals 1, c s, i (which is 20% higher than cc, j in the best case scenario) equals 1
c s, i
1 .2
+ 0.2 = 1.2. Thus,
=
= 1.2. Similarly, when pc, j 2 * equals 1, ps,i 2 * (which has a
1
cc, j
p s, i 2 *
price premium equal to 105.15% in the best case scenario) equals 1 + 1.05. Therefore,
p c, j 2 *
1 1.05
equals
= 2.05.
1
45

3.6.1 Conditions under which firms launch a sustainable product
Backward induction allows us to determine under which conditions the manufacturers
launch a sustainable product. The leader firm i considers the reaction of firm j before deciding
to launch a sustainable product. That is, the leader’s decision-making depends on whether the
follower will decide to launch a sustainable product later or not. From figure 3.2, the decision of
the leader to launch a sustainable product (move to stage 2) does not depend on only the
comparison of the leader’s profits in stages 1 and 2 (  c,i1 * and  s,i 2 * ), but also on the
comparison of the leader’s profits in stages 1 and 3 (  c,i1 * and  s,i 3 * ).

(

)

Do not launch
sustainable
product
(

)

Launch
sustainable
product
Follower
(Firm )

given

given
Leader
(Firm )

given

Launch
sustainable
product
Do not launch
sustainable
product
(

)

given

Figure 3.2 Backward induction decision tree for manufacturer’s decision
to launch a sustainable product.

46

From figure 3.2, there are three possible outcomes. The first one is that the follower decides to

3
2
launch a sustainable product if  s, j *   c, j * , and the leader also decides to launch a
3
1
sustainable product if  s,i *   c,i * . Payoffs in this case are profits in stage 3. The second
2
3
outcome is if the follower does not decide to launch a sustainable product (  c, j *   s, j * ),
2
1
and the leader decides to launch a sustainable product if  s,i *   c,i * . Payoffs in this case
are profits in stage 2. The last outcome is where no firm decides to launch a sustainable product.

1
3
That is, the leader does not decide to launch a sustainable product (either  c,i *   s,i * or
1

2

 c,i *   s,i * ). In this case, payoffs are equal to profits in stage 1. Next, according to
backward induction we will consider the follower’s decision making and then move to the
leader’s decision making about launching a sustainable product.

3.6.1.1 Follower’s decision making for launching a sustainable product
Consider the follower firm’s decision making by comparing the profits when launching a

3
2
sustainable product (  s, j * ) and when still producing a conventional product (  c, j * ). The
follower will decide to launch a sustainable product when there are higher profits from launching

3
2
a sustainable product (  s, j * >  c, j * ).

47

Given that the leader has already launched a sustainable product, the follower has to
decide whether it is more profitable to produce a “me too” product or not. This leads to the first
proposition (the derivations can be found in appendix D).

c s, j a s
Proposition 1: When  c   s   3 , c s,i  c s, j , and
, the follower will decide to

2
2
cc, j ac
launch a sustainable product and move to stage 3 if:
a. 0   s  0.75 , or
2
s 2
(  2   2) 2
 
c s, j
 
s
b. 0.75   2  1 and 1 
.

cc, j 0.25[( s  2)(( s  3)( s  4) s  8) s  16]
2
2
2
2
2

Explanation: Given the restrictions that the: 1) degrees of substitution in stages 2 and 3 are the
same (  c   s   3 ); 2) costs of both firms to produce sustainable products are the same
2
2
( c s,i  c s, j ); and 3) relative cost of a sustainable over conventional product

24

is equal to its

 c s, j a s 
 , then the follower decides to launch a
relative maximum willingness to pay 

 cc , j a c 


sustainable product. The intuition for proposition 1a is that the follower considers the consumers’
perception as a key factor for its sustainable product launch decision. That is, when consumers
realize that conventional and sustainable products are different ( 0   s  0.75 ), and have a
2
willingness to pay more for the sustainable product; a sustainable product market will mature

24

From now on, we will call the term

c s, j

as a relative cost of a sustainable product and

cc, j
as a relative maximum willingness to pay for a sustainable product.
48

as
ac

enough such that the follower launches a sustainable product to take market share from the
leader. The second factor the follower considers is the relative cost of a sustainable product

 c s, j

 cc, j



 . Proposition 1b shows that even if consumers do not realize the difference between two



types of products ( 0.75   s  1 ), the follower still decides to launch a sustainable product if the
2
marginal costs of the sustainable and conventional products are close enough, or when a relative
cost of a sustainable product is low enough



s 2


(  2   2) 2
 
 
1  cs, j 
.

s  2)(( s  3)( s  4) s  8) s  16] 
cc, j 0.25[(


2
2
2
2
2


Figure 3.3 presents simulation results for profit comparisons between stages 2 and 3 for
the follower firm assuming

25

c
s
as = 5, and varying the degrees of substitution (  2   2   3 ) for

 c s, j
three levels of relative costs of a sustainable product 
 cc, j



 . The relative cost levels were 1.2



(the best case scenario), and 1.66 (the worst case scenario). Figure 3.3 shows that over these
ranges, the follower firm will launch a sustainable product at all levels of relative costs as long as
the degree of substitution is not too high (below 0.97 for

25

c s, j
cc, j

= 1.2 (the best case scenario), and

as = 5 approximately corresponds with the demand for eggs/person/10 weeks. That is, when
as = 5, the minimum value of a total quantity in stage 2 is equal to 2.68 dozen eggs. According
to “Per Capita Consumption of Eggs (shell) in the U.S. 2009 | Statistic” (2011), egg consumption
per capita per year was 172.9 eggs. Therefore, on average a person consumes at least 2.68 eggs
per 68 days or about 10 weeks. as is stable across a wide range of values in terms of the shapes
of the curves but the scale changes in magnitude.
49

below 0.93 for

cost decreases (

c s, j
cc, j
c s, j
c c, j

= 1.66 (the worst case scenario). This also shows that as the new technology

also decreases), the higher range of  s allows the follower to get higher
2

profits when launching a sustainable product. To correspond with proposition 1, figure 3.3 shows
that for all values of

c s, j
c c, j

, the follower receives higher profit when launching a sustainable

product if  s is lower than 0.75 or consumers realize that the two types of products are different
2

 c s, j
(proposition 1a). Moreover, the lower relative cost of a sustainable product 
 cc, j



 , the higher



the chance that the follower will decide to launch a sustainable product in order to earn higher
profits (proposition 1b).

50

3

2

 s, j *  c, j *

phij3 phij2
2
2.0

1.5

──

1
1.0

c s, j

= 1.2
cz , j1.2
c
c

s
─ ─ z, j1.66
= 1.66
c c, j

0.5
0.2
0.2

0.4
0.4

0.6
0.6

0.8
0.8

1.0
1.0

s
g2c
2

0.75
Figure 3.3 Profit simulation between stages 2 and 3 for the follower firm
varying the degrees of substitution (  c   s   3 )
2
2
for three levels of relative costs of a sustainable product.

Next, we relax the first assumption such that  c   s   3 which is more realistic
2
2
meaning; consumers realize the difference in product types is more than the difference in brands.
For example, consumers feel that a cage-free egg and a conventional egg are more different than
a brand A egg and a brand B egg. Figure 3.4 shows graphs of the difference in follower’s profits

2
3
c s, j
in stages 2 and 3 (  s, j * -  c, j * ) when  3 is varied by setting
= 1.66, and as = 5.
c c, j
Figure 3.4 shows that given  c   s   3 , the shapes of the graphs of the difference in profits
2
2

3
2
(  s, j * -  c, j * ) are the same when we vary  3 . Therefore, the intuition of proposition 1 is
valid even when we relax the assumption  c   s   3 to be  c   s   3 .
2
2
2
2
51

3

2

 s, j *  c *
phij3 phij2 , j
33

── g3 3 0.20.2
 =

─ ─  3 = 0.4
- - - g3 30.40.6
- =

22
11

0.2
0.2

0.4

0.6

0.8
0.8

s
2
g2c
1.0
1.0

……..  30.60.8
g3 =

g3 0.8

1

Figure 3.4 Profit simulation

26

between stages 2 and 3 for the follower firm
varying the degrees of substitution in stage 2 (  c   s )
2
2
for four levels of degree of substitution (0.2, 0.4, 0.6, and 0.8) in stage 3.

Without any restrictions, comparative statics shows the difference in profits will increase
 ( 3 *  2 *)



s, j
c, j
if the cost to produce a sustainable product declines 
 0  . That complies with
cs, j





proposition 1b. When the follower has a technology to reduce the cost of a sustainable product,
3
2
the follower decides to launch a sustainable product to get a higher profit (  s, j *  c, j * > 0).

Next, profitability when there are sustainable eggs in the market already is explored. The
values, which are used in the simulation in the best (worst) case are: 1)

26

c s, i
cc, j

= 1.20 (1.66), 2)

Note,  3 is greater than  s ; therefore consider only the part where  3 is higher than  s .
2
2
52

p s, i 2 *
= 2.05 (1.48), 3) cc, j = 0.75 per dozen, 4)  c   s   3 , and 5) pc, j 2 * = 1.21 per
2
2
2*
p c, j
27

dozen . The variables in the model that cannot be observed are the levels of maximum
willingness to pay ( ac and as ) and the degree of substitution (  ). However, simulation can be
used to understand the unobserved variables. Figure 3.5 shows the relationship between

as
and
ac

 . That is, when consumers understand that two types of eggs are different (  approaches zero),
consumers would like to pay more for a sustainable product or they have a higher relative
maximum willingness to pay for a sustainable egg.

as ac
1.40
1.40
1.35
1.30
1.30
1.25
1.20
1.20
1.15
g2c
0.2
0.2

0.4
0.4

0.6

0.6

0.8
0.8

1.0
1.0

Figure 3.5 Simulation of the relative maximum willingness to pay for a sustainable product
 as 
  varying degrees of substitution (  =  c   s   3 ).
a 
2
2
 c

27

This is the wholesale price per dozen of conventional eggs in 2008 (United States Department
of Agriculture 2010a).
53

From the simulation of both the best and worst cases, the follower’s equilibrium price of a
sustainable egg in stage 3 is higher than the follower’s equilibrium price of a conventional egg in

2
3
3
2
stage 2 ( p s, j * > pc, j * ), and so is the follower’s equilibrium quantity ( q s, j * > qc, j * ).
Note, even in the worst case, the follower’s profit when launching the sustainable eggs is higher

2
3
than the follower’s profit when producing the conventional eggs (  s, j * >  c, j * );
therefore, the follower decides to launch the sustainable egg product.

3.6.1.2 Leader’s decision making for launching a sustainable product
The leader’s decision making to launch a sustainable product is conditional on the
follower’s decision making. That is, if the follower decides to launch a sustainable product, the
leader will compare its profits in stages 1 and 3 and decide to launch a sustainable product when
getting a higher profit (  s,i 3*   c,i1 * ). However, given that the follower does not decide to
launch a sustainable product, the leader has to compare its profits in stages 1 and 2 instead, and
decide to launch a sustainable product when in stage 2 profits are higher than in stage 1
(  s,i 2 *   c,i1 * ).

When the leader assumes that the follower will launch a sustainable product after the
leader launches is modeled in the next two propositions.

54

Proposition 2: When either (a)

c s, i
cc,i



c s, j

a
 s , and 1   3 or (b)
cc, j ac

c s, i
cc,i



c s, j

a
 s,
cc, j ac

cc,i  cc, j , and c s,i  c s, j , the leader will decide to launch a sustainable product and move to

stage 2 in order to get a higher profit (  s,i 3*   c,i1 * ).

Explanation: Assuming that the follower will launch a sustainable product later, proposition 2
expresses the conditions under which the leader launches a sustainable product by comparing the
leader’s profits in stages 1 and 3. There are two sets of conditions under which the leader will
decide to launch a sustainable product to get a higher profit (  s,i 3*   c,i1 * ). Those are, (a)
relative costs of a sustainable product of both firms are the same and equal to the relative
c
c s, j a s 
 and the degree of
maximum willingness to pay for a sustainable product  s,i 

 cc,i cc, j ac 



substitution in stages 1 and 3 are the same ( 1   3 ), and (b)

c s, i
cc,i



c s, j

a
 s and firms have
cc, j ac

the same costs to produce the same types of products ( cc,i  cc, j and c s,i  c s, j ).
The main idea of this proposition is that the leader firm decides to launch a sustainable
product when its cost in relative terms equals the consumer’s relative maximum willingness to
pay. Decomposing the cost of a sustainable product ( c s , i ) and the consumer’s maximum
willingness to pay for a sustainable product shows c s , i = cc, i + c , and as = ac + a where c
is an additional cost to produce a sustainable product and a is the additional maximum
willingness to pay for a sustainable product.

c s,i cc,i
a
or

 s can be rearranged as
ac
as
ac
cc,i
c s, i

55

cc,i  c c s,i cc,i

. Intuitively, the leader decides to launch a sustainable product when the

as
ac  a
ac

additional cost to produce a sustainable product can be covered by the additional maximum
willingness to pay for a sustainable product. That is, c has to be equal to

cc,i * a
ac

, or

c cc,i
< 1. Therefore, an egg manufacturer will be profitable when producing a sustainable

a ac
egg if the additional part costs less than an additional maximum willingness to pay such that

c cc,i
, holds.

a ac

56

Price

c

a
d
b
Quantity

0

Figure 3.6 Leader’s profits in stages 1 and 3
when the relative cost of a sustainable product of both firms are the same and equal to the
 c s, i c s, j a s 
 and degrees
relative maximum willingness to pay for a sustainable product 


 cc,i cc, j ac 


of substitution in stages 1 and 3 are the same ( 1   3 ). (Condition set (a) holds.)

3
Figure 3.6 illustrates proposition 2 with condition set (a) where D1,i and Ds,i are
c

demands for a conventional product from firm i in stage 1 and a sustainable product from firm i
3
in stage 3, respectively. D1,i has the same slope as Ds,i , which is equal to -1, but the y-intercept
c

will be lower because the maximum willingness to pay for a sustainable product is higher than
the one paid for a conventional product ( as > ac ). The equilibrium prices and quantities

*

*

*

1
1
3
3 *
( pc,i , qc,i , p s,i , and q s,i ) are arbitrarily determined in figure 3.6 since they are pre57

determined by solving the equilibrium in the oligopoly market (appendix B). Firm i ’s profits in
c s, i a s
1 *
3 *
stages 1 and 3 are cc, i ba pc,i , and c s , i dc p s,i , respectively. Graphically,
shows

cc,i ac

*

1
that the leader’s profit when launching a sustainable product is higher ( cc, i ba pc,i <
*

3
c s , i dc p s,i ). The graph and explanation of proposition 2 with condition set (b) is similar to
figure 3.6. The difference is that the costs for both firms are the same for conventional and
sustainable products ( cc,i  cc, j and c s,i  c s, j ) in this case, but the degrees of substitution in
stages 1 and 3 are not necessarily the same ( 1   3 ). Hence, the y-intercepts of the demands in
stage 3 are different, which results in differences in the equilibrium price and quantity.

The egg industry simulation results when comparing the leader’s profits in stages 1 and 3,
and assuming the additional cost of a sustainable product equals 66% (the worst cost case) are
shown in figures 3.7.1 and 3.7.2. Figure 3.7.1 shows the simulation results when 1   3 , but
neither cc,i  cc, j nor c s,i  c s, j (condition set (a)) while figure 3.7.2 shows the simulation
results when cc,i  cc, j and c s,i  c s, j , but 1   3 (condition set (b)). Figures 3.7.1, and
3.7.2 show when consumers realize products from the two brands are more different, hence it is
more profitable for the leader to launch a sustainable product (lower value of  , higher value of

3

1

 s,i *  c,i * ). For each value of  , figure 3.7.1 shows that the levels of the leader’s
3
1
difference in profits in stages 3 and 1 (  s,i *  c,i * ) are higher when the leader has a cost

58

advantage in the production of a sustainable product

28

( c s, i  c s, j ) than when it does not.

Moreover, without any restrictions, the comparative static results also show that the leader will
get a higher profit from launching a sustainable product when the leader has a lower cost of a

3
sustainable product. Lastly, both figures 3.7.1 and 3.7.2 also show that  s,i * is always greater
1
3
1
than  c,i * (  s,i *  c,i * > 0), and the larger value of the maximum willingness to pay for
3
1
a sustainable product ( as ) results in a larger value of profit difference (  s,i *  c,i * ). That
is, when consumers are willing to pay more for a sustainable product, the leader will get a higher
profit when launching a sustainable product.

28

From figure 3.7.1, the dashed lines illustrate the differences in profits of firm i in stages 1 and
3
1
3 (  s,i *  c,i * ) when firm i has a cost disadvantage in producing a sustainable product
3
1
( c s, i  c s, j ) while the solid lines illustrate  s,i *  c,i * when firm i has a cost advantage
in producing a sustainable product ( c s,i  c s, j ). Hence, the solid lines are above the dashed
lines for the same  3 .
59

3

c s, i  c s, j
….  3 = 0.2

1

 s,i 6*  c,i *

---  3 = 0.4
- -  3 =g3 0.2
0.6
- -  3 =g3 0.4
0.8

5

4

c s, i  c sg3 0.6
,j

─
─
─
─

3
3

2
1
1

3

4

5

6

7

8

0.2
 3 = g3 0.8
g3 0.2
0.4
 3 = g3 0.4
0.6
 3 = g3 0.6
0.8
 3 = g3 0.8

as
as

Figure 3.7.1 Profit simulation between stages 1 and 3
3
1
for the leader (  s,i *  c,i * ) when condition set (a) holds
and varying the maximum willingness to pay for a sustainable product ( as )

for four levels of degrees of substitution (0.2, 0.4, 0.6, and 0.8) in stage 3 (  3 ).

3
1
phii1
phii3*  c,i *
s, i
8
6

─

4

g1 0.9 &
 1 = 0.4 and  3 =g3 0.8
0.3

- -  1 = 0.9g1 0.4 3 =g3 0.3
and  & 0.8

2
4
4

6
6

8
8

10
10

a
as s

3
1
Figure 3.7.2 Profit simulation between stages 1 and 3 for the leader (  s,i *  c,i * )
when condition set (b) holds and varying the maximum willingness to pay for a sustainable
product ( as ) for two sets of degrees of substitution in stages 1 and 3,

 1 = 0.9 and  3 = 0.8, and  1 = 0.4 and  3 = 0.3.
60

Next, we will relax the assumption that

c s, i
cc,i



c s, j

a
 s to consider whether the leader should
cc, j ac

launch a sustainable product or not when relative costs of a sustainable product of both firms are
not equal to the relative maximum willingness to pay for a sustainable product
 c s, i c s, j a s 

.


 c c, i c c, j a c 



Proposition 3: When 1   3 , cc,i  cc, j , and c s,i  c s, j , the leader will decide to launch a
sustainable product and move to stage 2 if:
a.

b.

c s, i

a
 s , or
cc,i ac
c s, i

a
 s , and a s  c s,i 1  A or a s  c s,i  c s,i A where
cc,i ac

2
 1 1
2  4) 2 ( 2  2) 1  1 

( 1
8( 12  2)  
1
 y z

 y z
c s, i

 


, and
A2 2 
 0, y 
cc,i

2  2)  2
8( 12  2) 

 ( 2  4) 2 
z
 ( 2  4) 2  8( 1

 1

2
1


z


z2



a
z s .
ac

Explanation: The intuition for proposition 3a is that it is not only that the relative cost of a
sustainable product has to be equal to its relative maximum willingness to pay to make a leader
get higher profit when launching a sustainable product, but also the relative cost of a sustainable
product is less than its relative maximum willingness to pay. This is not surprising because the
61

lower relative cost of a sustainable product and the higher its relative maximum willingness to
pay result in a higher profit for producing a sustainable product. Therefore, a leader is more
willing to launch a sustainable product to get a higher profit when the relative cost of a
sustainable product is less than the relative maximum willingness to pay for a sustainable
 c s, i a s 
.
product 

 c c, i a c 



Price
3
a s ' 1q s,' j

a '
as '  as  s 
ac

3
a s   1q s, j

ac   1q1, j
c

3' *
p s,i
3 *
ps,i
*
p1,i
c
c s,i
cc , i

0

D1,i
c

3
Ds,i

a s c s ,i

ac cc,i

3'
Ds,i

e
c

a
d

f

b

*
* 3' *
q1,i q 3 q s,i
c
s,i

Quantity

Figure 3.8 Leader’s profits in stages 1 and 3 when a relative cost of a sustainable product is less
 c s, i a s 
,
than or equal to the relative maximum willingness to pay for a sustainable product 

 c c, i a c 


degrees of substitution in stages 1 and 3 are the same ( 1   3 ), and firms have the same costs to
produce the same types of products ( cc,i  cc, j and c s,i  c s, j ).

62

Figure 3.8 is an expansion of figure 3.6 where the demand in stage 3 when increasing the
maximum willingness to pay for a sustainable product ( a s ' ) has been added. The maximum
willingness to pay for a sustainable product increases until the relative cost of a sustainable
 c s, i a s ' 
 . When the demand in
product is less than its relative maximum willingness to pay 

 cc,i ac 



3' *
3'
stage 3 is changed to be Ds,i , the leader’s profit in stage 3 increases to be c s , i fe p s,i , which is
3' *
3 *
3
greater than the leader’s profit when the demand is Ds,i ( c s , i fe p s,i > c s , i dc p s,i ) and

certainly greater than the leader’s profit in stage 1. Therefore, when a relative cost of a
sustainable product is equal to or less than its relative maximum willingness to pay
 c s, i a s 

 , the leader decides to launch a sustainable product to get a higher profit.

 c c, i a c 



The intuition for proposition 3b is that even if the relative cost of a sustainable product is
higher than its relative maximum willingness to pay, the leader still decides to launch a
sustainable product to get a higher profit if the absolute term of maximum willingness to pay for
a sustainable product is sufficiently greater than its cost ( a s  c s,i 1  A ). That is, proposition
3a focuses on the concept of relative terms while proposition 3b focuses on the concept of
absolute terms.
The constant term A is a function of the degree of substitution (  ), the relative cost of a
 c s, i
sustainable product 
 c c, i



 , and the relative maximum willingness to pay for a sustainable



63

a 
product  s  . When the degree of substitution increases, the value of term A decreases
a 
 c

 A


   0  , which makes proposition 3b easier to satisfy; consequently, the leader firm is more



likely to launch a sustainable product. Intuitively, the more homogeneous the products, the
greater the probability that the leader firm innovates to make a differentiated product to capture
higher sale and profit. When the relative cost of a sustainable product increases, the value of term



A
 0  , which causes proposition 3b less likely to be satisfied. That
A also increases 
  cs,i / cc,i








is, when the leader has a higher cost to produce the sustainable product compared to the cost of
the conventional product, the leader is less likely to launch the sustainable product. Lastly, when
the relative maximum willingness to pay for a sustainable product increases, the value of term A


A
 0  , which results in proposition 3b becoming easier to satisfy. That is, it
decreases 
  a / a 

s c



is more attractive for the leader to launch the sustainable product when consumers would like to
pay more for the sustainable product compared to the amount paid for the conventional product.
To test that proposition 3b is a rare case or not, we use data from the egg industry: cc, i =
0.745, and we assume as = 5 and 1   3 = 0.9 (case 1 in table 3.2). Given the maximum
willingness to pay for a sustainable product is double the maximum willingness to pay for a
a

conventional product  s  2  , even the cost of a sustainable product is quadruple the
a

 c

 c s, i

conventional cost 
 4.43 , the leader profits are higher when launching a sustainable egg.
 cc,i




Therefore, proposition 3b is also important to the leader’s decision making and is not a rare case.
64

More simulations are shown in table 3.2 when varying the relative maximum willingness to pay
a 
for a sustainable product  s  . In addition, from table 3.2, the degrees of substitution ( 1   3 )
a 
 c

are changed in case 2, and the maximum willingness to pay for a sustainable product ( as ) are
changed in case 3.

Table 3.2 The egg industry simulations to test whether proposition 3b is a rare case or not when
varying the degree of substitution ( 1   3 ), maximum willingness to pay for a sustainable
a 
product ( as ), and the relative maximum willingness to pay for a sustainable product  s  .
a 
 c
c s, i
as
The highest value of
that the leader still gets a
1   3
as
Case
cc,i
ac
higher profit when launching a sustainable product
1
0.9
5
1.11
1.84
0.9
5
2.00
4.43
0.9
5
5.00
6.38
2
0.5
5
1.11
1.68
0.5
5
2.00
4.36
0.5
5
5.00
6.37
3
0.9
10
1.11
2.70
0.9
10
2.00
7.89
0.9
10
5.00
11.79

Figure 3.9 shows the simulation results of the egg industry when the degree of
substitution in stages 1 and 3 are the same and equal to 0.8 ( 1   3 = 0.8) for the relationship

3
1
between the leader’s difference in profits (  s,i *  c,i * ) and the maximum willingness to
pay for a sustainable product ( as ). When the relative cost of a sustainable product is less than or
 c s, i a s 
 , the
equal to the relative maximum willingness to pay for a sustainable product 

 c c, i a c 



65

leader’s profit in stage 3 is always higher than the leader’s profit in stage 1 (all dashed lines have

3
1
c s, i a s
positive values of  s,i *  c,i * ), which is consistent with proposition 3a. When
is

cc,i ac
high enough, i.e.

c s, i
cc,i

= 1.66 (the worst case scenario) and

as
= 1.1 as in figure 3.9, the leader
ac

does not decide to launch a sustainable product when the maximum willingness to pay for a
sustainable product ( as ) is a low value (less than 4.70 in Figure 3.9) since the leader gets a lower
profit when launching a sustainable product. Moreover, the higher value of the maximum
willingness to pay for a sustainable product ( a s  ), the greater the probability that the leader

3
1
will get a higher profit when launching a sustainable product (  s,i *  c,i *  ), which is
consistent with proposition 3b.

66

3

1

 s,i *  c,i *

c s , i / cc , i  a s / a c

6
6

… c s,i / cc,i = 1.2, as / ac = 3

--- c s,i / cc,i

5

= 1.66, as / ac = 3

c s,i / cc,i  a s / ac

44

--

c s, i / cc, i = as / ac = 1.2
─ ─ c s, i / cc, i = as / ac = 1.66

3

c s,i / cc,i  a s / ac

22

─ c s, i / cc, i = 1.2, as / ac = 1.1
1

─ c s , i / cc, i

4
4

6
6

8
8

10
10

= 1.66, as / ac = 1.1

ass
a

4.70

3
1
Figure 3.9 Profit simulation between stages 1 and 3 for the leader (  s,i *  c,i * )
varying the maximum willingness to pay for a sustainable product ( as )
 c s, i
for variety sets of the relative costs of a sustainable product 
 c c, i






a 
and the relative maximum willingness to pay for a sustainable product  s  .
a 
 c

Proposition 4 explores when the leader assumes that the follower will not launch a sustainable
product later. When the leader expects that the follower will not launch a sustainable product
later, the leader will compare its profits in stages 1 and 2 for the decision making to launch a
sustainable product.

67

Proposition 4: When  1   c   s , and cc,i  cc, j , the leader will decide to launch a
2
2
sustainable product and move to stage 2 if:
a.

b.

c s, i

a
 s , or
cc,i ac
c s, i

a
 s and a s  c s,i  ac  cc,i .
cc,i ac

Explanation: Proposition 4 shows that the leader should launch a sustainable product when the
relative cost of a sustainable product is less than or equal to its relative maximum willingness to
 c s, i a s 
 , or the additional cost to produce a sustainable product can be covered by the
pay 

 c c, i a c 



additional maximum willingness to pay for a sustainable product. If the concept of relative does
 c s, i a s
not hold in proposition 4a 

 c c, i a c



 , then the leader still decides to launch a sustainable



product to obtain a higher profit when the difference between the maximum willingness to pay
for a sustainable product and its cost is sufficiently high ( a s  c s,i  ac  cc,i ). That is, if the
relative term concept does not hold, the leader still decides to launch a sustainable product if the
absolute term concept holds. Hence, proposition 4 has the similar intuition as proposition 3.
Proposition 4b is not a rare case as shown in table 3.3. For example, when the maximum
willingness to pay for a conventional product ( as ) equals 5 (case 1 in table 3.3), even when the
maximum willingness to pay for a sustainable product is double the conventional product
 as


 2  , the relative cost of a sustainable product can be as high as 4.36
a

 c


68

 c s,i


 4.36  . More
 cc,i




simulations when varying the relative maximum willingness to pay for a sustainable product
 as 
  are shown in table 3.3. The maximum willingness to pay for a sustainable product ranged
a 
 c

from 5 to10 in case 2 to test the robustness of the model.

Table 3.3 The egg industry simulation to test whether proposition 4b is a rare case or not when
a 
varying the value of the relative maximum willingness to pay for a sustainable product  s  ,
a 
 c
and the maximum willingness to pay for a sustainable product ( as ).
Case
1

2

as
5
5
5
10
10
10

as
ac

The highest value of

1.11
2.00
10.00
1.11
2.00
10.00

c s, i
cc,i

that the leader still gets a higher

profit when launching a sustainable product
1.67
4.36
6.37
2.34
7.71
11.74

Figure 3.10 shows the relationship between the leader’s difference in profits in stages 1

2
1
and 2 (  s,i *  c,i * ) and the difference between the maximum willingness to pay for a
sustainable product and the cost of a sustainable product ( a s  c s,i ), which is consistent with
 c s, i a s 
 , the leader’s profit in stage 2 is always higher than the
proposition 4. That is, when 

 c c, i a c 



leader’s profit in stage 1 (all dashed lines have positive values of  s,i 2 *  c,i1 * ). Moreover,
when assuming the relative cost of a sustainable product is higher than its relative maximum
 c s, i a s
willingness to pay 

 c c, i a c



 , the higher value of a s  c s,i , the higher probability that



69

2
1
c
 s,i * is greater than  c,i * . For example, when determining cc, j = 0.75, s,i = 1.66 (the
c
c,i

worst case scenario),

as
=1.1, and degrees of substitution in stages 1 and 2 are the same and
ac

equal to 0.9 ( 1   c   s = 0.9), the leader’s difference in profits in stages 1 and 2
2
2
2
1
(  s,i *  c,i * ) can be negative or positive. It will be a positive value when the difference

between the maximum willingness to pay for a sustainable product and the cost of a sustainable
product ( a s  c s,i ) is sufficiently high ( a s  c s,i > 4.17), which is consistent with proposition
4b.

70

c s , i / cc , i  a s / a c

 s,i 2 *  c,i1 *

… c s,i / cc,i = 1.2, as / ac = 3

--- c s,i / cc,i

25

= 1.66, as / ac = 3

c s,i / cc,i  a s / ac

20
20

--

c s, i / cc, i = as / ac = 1.2
─ ─ c s, i / cc, i = as / ac = 1.66

15

c s,i / cc,i  a s / ac

10
10

─ c s, i / cc, i = 1.2, as / ac = 1.1
─ c s, i / cc, i = 1.66, as / ac = 1.1

5

2
2

3
3

4
4

5
5

6
6

7
7

8
8

as
asm  cs,i

1
1
1
1 4.17 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
Figure 3.10 Profit simulation between stages 1 and 2 for the leader (  s,i *  c,i * )
1
1
1
1
1
1
1
2
2
2
2
2
2
varying the difference between the maximum2
willingness to pay for a sustainable product and
 c s, i 

the cost of a sustainable product ( a s  c s,i ) for a variety of relative cost 
 c c, i 


a 
and relative maximum willingness to pay  s  .
a 
 c

Consider the egg industry situation, proposition 1, and the simulation results from section
3.6.1.1, the follower’s launch decision making is known: the follower decided to launch after the
leader has launched. Hence, in this step, the leader makes a launch decision given the

2
3
information that the follower decided to launch a sustainable egg (  s, j * >  c, j * ); thus, the
leader compares its profits in stages 1 and 3. Proposition 3 considers the leader’s decision

71

29

making . The first step to do that is to check whether proposition 3a holds. Figures 3.11.1 and
3.11.2 show values of

c s, i
cc,i

when  is varied (Note,

and

as
for the worst case and the best case from the simulations
ac

as
, and  are not directly observed). Figure 3.11.1 shows that
ac

c
a
proposition 3a does not hold for the worst case for all values of  since s is greater than s
ac
cc
while figure 3.11.2 shows that proposition 3a holds for the best case when  is low (  is less
than 0.88).

29

The reasons are: 1) proposition 3 compares the leader’s profits in stages 1 and 3, and 2)
proposition 3 does not require the equality of the relative cost of a sustainable product and the
 c s, i a s 
 , which is flexible
relative maximum willingness to pay for a sustainable product 

 c c, i a c 


and appropriate to use in our analysis.
72

cs a s
,
cc ac
csi cci,as ac

cs
cc

1.6
1.6
1.5
1.4
1.4
1.3
1.2
1.2

as
ac

1.1


g
0.2

0.2

0.4

0.6

0.8

0.8

0.6

0.4

1.0

1.0

a
c
Figure 3.11.1 The relationship between s or s and  for the worst case scenario.
cc
ac

csi cci,as ac
2.2
2.2

The range that
proposition 3a does
not hold

The range that proposition
3a holds

2.0
1.8
1.8
1.6
1.4
1.4
1.2

g
0.2
0.2

0.4
0.4

0.6
0.6

0.8
0.8

1.0
1.0

0.88

a
c
Figure 3.11.2 The relationship between s or s and  for the best case scenario.
cc
ac
73

Since proposition 3a does not hold in general, we check whether proposition 3b holds.
Figures 3.12.1 and 3.12.2 show that proposition 3b holds for both the worst and the best case
since the values of a s  c s, i are greater than c s,i A for all  . Consequently, the leader’s profit

3
1
in stage 3 (  s,i * ) is higher than the leader’s profit in stage 1 (  c,i ); therefore, the leader
30

decides to launch the sustainable egg into the market .

a s  c s ,i , c s ,i * A

as csi,A csi
66

a s  c s, i

5

c s ,i * A

4
4
3
2
2


g2c
0.2
0.2

0.4

0.6
0.6

0.4

0.8
0.8

1.0
1.0

Figure 3.12.1 The relationship between a s  c s, i or c s,i * A and 
for the worst case scenario.
30

The leader’s equilibrium price of a sustainable egg in stage 3 is greater than the leader’s
3
1
equilibrium price of a conventional product in stage 1 ( p s,i *  pc,i * ) while the leader’s
equilibrium quantity in stage 3 is not necessarily greater than the leader’s equilibrium quantity in
stage 1. That is, the equilibrium quantity in stage 3 is greater than the equilibrium quantity in
3
1
q s,i *  qc,i * ) when consumers realize the difference between conventional and
stage 1 (
sustainable eggs enough or when  is less than 0.57 for the worst case and 0.86 for the best case
scenario.
74

a s  c s, i , c s, i * A

as csi,A csi
12
12
10

a s  c s, i

8
8
6
4
4
2

c s,i * A

g2c
0.2
0.2

0.4
0.4

0.6
0.6

0.8
0.8

1.0
1.0

Figure 3.12.2 The relationship between a s  c s, i or c s,i * A and 
for the best case scenario.

In sum, according to the prices and costs of the conventional and the sustainable eggs
from the literature, both the leader and the follower egg manufacturers should launch a
sustainable egg product to get higher profits. This corresponds with what we observe, there is
an expansion of the sustainable egg market during the past few years, even store brands have
launched cage-free and/or free-range eggs in many regions of the U.S (Patterson 2011).

After determining the conditions that firms should launch sustainable products to get a
higher profit, the conditions that encourage a firm to be a leader firm are explored.

75

3.6.2 Conditions that encourage a firm to be a leader
This part starts with the standard findings and propositions from comparing profits of
both firms in each stage. Simulations of the egg industry are used to explain the intuition and
make the standard findings and propositions more understandable. The study concludes with the
conditions that encourage a firm to be a leader to launch a sustainable product to the market.

3.6.2.1 Comparison of profits in stage 1
Standard Finding 1: In stage 1, marginal costs

31

determine which firm has a higher profit.

Moreover, the difference in marginal costs and the degree of substitution (  1 ) determines the
amount of difference in profits. The larger the difference in marginal costs, and/or the larger
degree of substitution lead(s) to a larger difference in profits.

Explanation: Since  1 is symmetrical for both inverse demand functions in stage 1
( pc, i  ac  qc, i   1qc, j , and pc, j  ac  qc, j   1qc,i ), both firms face the same demand
function. Therefore, the firm who has a lower marginal cost will achieve a higher profit.
Consumers can easily switch to the cheaper product when the degree of substitution (  1 ) is high
meaning the two products are similar.
The egg industry sheds light on the relationships between the two firms in stage 1. The
first simulation assumes

31

32

that ac = 3.01 when

c c, i
cc, j

equals 1, the marginal cost for both

In our analysis, the marginal costs are the same as the variable costs.

32

We assume ac = 3.01 to be consistent with the demand for eggs/person/10 weeks. The details
are shown in footnote 25.
76

conventional goods are the same, and when

c c, i
cc, j

is less (more) than 1, cc,i is less (more) than

cc, j , figures 3.13.1 and 3.13.2 are obtained. Figure 3.13.1 shows that when the marginal cost of

firm i is lower (higher) than the marginal cost of firm j or

c c, i
cc, j

<1(

c c, i
cc, j

> 1), the profit of

1
1
1
1
firm i is greater (lower) than the profit of firm j or  c,i *  c, j * > 0 (  c,i *  c, j * < 0).
From figure 3.13.2, when the degree of substitution (  1 ) changes, the sign of the difference in

1
1
profits of the two firms (  c,i *  c, j * ) does not change; however, when  1 increases, the
1
1
absolute value of  c,i *  c, j * will increase.

─
-…

profiti1 profitj1
0.2
0.2

= 0.8
= 0.5
= 0.2

0.1
0.1

k
0.9
0.9

1.0

1.0

1.1
1.1

1.2
1.2

0.1

Figure 3.13.1 Profit simulation between two firms in stage 1 varying the relative cost
 cc,i 

 for three levels of degree of substitution in stage 1 (  1 ), 0.8, 0.5, and 0.2.
 cc, j 


77

─

phii1 phij1

= 0.5

--

=1

….

=2

0.5
0.5

0

g1
0.2

0.2

0.4

0.6

0.4

0.6

0.8

0.8

1.0

1.0

0.5

Figure 3.13.2: Profit simulation between two firms in stage 1
varying the degree of substitution in stage 1 (  1 )

 cc,i
for the three levels of the relative cost 
 cc, j



 , 0.5, 1, and 2.



3.6.2.2 Comparison of profits in stage 2
Standard Finding 2: When the degree of substitution of a conventional and sustainable product
are the same in stage 2 (  c   s ), and the levels of maximum willingness to pay for two types
2
2
of products are the same ( as  ac ), the firm that produces a conventional product will have a
higher profit than the firm that produces a sustainable product.

Explanation: A firm that produces a sustainable product has a higher cost than a firm which
produces a conventional product ( c s,i  cc, j ). However, consumers do not realize the
difference between the two types of products (  c   s ) and would like to pay the same amount
2
2

78

for both types ( as  ac ). Consequently, the firm that produces a conventional product gets a
higher profit than the firm that launches a sustainable product in this stage.

Next, the restrictions  c   s and as  ac are relaxed in order to compare the profits of
2
2
two firms in stage 2 in the general case.

Proposition 5: In stage 2, the firm that launches a sustainable product will get a higher profit
than the profit of the firm that produces a conventional product if:
c s, i

a.

a
 s , or
cc, j ac

b.

as
a
B
 s , and
c s,i
cc, j ac
c s, i

c
c
c
z  2 (1  xy   2 ( y  1))  2 x( y  1) 



 , x  2 , y 
where B 
s
c
c
2
y  2 (1  xz   2 ( z  1))  2 x( z  1) 





c s, i

a
, and z  s .
ac
cc , j

Explanation: The intuition of this proposition is similar to the intuition of proposition 3 except
proposition 3 compares the profits of the same firm when producing different types of products
(different stages) while proposition 5 compares the profits of different firms, which produce
different types of products in the same stage. When the relative cost of a sustainable product is
less than or equal to its relative maximum willingness to pay, the firm which launches a
sustainable product will get a higher profit than the firm that still produces a conventional
product (proposition 5a). This implies that the additional cost to produce the sustainable product
can be covered by the additional maximum willingness to pay for the sustainable product.
79

However, when proposition 5a does not hold, the firm that launches a sustainable product still
obtains a higher profit than the firm which produces a conventional product if the maximum

 a

willingness to pay over the cost of a sustainable product is sufficiently high  s  B  .
 c s ,i



The constant term B is a combination of the degree of substitution of a sustainable

c
  c  , the relative degree of substitution ( 2 ), the relative
product to a conventional product  
s
 2
2
 c s, i
cost of a sustainable product 
 c c, i



 , and the relative maximum willingness to pay for a



a 
sustainable product  s  . When the sustainable product is substituted for the conventional
a 
 c



 B

product to a higher degree, the value of constant term B decreases 
 0  , which makes
  c

 2

proposition 5b easier to satisfy. Intuitively, when the sustainable product is more substitutable
for the conventional product, consumers are more likely to purchase the sustainable product,
which causes the demand for the sustainable product to increase. Hence, the leader is more likely
to get a higher profit than the follower in stage 2. Similarly, when the relative degree of





B
substitution increases, the value of term B decreases 
 0  . This results in the
   c /  s 

  2 2


 

relative demand for the sustainable product over the conventional product to increase making the
leader more likely to receive a higher profit than the follower in stage 2. Next, when the relative

80



B
cost of the sustainable product increases, the value of term B increases 
 0  , and
  c s,i / cc, j








it is more difficult for proposition 5b to be satisfied. It is not surprising that the leader who
launches the sustainable product is less likely to have more profit than the follower if the relative
cost of a sustainable product increases. Lastly, when the relative maximum willingness to pay for



B
the sustainable product increases, the value of term B decreases 
 a / a   0  , which makes

s c


proposition 5b more likely to be satisfied. That is, when consumers would like to pay more for
the sustainable product compared to the conventional product, the leader firm who markets the
sustainable product is more likely to have a higher profit than a profit of the follower firm who
markets the conventional product in stage 2.
In addition, proposition 5b is not a rare case as shown in table 3.4. Assuming cc, j =

s
c = 0.9, and  2 = 0.9 (case 2 in table 3.4), the cost of the sustainable product
0.745, as = 5, 
2
c
2
can be higher than the cost of the conventional product by 4.45 times, which makes the leader’s
profit higher than the follower’s profit in stage 2. Table 3.4 also shows simulations when varying
a 
the relative maximum willingness to pay for the sustainable product  s  , and when the
a 
 c

 c 
 
relative degree of substitution in stage 2  2  and the degree of substitution of a sustainable
 s 
 2
product to a conventional product in stage 2 (  c ) are varied (case 1).
2

81

Table 3.4 Egg industry simulation of proposition 5b when varying the relative maximum
a 
willingness to pay  s  , the relative degree of substitution
a 
 c

 s 
 2
 c  , and the degree of
 
 2

substitution in stage 2 (  c ).
2

Highest value of
Case

1

2

s
2
c
2

0.5
0.5
0.5
0.9
0.9
0.9

c
2

0.5
0.5
0.5
0.9
0.9
0.9

as
ac
(1)

1.11
2.00
5.00
1.11
2.00
5.00

c s, i
cc,i

that the

sustainable product launching firm
gets a higher profit than the profit
of the firm which producing a
conventional product
(2)
2.20
4.60
6.40
1.88
4.45
6.38

c s, i
cc, j

as
ac

or (2)/(1)

1.98
2.30
1.28
1.69
2.23
1.28

Figure 3.14 shows the simulation results for the relationship between the difference in
profits of two firms in stage 2 (  s,i 2 *  c, j 2 * ), and the maximum willingness to pay for a

s
c = 0.9,  2 = 0.9, and a s = 1.4772. When
sustainable product ( as ) assuming cc, j = 0.745, 
2
c
ac
2
the relative cost of a sustainable product is less than or equal to the relative maximum
willingness to pay for a sustainable product (

c s,i
cc , j



c s, i a s
1 as
in figure 3.14), the
* , and

2 ac
cc, j ac

profit of the firm that launches a sustainable product is always higher than the profit of the firm
that produces a conventional product, which represents proposition 5a. When the relative cost of
82

a sustainable product is higher than the relative maximum willingness to pay for a sustainable
product (

c s, i

c s, i
a
a
 3 * s , and
 5 * s in figure 3.14), whether the profit of the firm that
cc, j
ac
cc, j
ac

launches a sustainable product is higher than the profit of the firm which produces a
conventional product or not depends on the value of maximum willingness to pay for a
sustainable product (proposition 5b). That is, the higher value of the maximum willingness to
pay for the sustainable product, the more likely that the firm launching the sustainable product
gets higher profits.

phii2 phij2
7

…

66
1
15
3
44
1
13
3
22
1
11
3

----

─

as
3

4

5
5

6

3
4
1
1
1
3
1
1
3
3
Figure 3.14 Profit simulation between two firms in stage 2 (  s,i 2 *  c, j 2 * ) varying
 c s, i
the maximum willingness to pay ( as ) for variety of relative costs 
 c c, i

a 
and the relative maximum willingness to pay  s  .
a 
 c

83






3.6.2.3 Comparison of profits in stage 3
Standard Finding 3: In stage 3, when both firms have the same (marginal) cost ( c s,i  c s, j ),
the follower firm will get a higher profit with a lower price and a higher quantity. This is the
same result as (Boyer and Moreaux 1987; Shy 1995).

Explanation: Under the Stackelberg price leadership model, the leader sets the price first, and the
follower sets its price after observing the price in the new market. Under the same marginal cost,
the follower will undercut the price of the leader in order to get a higher market share and a
higher profit. In the Stackelberg price leadership model, the follower gets a second-mover
advantage in the sense that the follower has more information about the leader’s price and can set
their price accordingly.

From the comparisons of profits of the two firms in each stage, it can be concluded that
there are three factors that encourage a firm to be a leader. First, if a firm has a disadvantage in
cost to produce a conventional product, a firm has an incentive to switch to produce and market a
sustainable product. This is because a firm can increase its market share and profit by launching
a sustainable product instead of producing a conventional product, which is dominated by
another firm who has a lower cost. Second, a firm would like to launch a sustainable product first
to get a higher profit if the relative cost of the sustainable product is less than or equal to its
relative maximum willingness to pay. That is, when the consumers demand is strong for the
sustainable product while the cost of the sustainable product is not too high, the probability of
the leader launching is higher in stage 2. Lastly, a cost advantage for producing a sustainable
product encourages a firm to be a leader by launching first. If a leader firm has the same or

84

higher production cost for the sustainable product than the follower, the leader firm will lose the
market share, and has less profit than the follower when the follower also switches to market a
sustainable product. This is because the follower can observe the leader’s sustainable product
price and then undercut the price to get higher quantity, market share, and profit. Thus, if a firm
has a cost disadvantage for producing a sustainable product, it is better to wait to launch a
sustainable product later to get a higher profit.

Figure 3.15 is a flow chart that summarizes the relationship of all propositions, standard
findings, and their restrictions. The boxes named stage 1, stage 2, and stage 3 show the findings
from the comparisons of the profits of two firms in stages 1, 2, and 3, respectively. The shaded
boxes show the propositions and their restrictions regarding the movement to the next stage.
There are two sets of boxes that contain the propositions and their restrictions regarding moving
to stage 2. This implies that there are two separate ways to move to stage 2: 1) comparing the
leader’s profits in stages 1 and 2, and 2) comparing the leader’s profits in stages 1 and 3. For the
left path, we compare the leader’s profits in stages 1 and 2 when the follower will decide not to
launch the sustainable product; hence, the boxes on the left hand side show the progression to
stage 2. For the right path, we compare the leader’s profits in stages 1 and 3 when the follower
will decide to launch the sustainable product, so the boxes on the right hand side show the move
to stage 3.

85

STAGE 1
Standard Finding 1 (no restriction)

Move to stage 2 (

)

Move to stage 2 (

Common Assumptions:
Proposition 4
(

)

Common Assumptions:
Proposition 2

, and

)

(

, and either

or

, and

Proposition 3
(
,

,
)

, and

STAGE 2

STAGE 2

Proposition 5 (no restriction)

)

Proposition 5 (no restriction)
Standard Finding 2

Standard Finding 2

(

and

(  c   s and as  ac )
2
2

)

Move to Stage 3 (

)

Proposition 1
(

,

, and

)

STAGE 3
Standard Finding 3 (

Figure 3.15 Summary of decision making arranged by propositions,
standard findings, and their restrictions.

86

)

3.7 Conclusion
A food manufactures’ decision making model is constructed to illustrate the
launching strategy of a sustainable food product. The factors that influence firms to
launch a sustainable product are from both demand and supply side. That is, on the
demand side, food manufacturers decide to launch a sustainable product when consumers
realize and prefer the difference between the conventional and the sustainable products;
moreover, they prefer and are willing to pay more for the sustainable product. On the
supply side, the firms get higher profits when launching the sustainable product if the
relative cost of the sustainable product is low enough; that is less than or equal to the
relative maximum willingness to pay for the sustainable product. This implies that the
additional cost to produce the sustainable product has to be covered by the additional
maximum willingness to pay for the sustainable product to induce the firms to launch the
sustainable product.
The study also determines the conditions that encourage a firm to be a leader. A
firm decides to be a leader when it has a cost disadvantage in producing a conventional
product, but has a cost advantage to producing a sustainable product. Moreover, the firm
decides to be a leader when the relative cost of the sustainable product has to be less than
or equal to its relative maximum willingness to pay, which is the same condition that
encourages the firm to launch the sustainable product.
The leader firm and the follower firm have different criterion to launch a
sustainable product. On one hand, the leader firm has to determine first whether the
relative maximum willingness to pay for the sustainable product is higher than the
relative cost of the sustainable product (the relative term concept), and the leader firm

91

decides to launch the sustainable product and attains a higher profit when this relative
condition holds. If not, the leader has to determine whether the absolute value of the
maximum willingness to pay for the sustainable product is sufficiently higher than the
absolute value of the cost of the sustainable product or not, and decides to launch when
the absolute concept holds. On the other hand, the follower’s most influential factor is the
degree of substitution; consumers have to realize the difference between the sustainable
and conventional product. If the consumers perceive the two products to be close but not
the same, then the follower firm will decide to produce a me too product only when the
costs of productions of the two goods are sufficiently close. In addition, the simulation
results show that egg manufacturers, both the leader and the follower firms, should
launch sustainable egg products to get higher profits. As a consequence, both national
and store brand manufacturers have increasingly marketed sustainable eggs the past few
years.

92

APPENDICES

93

APPENDIX B: PROFIT DERIVATION BY STAGE

Stage 1:
In this stage, both firms produce conventional products and simultaneously
choose a price as a strategy. Firm i ’s and firm j ’s inverse demand functions are:
pc, i  ac  qc, i   1qc, j , and

(3.1)

pc, j  ac  qc, j   1qc,i .

(3.2)

Hence, the demand functions are: qc,i 

q c, j 

(1   1 )ac   1 pc, j  pc,i
, and
1   1 2

(1   1 )ac   1 pc,i  pc, j
. Then, we can set firm i ’s profit function (  i ) as:
2
1   1 

 (1   1)ac   1 pc, j  pc,i 
( p  c ) , where c and c
c, i
c, j are marginal costs

 c,i c,i
2
1   1 



i  

to produce a conventional product for firm i and firm j , respectively. The first order
condition (FOC) is then:
FOC.

 i
pc,i

= 0,

 (1   1)ac   1 pc, j  pc,i   pc,i  cc,i



  1   2
2
1   1 
1

 

We also check the second order condition (SOC) which is






= 0.

 2 i
2
=
< 0. This
pc,i 2  1 2  1

confirms a maximum. Rearranging and with some substitutions, the reaction functions
are:
94

p c, i

 (1   1 )ac   1 pc, j  cc,i
=

2



 , and



(B.1)

p c, j

 (1   1 )ac   1 pc,i  cc, j
=

2



.



(B.2)

From solving (B.1) and (B.2) simultaneously, we get the optimal prices in stage 1
( pc,i1 * , and pc, j1 * ) as follows:

pc,i1 *

  2     2 a   c


1
1  c 1 c, j  2cc,i 


=
 , and
 4   2 






1 




(B.3)

p c, j 1 *

  2     2 a   c  2c

1
1  c 1 c,i
c, j

= 
 4   2 


1 




(B.4)



.



By plugging the equilibrium prices into the demand functions, the equilibrium quantities
in stage 1 ( qc,i1 * , and qc, j1 * ) are as follows:

qc,i1 *

2
2
(2   1   1 )ac   1cc, j  ( 1  2)cc,i
=
, and
2
4
4  5 1   1

(B.5)

q c, j 1 *

2
2
(2   1   1 )ac   1cc,i  ( 1  2)cc, j
=
.
2
4
4  5 1   1

(B.6)

Then, firm i and firm j ’s profits in stage 1 (  i

95

1

1
* , and  j * ) are:

1
 c,i *

1

 c, j *

 

2 
2
ac  2   1   1    2   1 cc,i   1cc, j 


 

= 
2
  2  4  1   2 
 1
 
1

 


2

 

2 
2
ac  2   1   1    2   1 cc, j   1cc,i 


 

= 
2
  2  4  1   2 
 1
 
1

 


2

, and (B.7)

.

(B.8)

Stage 2:
The leader firm which is assumed to be firm i launches a sustainable product to
the market, while the follower (firm j ) still supplies a conventional product. Both firms
choose price simultaneously as the strategy. An inverse demand function of firm i who
produces the sustainable product is,

s
ps,i  as  qs,i   2 qc, j ,

(3.3)

and an inverse demand function of firm j who produces the conventional product is,

c
pc, j  ac  qc, j   2 qs,i ,

(3.4)

where  s   c . From the inverse demand functions, the demand functions are:
2
2

c
a   sa   p   s p
 a   ca   p
 s
 c
s, i
c, j   2 p s , i
2 c
2 c, j
2 s




q s, i 
, and qc, j 
. Then,
s c
s c
1   2 2
1   2 2
firm i ’s profit function is,  i  ( ps,i  cs,i )qs,i

as   2sac  ps,i   2s pc, j and the first order condition is:
 ( ps,i  cs,i )
s c
1   2 2

96

 i
p s,i

= 0,

s
s
( c s , i  p s , i )   a s   2 a c   p s , i   2 p c, j )




s c
1  2 2

FOC.

= 0.

 2 i
2
The SOC is
=
< 0; thus, confirming a maximum profit in equilibrium.
c s
p s,i 2  2  2  1
a   sa   c   s p
 s
2 c  s, i
2 c, j


Rearranging the FOC, p s, i =
, and pc, j =
2
c
 a   ca   c
 c
2 s  c, j   2 p s,i


are obtained. Next, from the reaction functions of firm i
2
and j , the equilibrium prices in stage 2 ( p s,i 2 * , and pc, j 2 * ) are:

 2a   s a   s  c a   2c   s c

s, i
2 c
2 2 s
2 c, j

2* =  s
p s,i
, and
 4   s c 

2 2



(B.9)

c
 2a   c a   s  c a   2c

c, j   2 c s , i
2 s
2 2 c

2* =  c
.
p c, j
 4   s c 

2 2



(B.10)

Then, the equilibrium quantities in stage 2 ( q s,i 2 * and qc, j 2 * ) are:

q s,i 2 * =

s
s c
s
s c
(2a s  ac 2  a s  2 2 )   2 cc, j  ( 2 2  2)cs,i
, and
2 c 2
s c
s
4  5 2 2    2    2 
   
   

97

(B.11)

qc, j 2 * =

c
s c
c
s c
(2ac  a s  2  ac 2 2 )   2 c s,i  ( 2 2  2)cc, j
.
2 c 2
s c
s
4  5 2 2    2    2 
   
   

(B.12)

Also when the equilibrium prices and quantities are plugged back into the profit
functions, the equilibrium profits of both firms (  i 2 * , and  j 2 * ) are:

2
  s c

s
s
s c


 a s   2 2  2   ac  2    2 cc, j   2   2 2 c s,i 






 , and (B.13)
 s,i 2 * = 
2
  s  c  4  1   s  c 
 2 2
 
2 2

 


 c, j 2 * =

  s c

c c
s c


 ac   2 2  2   a s  2    2 c s,i   2   2 2 cc, j 




 

2

  s  c  4  1   s  c 
 2 2
 
2 2

 


2
.

(B.14)

Stage 3:
In this stage, the follower firm (firm j ) also supplies a new sustainable product.
Both firms still choose prices as a choice variable, but the game is now a sequential in
this stage. The leader will choose its own price first and then the follower will decide to
choose their price later. That is, the leader will put the reaction function of firm j into its
own objective function in order to protect the new market. An inverse demand function
of firm i who is the leader is,
p s, i  a s  q s, i   3 q s, j ,

(3.5)

and an inverse demand function of firm j who produces a sustainable product is,
p s, j  a s  q s, j   3 q s, i .

(3.6)

98

The demand functions in this stage are: q s,i 

(1   3 )a s   3 p s, j  p s,i
, and q s, j 
1   3 2

(1   3 )a s   3 p s,i  p s, j
. From backward induction, we find firm j ’s reaction
1   3 2
function in order to put it into firm i ’s objective function later. Firm j ’s profit function



is:  j  ( p s, j  c s, j )q s, j  p s, j  c s, j

 (1   3 )as   3 p2s,i  ps, j . Firm
1   3 

j ’s first

order condition is as follows:
FOC.

 j
p s, j

= 0,

(cs, j  p s, j )  (1   3 )a s   3 ps,i  p s, j
= 0.
2
1  3
 2 j
2
The SOC is
=
< 0 which implies the maximum equilibrium profit.
p s, j 2  3 2  1
From the FOC,
(1   3 )a s   3 p s,i  c s, j
2

= p s, j .

(B.15)

Firm j ’s reaction function is shown in equation (B.15). Firm i ’s profit function is,

 i  ( p s, i  c s, i ) q s, i   p s, i  c s, i 

(1   3 )a s   3 p s, j  p s,i
. Then, we substitute the
1   3 2

reaction function of firm j (equation (B.15)) into firm i ’s profit function,  i 

99

 p s, i  c s , i

 1   2
3



(1   )a    (1   3 )as   3 ps,i  cs, j
3 s
3


2





  ps,i  . From the first order





condition, the equilibrium price of firm i in stage 3 is:

2
2
(2   3   3 )a s   2   3 cs,i   3cs, j




p s,i * =
.
2)
2(2   3

3

(B.16)

3
Substituting p s,i * into the reaction function of firm j , we get the equilibrium price of
firm j in stage 3 as:

2
3
2
2
(4  2 3  3 3   3 )a s   4   3 cs, j  (2   3 ) 3cs,i




p s, j * =
.
2)
4(2   3

3

(B.17)

2   3 2
We also check SOC of firm i ’s profit function; that is
< 0 which guarantees
 3 2  1
the maximum equilibrium profit of firm i . Substituting the equilibrium prices into the
demand function, so the equilibrium quantities are:

q s,i 3 * =

2
2
(2   3   3 )a s   2   3 cs,i   3cs, j




, and
2
4(1   3 )

(B.18)

 4  2  3 2   3 a   4  3 2 c
  2   2  3c s, i

3
3
3 s 
3  s, j 
3






qs, j3 * =
. (B.19)
 2  3 2   4 
4
3
3


Finally, the equilibrium profits in stage 3 (  s,i 3 * , and  s, j 3 * ) are:

100

2
 ( 2     2 ) a  ( 2   2 )c   c


3 s, j 
3
3 s
3 s, i
 , and
 s,i 3 * = 
8 2  3 2   4 

3
3



(B.20)

  4  2  3 2   3 a   4  3 2 c
  2   2  3c s, i  2


3
3
3 s 
3  s, j 
3






 .
 s, j 3 * = 
2
16  2  2  1   2 
 3
 
3

 

(B.21)

101

APPENDIX C: DEMAND UNCERTAINTY: AN EXPANSION OF THE MODEL

The author focuses on an expansion of the model for demand uncertainty. That is
whether consumers are willing to pay more for a sustainable product is still questionable
for food manufacturers. This issue is important for manufacturers’ decision making when
considering launching a sustainable product. Many market entry articles are concerned
about demand and/or profitability uncertainty (Maggi 1996; Hirokawa and Sasaki 2001;
Jeitschko and Creane 2009).
The reasons why the uncertainty on the demand side is explicitly represented in
the model are: 1) the author would like to make the model tractable and as simple as
possible; therefore, the uncertainty on a supply side is not included in the model, and 2)
the firm can control costs, but not consumer demand. Uncertainty is modeled only in
stage 2 because for three reasons. First, stage 1 is the current situation. Firms know the
existing demand; hence, there is no uncertainty. Second, in stage 2 firms never had a
sustainable product in the target market, so they cannot predict the consumers’ maximum
willingness to pay for the new product. Third, since the sustainable product was launched
in stage 2, the firms know the consumers’ maximum willingness to pay for the new
product already. Therefore, there is no uncertainty in stage 3.
Many authors applied the real option method to consider uncertainty of launching
a product or investing in research and the development of a new product (Smit and
Ankum 1993; Grenadier 2000; Botteron, Chesney, and Gibson-Asner 2003; Russo,
Cardillo, and Perito 2003; Schwartz and Trigeorgis 2004; Kijima and Shibata 2005).
However, the real options is not suitable to our model since the real options is usually set

102

up to deal with a pattern of continuous time and infinite period while the model in this
paper is a three-stage discrete time game.
The idea of the binary distribution of maximum willingness to pay for the
sustainable product ( a s ) is adapted from the demand function in Jeitschko and Creane
(2009). An inverse demand function for a sustainable product in stage 2 is adjusted from

s
Jeitschko and Creane (2009) to be p s, i  a s  (1   )ac   q s, i   2 qc, j where 
represents the leader firm’s expectation that consumers would like to pay more for the
new sustainable product, and  has a value between zero and one. According to the
inverse demand function defined above, the expected maximum willingness to pay for
the sustainable product is defined in Figure 3.16.

The expected maximum willingness
to pay for a sustainable product

0

1

Figure 3.16 The relationship between the expected maximum willingness to pay
for the sustainable product ( a s ) and the probability that the leader firm expects that
consumers are willing to pay more for the sustainable product (  ).

The results when uncertainty is added into the model expand the standard findings and
propositions.
103

Standard Finding 4: When the levels of maximum willingness to pay for the sustainable
and conventional product are the same ( a s  ac or   0 ), the case with risk and without
risk are the same (  disappears). This is the case because the maximum willingness to
pay for the sustainable and conventional product are the same, and uncertainty
represented by  disappears.

Explanation: This assumption represents the worst situation; that is the leader firm
expects to get zero premium from the new product. Therefore, there is no uncertainty
defined in the model since the firm assumes the lowest maximum willingness to pay for
the sustainable product already.

Proposition 6: When

c s, i

a
c
s
 s ,  1   2   2 , and cc,i  cc, j , a leader firm i will
c c, j a c

decide to launch the sustainable product and move to stage 2 if  

c s, i
as

.

Explanation: From proposition 6, the leader will get a higher profit when launching the
sustainable product if the probability to get the premium from consumers is high enough
such that  

c s, i
as

. Intuitively, proposition 6 states that the leader will decide to launch a

sustainable product when their expectation to get a price premium from the new
sustainable product is higher than the ratio of the cost over the maximum willingness to
pay for the sustainable product. When  is equal to 1 or when the leader expects that
consumers certainly have a willingness to pay more for the sustainable product, the leader
104

 c s, i
will launch the sustainable product to get a higher profit 1 

as



 which is consistent



with proposition 4a. When  is equal to 0, or the leader does not expect to get the price
premium from the sustainable product, the leader will not launch the sustainable product

c

 0  s, i

as



.



For the simulation part, it was determined that

cs,i
= 1.66, c c, i = 0.75 (see
cc, j

c
s
section 3.5), and  1   2   2 = 0.7. Figure 3.17 shows the simulation results, which are
consistent with proposition 6. That is  s, i 2 *  c, i1 * is a positive value when  

or the leader will get a higher profit when launching the sustainable product if  

105

c s, i
as

c s, i
as

.

,

1.2
1.2
1.0
0.8
0.8

--

0.4
0.4
0.2

0.2

0.2

0.4
0.4

0.2

0.6
0.6

0.8
0.8

= 0.3
csioveras 0.3

─

0.6

= 0.5
csioveras 0.5

1.0
1.0

11
1

Figure 3.17 Leader profit simulation between stages 1 and 2 (  s, i 2 *  c, i1 * )
varying the probability of leader’s expectation that consumers are willing to pay more for
the sustainable product (  ) for two levels of the cost over
 c s, i 

the maximum willingness to pay 
 a  , 0.3 and 0.5.
 s 

Proposition 7: When

c s, i

a
 s , and the degree of substitution of the conventional and
c c, j a c

sustainable product are the same in stage 2 (  c   s ), the leader will get a higher profit
2
2
than the follower if  

c s, i
as

.

Explanation: Proposition 7 is similar to proposition 6, except it compares the leader’s
profit and the follower’s profit in stage 2 instead of comparing the leader’s profits in
stages 1 and 2. The leader will get a higher profit than the follower when  

c s, i
as

when the expectation of a price premium from the new sustainable product is high

106

or

enough or higher than the cost over the maximum willingness to pay for the sustainable
product. When the leader expects that consumers have a willingness to pay more for the

 c s, i
sustainable product (  = 1), the leader has a higher profit than the follower 1 

as



,



which is consistent with proposition 5a. When  is equal to 0, the leader gets less profit

c s, i 

 , which is consistent with standard finding 4.
than the follower  0 

as 


Assuming that

c s, i
c c, j

c
s
= 1.66, c c, i = 0.75, and  1   2   2 = 0.7, figure 3.18

shows the simulation results, which are consistent with proposition 7. That is

 s, i 2 *  c, j 2 * is a positive value when  

c s, i
as

meaning that the leader will get a

higher profit when launching a sustainable product if  

107

c s, i
as

.

1.5
1.5

1.0
1.0

--

= 0.3
csioveras 0.3

─

= 0.5
csioveras 0.5

0.5
0.5

0.20.2

0.4

0.4

0.6
0.6

0.8
0.8

1.0
1.0

Figure 3.18 Profit simulation between two firms in stage 2 (  s, i 2 *  c, j 2 * )
varying the probability that the leader expects that consumers are willing to pay more for
the sustainable product (  ) for two levels of the cost over
 c s, i 

the maximum willingness to pay 
 a  , 0.3 and 0.5.
 s 

Figure 3.19 is an expansion of figure 3.15. It shows all standard findings,
propositions and their restrictions when including risk variable (  ) in the model.

108

STAGE 1
Standard Finding 1 (no restriction)
Move to stage 2 (
No Risk ( )
Common
Assumptions:
Proposition 4
(

)

Move to stage 2 (

)

Risk ( )
Common
Assumptions:
Proposition 6
(

,

and

Common Assumptions:
Proposition 2

or

(

,

)
, and

, and either
, and

,
)

Proposition 3
(
,

, and

)

)

STAGE 2
Standard Finding 4 ( a s  ac )
Standard Finding 2
( c   s )
2
2
No Risk (  )
Risk (  )
Proposition 5 Proposition 7
c s, i a s

(no restriction) (
, and
c c, j a c

STAGE 2
Standard Finding 4 (
Standard Finding 2
(

)

)

No Risk ( )
Proposition 5

Risk ( )
Proposition 7

(no restriction)

(

, and

c
s
2 2)

)
Move to Stage 3 (

)

Proposition 1
(

,

, and

)

Standard Finding 3 (

)

STAGE 3

Figure 3.19 Summary of decision making arranged by propositions, standard findings
and their restrictions when uncertainty is introduced.

109

APPENDIX D: THE PROOF AND DERIVATION FOR PROPOSIONS AND
STANDARD FINDINGS

Proposition 1 Proof

 s, j 3 * -  c, j 2 * =



3
 (  s 
 2

   





ac  cc, j 2
2
2
1
16(  s   1) 
 2
 
z

2
 s 2 1
2
s  21 
 
 
 3  s   2 s  4) 2 16  2  z   2
 2
z 
2
 
 .
 

2
2
  s   2) 2
  s   4) 2

( 
( 

 2
 2



ac  cc, j 2
2
1
16(  s   1) 
 2
 
z

is a negative value since 0   c  1 .
2
2

2
 s 2 1
s  21
3
2

16   

2
(  s   3  s   2 s  4) 2
 2  z
 2
 2
z
2
 will be a negative value
 
 
 

2
2
(  s   4) 2
(  s   2) 2
 2
 2
 
 
21

1
3
2
16   s 
 s 2 
 2
  s   3  s   2 s  4) 2
2
( 
  z
 2
z
2


 2
 
if
is greater than
2
2
(  s   4) 2
(  s   2) 2
 2
 2
 
 

110

2
. The

3
2
(  s   3  s   2 s  4) 2
 2
 2
2
 
conditions to make  
greater than
2
(  s   2) 2
 2
 

21

1
16   s 
 s 2 
 2
2
  z
z


2
(  s   4) 2
 2
 

2
are shown in proposition 1.

Proposition 2 Proof
c s, i c s, j
a
Given that
 s , and  1   3 .

c c, i c c, j
ac

2
 c   c ( 2  2)  a ( 2    2) 
 s, j 3

s, i 3
s 3
3

 s, i *  c, i * = 
2
8  3 2  4    3 4  3 3 2  2 

 


 

3

1


 1 2 
 1 2 
   3 4  8 3 2     1  16    1  . Since both terms are positive values,

 z 

 z 









3

1

*
 *, i   c, i is always greater than zero.
s

111

Given that

c s, i
c c, i



c s, j

a
 s , c c, i  c c, j and c s, i  c s, j .
c c, j
ac

2

1 

8( 1  1)  
3
1
 ( 3  1)( 3  2) 2
1
z 

 s, i *  c, i * = a s  c s, j 2 
 . The first
8
  33   3 2  2 3  2 ( 1  2) 2 ( 1  1) 









term is a positive value and the second term is a positive value since

2
1
8( 1  1) 
3
( 3  1)( 3  2) 2
z
* 1
is greater than
. Therefore,  *, i   c, i is
s
 33   32  2 3  2
( 1  2) 2 ( 1  1)

always greater than zero.

Proposition 3 Proof


1
1 2 

2 (  2) 2 8(c s, j    a s  ) 
 y
3
1
 ( a s  c s, j )
3
1
z 
 

 s, i *  c, i * = ( 3  1)
 . The
8
 33   3 2  2 3  2
( 3  2) 2 ( 3  1) 






first term is a negative value.

(a s  c s, j ) 2 ( 3  2) 2

 33   3 2  2 3  2

is a negative value while

1
1
8(c s, j    a s  ) 2
 y
z
 
is a positive value. The second term will be a negative value
( 3  2) 2 ( 3  1)

112

1
1
8(c s, j    a s  ) 2
2 (  2) 2
 y
( a s  c s, j )
z
3
 
when
is greater than
or when the
( 3  2) 2 ( 3  1)
 33   3 2  2 3  2

conditions shown in proposition 3 are satisfied.

Proposition 4 Proof

2

1

 s, i *  c, i * =

1
2
2 
  c 2
 
     4     c   1
 2
 2 
  


 


2

2
2 
 
 c 2
 c 2
1
1
c  2  c
c 1  2  a  c 2   c 1  2 
  s , i  a s    c s, i      
 s  
 
 2    2
2 y
2 z
 
 
 2 

 2 


y
z







 



1
. The first term (
) is a negative value since 0   c  1 . The
2
2
  c 2
   c 2 
     4       1
 2 
  2 


 


2
2
 c 2
c  2   c 1  a 1  is less than
    
  s, i
second term is a negative value when
s 
2
 2 
 
y
z


 

2

2
2




 c    c    c 1  2   a    c    c 1  2   . The conditions that make the
 
s  2 
2 y
2 z


 s, i   2 


 


second term a negative value are shown in proposition 4.

113

Standard Finding 1 Proof
The difference of profits in stage 1 between two firms (  c,i1 * -  c, j1 * ) is

(2a c  cc, i  cc, j )(cc, i  cc, j )
. The denominator is a negative value since 0 <  1 < 1
2 4
1
and the first term of the numerator is a positive value since a c  c c, i and a c  c c, j . If
c c, i is greater than cc, j ,  c,i1 * -  c, j1 * will be less than 0. This means that a firm

that has a higher cost also has a lower profit, and vice versa. Moreover,  1 has no effect
on the sign of  c,i1 * -  c, j1 * , but does have an effect on the amount of difference in
profits.

Standard Finding 2 Proof
When  c   s , and a s  ac ,  s,i 2 * -  c, j 2 * is
2
2


 2 
 1  


 1 

c s, i  c s, i     1  2a s     1 
 y  
 y 


  



which equals
2
 c   4
 2
 

  1  
 1  
c s, i     1 c s, i     1  2a s ) 
  y  
 y  

   
  
 . The denominator is a negative value since
2
 c   4
 2
 

114

1
c
0   2  1 . The numerator is a positive value since 0     1 and c s, i  a s . Thus,
 y
 

 s,i 2 * -  c, j 2 * is less than zero.

Proposition 5 Proof

2

2

 s, i *  c, j *


21
21




1
1
  c s, i    c 
c
 2  as   c 
c
 2
 2
 2
2 xy
2 xz
  x

  x







=
2
 c 2 1
  c 2 1 
 
 4  
 1
 2  x
  2  x 

 


2

+

2

 c 2 1
 c 2 1
1
1 
c
c
 c  
  2  2   as    2 
  2  2 
   xz
 s,i   2  xy
y
z 





.
2
21
21 
 c 
  c
 
 4  2 
 1
 2  x
   x 
 

 

The first term is a positive value while the second term is a negative value since

2
2
c
0   2  1 and 1  x .  s, i *  c, j * will be greater than zero when the absolute
value of the first term is greater than the absolute value of the second term; that is when
the conditions in proposition 5 are satisfied.

115

Standard Finding 3 Proof

2
2
3
3 * -  3 * = (a s  c s, j ) (3 3   3  4) 3 . The denominator is a positive
 s,i
s, j
16( 3  1)( 3 2  2) 2
value while the numerator is a negative value since (3 32   3  4) is less than zero.
Therefore,  s,i 3 * -  s, j 3 * is a negative value.

Standard Finding 4 Proof

s
From appendix C, p s, i  a s  (1   )ac   q s, i   2 qc, j . Thus,
p s, i  ac  (1   )ac   q s, i   s qc, j  ac  q s, i   s qc, j .
2
2

Proposition 6 Proof

 s,i 2 * -  c,i1 * =

1
2

2

(    4) 2 (    1)
 1
 1
 
 





2
[ a c  cc, j 2   1 2   1  2  




2


2    1   2   a  1   2   1    2  1     2  2    z 2 ].
 cc, j   1 


 1

1 
c  1
1






z
z





The first term is a negative value while the second term can be a positive or negative
value. Thus  s,i 2 * -  c,i1 * will be a positive value when the second term is a
negative value or when the condition in proposition 6 is satisfied.

116

Proposition 7 Proof

 1 
1 
1

  1 c s, i   1  a s  (  2)    (c s, i  a s )


 z 
z 
z

 s,i 2 * -  c, j 2 * =
. The
2
 c   4
 2
 
denominator is a negative value since 0   c  1 . The numerator is a negative value
2


1 
1

when  c s, i   1  a s  (  2)    (c s, i  a s ) is a negative value since 1  z . The


z

z



1 
1

condition that makes  c s, i   1  a s  (  2)    (c s, i  a s ) a negative value is


z

z


shown in proposition 7.

117

APPENDIX E: SIMULATION STEPS AND EXAMPLE

The Steps for Simulation
The information from the egg industry is used for simulation. The steps for simulation are
described as follows:
Step 1: Find the difference in profits.
Step 2: Substitute the value of the parameters by using the information from the egg
industry.
Step 3: Substitute the restriction(s) of each proposition/standard finding.
Step 4: Arbitrarily determine the value of some unobserved parameters, such as the
degrees of substitution, and the levels of maximum willingness to pay for the
sustainable and conventional product. These depend on the restriction(s) of each
proposition/standard finding and in which parameter is of interest.
Step 5: Plot a graph where the y-axis represents the difference in profit and the x-axis is
an unobserved variable which is related to the proposition/standard finding.

Simulation Example
An example used to demonstrate the steps for simulation is how to determine the shape in
figure 3.3 which explains proposition 1. Proposition 1 shows under which conditions the
follower decides to launch a sustainable product. The steps for simulation are as follows:

118

Step 1:
Compare the follower’s profits in stages 2 and 3. That is, the follower will launch a
sustainable product when its profit in stage 3 is higher than its profit in stage 2
(  s, j 3 *   c, j 2 * ) or when the difference in profits for figure 3.3 (  s, j 3 *  c, j 2 * )
is a positive value. From Mathematica  s, j 3 *  c, j 2 * is

2
 2(a  c )   c (a  c )   c  s (a  c ) 


c
c, j
s, i
c, j 
2 s
2 2 c

2
  c  s  4    c  s  1
 2 2
  2 2


 


2

 2   2   c  4  3 2   a  4    2  3   2   
  3 c s, i 
3  s, j 
3  s
3
3
3 









.
2
16  3 2  2    3 2  1

 


 


Step 2:
The cost for conventional eggs ( cc, j ) from the egg industry is $0.75 per dozen, and the

 c s, i
relative cost 
 c c, j



 of the sustainable product for the worst case scenario is 1.66.



Substituting these values into  s, j 3 *  c, j 2 * equation,  s, j 3 *  c, j 2 * is equal to

1.49   2c as  0.75 2s  1.24 ac  2c 2s  2 2  2c 2s  42  2c 2s  1




0.06 2.47 3  1.24 3 2  as  3  3.24 3  1 3  1.24  cs, j 3 3 2  4

 3 2  22  3 2  1
119

2 .

Step 3:
Substitute the restrictions of proposition 1 into  s, j 3 *  c, j 2 * from step 2. The

c s, j a s
restrictions are  c   s   3 , c s, i  c s, j , and
. Then,  s, j 3 *  c, j 2 * is

2
2
c c, j a c



1
 s 2 
16   2   1
 

 


2 

2 

2

s
s
s
(1) as  1.242  2  3.24  2  1  2  1.24

 

2
 s2

  2  2











 

2



s
s
s
s 2 
16  0.75  2  0.81  2  2.47  as   1.20   2  0.60  2  




 .

2
 s2

  2  4





 

Step 4:
We assume that a s = 5; then substitute this value into  s, j 3 *  c, j 2 * in step 3. Thus,

 s, j 3 *  c, j 2 * equals







1
2
4
  c 2 
     1 8  6  c     c  
 2
 2 
 2 

 
  









s
s
s
s
s
[0.89  2  6.30  2  1.59  2  1.15  2  0.93  2  0.96

s
s
s
s
s 2
 2  1.31  2  1.83  2  2.11  3.13  2.24 2   2 ] .













120

 

2

Step 5:
From step 4,  s, j 3 *  c, j 2 * is a function of only  s . Therefore, we can draw the
2
graph of  s, j 3 *  c, j 2 * by determining  s, j 3 *  c, j 2 * as the y-axis and  s as
2
the x-axis as shown in figure 3.20.

phij3 phij2

2
2.0

1.5

1
1.0

0.5

g2c
0.2
0.2

0.4
0.4

0.6
0.6

0.8
0.8

1.0
1.0

Figure 3.20: Profit simulation between stages 2 and 3 for the follower
varying the degrees of substitution (  c   s   3 ) when the level of relative costs of a
2
2
 c s, j 
 equals 1.66 (the worst case scenario).
sustainable product 
 c c, j 


When these steps are repeated for the case that
illustrated as in figure 3.3.

121

cs , j
cc, j

equals 1.2, the graph can be

REFERENCES

122

REFERENCES

Acs, Zoltan J., and David B. Audretsch. 1987. “Innovation, Market Structure, and Firm
Size.” The Review of Economics and Statistics: 567–574.
Benedetto, C. Anthony. 1999. “Identifying the Key Success Factors in New Product
Launch.” Journal of Product Innovation Management 16 (6): 530–544.
Bennett, Richard. 1998. “Measuring Public Support for Animal Welfare Legislation: A
Case Study of Cage Egg Production.” Animal Welfare 7 (1): 1–10.
Botteron, Pascal, Marc Chesney, and Rajna Gibson-Asner. 2003. “Analyzing Firms’
Strategic Investment Decisions in a Real Options’ Framework.” Journal of
International Financial Markets, Institutions and Money 13 (5): 451–479.
Boyer, Marcel, and Michel Moreaux. 1987. “Being a Leader or a Follower: Reflections
on the Distribution of Roles in Duopoly.” International Journal of Industrial
Organization 5 (2): 175–192. doi:http://dx.doi.org/10.1016/S01677187(87)80018-8.
Broder, Christina. 2010. “Accenture and UN Global Compact Study Finds Sustainability
Now Number One Priority for CEOs”. Big4.com.
http://www.big4.com/news/accenture-and-un-global-compact-study-findssustainability-now-number-one-priority-for-ceos/.
Bröring, Stefanie. 2007. “Innovation Strategies in the Emerging Nutraceutical and
Functional Food Industry.” In .
https://www.ifama.org/events/conferences/2007/cmsdocs/1104_session2_Trovato
re.pdf.
Debruyne, Marion, Rudy Moenaertb, Abbie Griffinc, Susan Hartd, Erik Jan Hultinke, and
Henry Robben. 2002. “The Impact of New Product Launch Strategies on
Competitive Reaction in Industrial Markets.” Journal of Product Innovation
Management 19 (2): 159–170.
Dixit, Avinash. 1979. “A Model of Duopoly Suggesting a Theory of Entry Barriers.” Bell
Journal of Economics 10 (1): 20–32.
Dockner, Engelbert, and Steffen Jørgensen. 1988. “Optimal Pricing Strategies for New
Products in Dynamic Oligopolies.” Marketing Science 7 (4): 315–334.
Fulton, Murray, and Konstantinos Giannakas. 2004. “Inserting GM Products into the
Food Chain: The Market and Welfare Effects of Different Labeling and

123

Regulatory Regimes.” American Journal of Agricultural Economics 86 (1): 42–
60. doi:10.1111/j.0092-5853.2004.00561.x.
Grenadier, Steven R. 2000. “Option Exercise Games: The Intersection of Real Options
and Game Theory.” Journal of Applied Corporate Finance 13 (2): 99–107.
Henard, David H., and David M. Szymanski. 2001. “Why Some New Products Are More
Successful Than Others.” Journal of Marketing Research 38 (3): 362–375.
Hirokawa, Midori, and Dan Sasaki. 2001. “Endogenously Asynchronous Entries into an
Uncertain Industry.” Journal of Economics & Management Strategy 10 (3): 435–
461. doi:10.1111/j.1430-9134.2001.00435.x.
Hitsch, Günter J. 2006. “An Empirical Model of Optimal Dynamic Product Launch and
Exit Under Demand Uncertainty.” Marketing Science 25 (1): 25–50.
“Impacts of Banning Cage Egg Production in the United States A Report Prepared for
United Egg Producers.” 2009. Promar International.
http://www.unitedegg.org/information/pdf/Promar_Study.pdf.
Jegtvig, Shereen. 2013. “Nutrition FAQ: How Much Protein Is in an Egg?” About.com
Nutrition. March 3.
http://nutrition.about.com/od/askyournutritionist/f/eggs_protein.htm.
Jeitschko, Thomas D., and Anthony Creane. 2009. “Endogenous Entry in Markets with
Adverse Selection” (June 3): 1–38.
Kijima, Masaaki, and Takashi Shibata. 2005. “Real Options in an Oligopoly Market.”
The Kyoto Economic Review 74 (1): 47–64.
doi:http://dx.doi.org/10.11179/ker.74.47.
Lee, Yikuan, and Gina Colarelli O’Connor. 2003. “New Product Launch Strategy for
Network Effects Products.” Journal of the Academy of Marketing Science 31 (3):
241–255. doi:10.1177/0092070303031003003.
Lint, Onno, and Enrico Pennings. 2001. “An Option Approach to the New Product
Development Process: a Case Study at Philips Electronics.” R&D Management 31
(2): 163–172.
Maggi, Giovanni. 1996. “Endogenous Leadership in a New Market.” RAND Journal of
Economics 27 (4): 641–659.
Murto, Pauli, and Jussi Keppo. 2002. “A Game Model of Irreversible Investment Under
Uncertainty.” International Game Theory Review 4 (02): 127–140.
doi:10.1142/S0219198902000604.

124

Narver, John C., Stanley F. Slater, and Douglas L. MacLachlan. 2004. “Responsive and
Proactive Market Orientation and New‐Product Success*.” Journal of Product
Innovation Management 21 (5): 334–347. doi:10.1111/j.0737-6782.2004.00086.x.
O’Donnell, Fiona. 2008. “Green Living - US - February 2008”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/display/295904/#.
Oberholtzer, Lydia, Catherine Greene, and Enrique Lopez. 2006. “Organic Poultry and
Eggs Capture High Price Premiums and Growing Share of Specialty Markets”.
LDP-M-150-01. Outlook Report from the Economic Research Service. United
States Department of Agriculture.
http://www.ers.usda.gov/media/866670/ldpm15001_002.pdf.
Patterson, Bill. 2011. “Eggs - US - May 2011”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/display/543044/?highlight=true#.
“Per Capita Consumption of Eggs (shell) in the U.S. 2009 | Statistic.” 2011. Accessed
April 22. http://www.statista.com/statistics/183685/per-capita-consumption-ofeggs-shell-in-the-us-since-2000/.
Poddar, Sougata, and Bibhas Saha. 2010. “Product Innovation and Stability of
Collusion.” Economics Bulletin 30 (2): 1392–1400.
Robertson, Thomas S., Jehoshua Eliashberg, and Talia Rymon. 1995. “New Product
Announcement Signals and Incumbent Reactions.” Journal of Marketing 59 (3):
1–15.
Russo, Carlo, Concetta Cardillo, and Maria Angela Perito. 2003. “Product Innovation and
Imperfect Competition in the Italian Fruit-drink Industry.” International Food and
Agribusiness Management Review 6 (1): 70–83.
Satimanon, Thasanee, and Dave D. Weatherspoon. 2010. “Hedonic Analysis of
Sustainable Food Products.” International Food and Agribusiness Management
Review 13 (4): 57–74.
Schwartz, Eduardo S., and Lenos Trigeorgis. 2004. Real Options and Investment Under
Uncertainty: Classical Readings and Recent Contributions. The MIT Press.
Shy, Oz. 1995. Industrial Organization: Theory and Applications. The MIT press.
Singh, Nirvikar, and Xavier Vives. 1984. “Price and Quantity Competition in a
Differentiated Duopoly.” RAND Journal of Economics 15 (4): 546–554.
Smit, Han TJ, and L. A. Ankum. 1993. “A Real Options and Game-theoretic Approach to
Corporate Investment Strategy Under Competition.” Financial Management 22
(3): 241–250.
125

Stokes, Jeffrey R., Rekha M. Rajagopalan, and Spiro E. Stefanou. 2008. “Investment in a
Methane Digester: An Application of Capital Budgeting and Real Options.”
Applied Economic Perspectives and Policy 30 (4): 664–676.
Sumner, Daniel A., J. Thomas Rosen-Molina, William A. Matthews, Joy A. Mench, and
Kurt R. Richter. 2008. Economic Effects of Proposed Restrictions on Egg-laying
Hen Housing in California. University of California Agricultural Issues Center.
Tauer, Loren W. 2006. “When to Get in and Out of Dairy Farming: a Real Option
Analysis.” Agricultural and Resource Economics Review 35 (2): 339–347.
Tzouramani, Irene, and Kostadinos Mattas. 2004. “Employing Real Options
Methodology to Evaluate the Organic Agriculture Scheme in Greece.” In , 1–10.
Vienna, Austria.
http://ecsocman.hse.ru/data/568/674/1219/87EAAE_Tzouramani.pdf.
United Egg Producers. 2011. “The Egg Industry and Animal Welfare A Science – Based
Approach.” Accessed March 15.
http://www.unitedegg.com/information/pdf/Egg_Industry__Animal_Welfare_Bro
chure.pdf.
United States Department of Agriculture. 2010a. “Prices Paid by First Receiver for
Organic and Conventional Eggs 2004-08.”
http://www.usda.gov/wps/portal/usda/!ut/p/c5/04_SB8K8xLLM9MSSzPy8xBz9
CP0os_gAC9-wMJ8QY0MDpxBDA09nXw9DFxcXQcAA_1wkA5kFaGuQBXeASbmnu4uBgbe5hB5AxzA0UDfzyM_N1W_IDs7zdF
RUREAZXAypA!!/dl3/d3/L2dJQSEvUUt3QS9ZQnZ3LzZfUDhNVlZMVDMx
MEJUMTBJQ01IMURERDFDUDA!/?navid=SEARCH&mode=simple&q=Price
s+paid+by+first+receiver+for+organic+and+conventional+eggs+200408&site=usda&x=28&y=4.
United States Department of Agriculture, National Nutrient Database for Standard
Reference Release 25. 2010b. “Nutrient Data for 01123, Egg, Whole, Raw,
Fresh.”
United States Department of Labor, Bureau of Labor Statistics. 2011. “Consumer Price
Index – Average Price Data, Item: Eggs, Grade A, Large, Per Doz.”
Yoon, Eunsang, and Gary L. Lilien. 1985. “New Industrial Product Performance: The
Effects of Market Characteristics and Strategy” 2 (3) (September): 134–144.
doi:http://dx.doi.org/10.1016/0737-6782(85)90033-5.
Zelman, Kathleen M. 2009. “Good Protein.” http://www.webmd.com/fitnessexercise/guide/good-protein-sources.

126

CHAPTER 4: CONCLUSION

This dissertation studies sustainability within the food system. It estimates the: 1)
value of sustainable attributes, 2) the food manufacturers’ decision making on whether to
launch a sustainable product, and 3) then simulates results from the U.S. egg industry.
The first study informs food firms about price premiums of sustainable attributes as one
of the deciding factors on launching a sustainable product. The second paper constructs a
systematic product-launching decision model and suggests under which conditions firms
decide to launch a sustainable product and/or encourage a firm to be a leader that markets
a sustainable product first.
The first paper found that the welfare-managed attributes, free-range and/or cagefree attribute were positively and significantly related to the price of eggs. The welfaremanaged attribute has a price premium equal to 3.57 cents over a standard conventional
egg. The attribute that has the highest premium was the organic attribute which increases
the price per egg by 16.50 cents. Surprisingly, the premium of the stack attributes,
organic and welfare-managed attributes together, equals only 11.26 cents per egg which
was less than the premium of the organic attribute. This implies that consumers may be
confused about the difference between organic and welfare-managed attributes and are
not willing to pay for both labeled attributes simultaneously. Consequently, in the short
run, egg manufacturers could maximize profit by offering and labeling either organic or
welfare-managed eggs until consumers perceive the difference between these two
attributes.

127

The second paper constructed a model for food manufacturers’ decision making
to launch a sustainable product by employing game theory. There are two main
conditions that influence a leader food firm to launch a sustainable product. The first
condition is that the relative cost of a sustainable product over a conventional product is
lower than or equal to the relative maximum willingness to pay for a sustainable product
over a conventional product. This implies that an additional cost to produce a sustainable
product has to be covered by an additional maximum willingness to pay for a sustainable
product to induce the leader food firms to launch a sustainable product. The second
condition is that if the first condition is not satisfied, it is profitable for the leader firm to
launch a sustainable product only when an absolute value of the maximum willingness to
pay for a sustainable product is sufficiently higher than the absolute cost of the
sustainable product.
The conditions to support a follower food firm to launch a sustainable “me-too”
product are different from a leader’s conditions. Firstly, the follower food firm decides to
launch a sustainable product when consumers realize that there is a difference between
conventional and sustainable products. In addition, consumers have to prefer and be
willing to pay more for the sustainable product. Secondly, even if the first condition does
not hold, the follower food firm will still decide to launch a sustainable product when the
costs of a sustainable and conventional product are sufficiently close.
The second study also determines under which conditions food firms are
encouraged to be a leader in the sustainable product market. A firm decides to be a leader
when it has a cost advantage to produce a sustainable product, and a cost disadvantage to
produce a conventional product. Moreover, the firm decides to be a leader when the

128

relative cost of a sustainable product over a conventional product is less than or equal to
the relative maximum willingness to pay for a sustainable product over a conventional
product which is the same condition that encourages the leader firm to launch a
sustainable product. Simulations from the egg industry show that both a leader and a
follower attain higher profits when they decide to launch a sustainable egg product. These
results are supported by observation and data from the U.S. Egg Industry. That is, from
the Mintel’s Consumer Survey Data, the percentage of consumers who buy free-range or
cage-free eggs increased from 9% in 2008 to be 14% in 2011 (Mintel, 2008 and Mintel,
2011).

129

REFERENCES

130

REFERENCES

Patterson, Bill. 2008. “Eggs - US - March 2008”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/sinatra/oxygen_academic/search_r
esults/show&/display/id=296610.
Patterson, Bill. 2011. “Eggs - US - May 2011”. Mintel.
http://academic.mintel.com.proxy1.cl.msu.edu/display/543044/?highlight=true#.

131