Download Kanniainen 01 ee08 6538261 en

Firms’ Ethics, Consumer Boycotts,
and Signalling
Amihai Glazer, Vesa Kanniainen
& Panu Poutvaara
Firms’ Ethics, Consumer Boycotts, and
Signalling ¤
Amihai Glazery
Vesa Kanniainenz
Panu Poutvaarax
February 4, 2008
Abstract
This paper develops a theory of consumer boycotts. Consumers care
both about the products they buy and about the images of the …rms producing those goods. To a¤ect a …rm’s ethical behavior, moral consumers
may refuse to buy from an unethical …rm. Consumers who do not care
about ethical behavior may join the boycott to (falsely) signal that they do
care. In the choice between an ethical or unethical behavior, it is optimal
to play a mixed strategy when the cost of it is not small or substantial.
For low and high cost of ethical investment, a pure strategy is optimal. In
particular, when the cost is low, the ethical behavior arises from prisoners’
dilemma.
Key words: …rm’s ethical code, consumer morality, boycotts
JEL classi…cation: M14, D43
1
Introduction
A …rm deviating from an accepted social norm like environmental protection,
sound personnel policy, or avoidance of child labor, may risk punishment by
consumers. Examples of such punishments are many. Shell Oil su¤ered damage
to its image from the military action of the Nigerian government against domestic protests aimed at protecting the delta of its river. Nestle su¤ered from
lost reputation after selling inappropriate milk to pregnant mothers in developing countries. The accounting …rm Arthur Andersen never recovered from
¤ Prepared for the CESifo Ethics and Economics conference, February 7-9, 2008. The initial
draft of this was presented at the NORIO VI Conference in Stockholm, June 1-2, 2007 and in
the EBEN Conference on Business Ethics, Leuven September 19-20, 2007.
y Department of Economics, University of California– Irvine, Irvine, California 92697, USA.
E-mail: [email protected]
z Department of Economics, Arkadiankatu 7 (P.O. Box 17), 00014 University of Helsinki,
Finland. E-mail: [email protected].….
x Department of Economics, Arkadiankatu 7 (P.O. Box 17), 00014 University of Helsinki,
Finland. E-mail: panu.poutvaara@helsinki.….
1
its involvement with Enron. When Turku Mustard was bought by a Swedish
…rm which moved production to Sweden, Finnish consumers boycotted their
favorite mustard, causing its share price to fall by about ten percent. In 2005,
an Estonian ship, owned by Tallink, was caught releasing waste into the Baltic
Sea. Two days after passengers organized a boycott against the …rm, the …rm
announced a change in its policy, pledging to safely release waste into containers
in Helsinki. The plan of the …rm producing the British condiment HP sauce to
move production to the Netherlands caused a consumer boycott.
A person who joins a consumer boycott is typically willing to pay a higher
price for a good produced by a …rm not boycotted. Moreover, those organizing
the boycott often want to see other consumers join.1 The internet and other
modern means of communication provide consumers with new instruments to
in‡uence the ethical behavior of other producers (Andersen, 1999).2 The consumers’ concerns can induce a …rm to devote attention to its image. Indeed, the
internet home pages of many …rms describe their work in helping development
projects and controlling environmental damage.3 4
Our paper asks how many and under what conditions consumers join a boycott. We also ask how …rms behave under the threat of a boycott. We analyze
a model in which competing …rms produce identical products but can choose
their corporate ethics, di¤erentiating their image among consumers. Consumers
observe the behavior of …rms and may avoid buying from a …rm which violated
some ethical position. Some consumers join a boycott because of their moral
outrage. Other, amoral, consumers may join a boycott because of the private
bene…t of appearing to have moral considerations. We also ask how …rms behave
under the threat of a boycott.5
We examine in particular whether a boycott can e¤ectively direct corporate ethics and how competition determines the market outcome. The existing
literature is meager. A few papers study corporate ethics but none studies consumers’ actions. In our paper the e¢ciency of a boycott is determined by the
moral reaction of consumers and the cost imposed on consumers who do not
1 The Danish …rm Arla is an interesting special case in that its boycott in the Islamic world
led to a counter boycott among the western consumers regarding the initial boycott unfair.
2 Consumers’ in‡uence can be thought to be the greatest in industries where products are
not too di¤erentiated and where competition is strong. Those features can be expected to
be measured by price elasticities. Elastic demand points to high substitutability and a low
switching cost. Research supports the proposition that consumers can in‡uence …rms, see
Morales (2005). Cronberg (1986) analyzes consumers’ in‡uence on new technology. Reasons
for why boycotts arise are studied by Klein, Mith, and John (2004). Information on how
consumers react to other aspects than the price—a …rm’s image—can apparently be based on
case studies only. Stock prices and consumer prices can there be helpful indicators.
3 Klein, John and Smith (2001 and 2004) explore the motivations for consumer boycotts.
The deviating …rms may su¤er from a boycott by its competitor, cf. Brennan (1992).
4 Yet, there may be switching costs, especially if transaction costs make consumers commit
to some products. Switching costs are analyzed by Klemperer (1995). See also Antheon,
Camarero and Carrero (2007).
5 A …rm can respond to a boycott by playing tough or weak, depending on how much it
values a good image. The reaction of a …rm may result in a particular reputation and may
in‡uence the success or failure of a future boycott. We ignore such reputation building in the
current paper.
2
join the boycott. By joining the boycott, an individual can signal - falsely - his
type in his reference network. Consumer opportunism and the option of free
riding, however, reduce the success of a boycott.
For concreteness, we shall say that a …rm behaves ethically if and only if it
invests in pollution abatement. We shall …rst consider the investment decision
as …xed for the …rm, not subject to its choice. Such an approach is useful in
considering the e¤ects of di¤erent regulations in di¤erent jurisdictions. Thus,
we can consider competition between a …rm in a country which imposes strict
environmental regulations and a …rm in a country with lax regulations. In the
absence of consumer reactions, the …rm subject to lax requirements will earn
higher pro…ts. But if some consumers prefer to buy from a non-polluting …rm,
then the opposite may hold. Indeed, consumer behavior may lead countries to
engage in a “race to the top," with one country not wanting to appear to have
laxer requirements than another.
What makes the problem non-trivial is the joint consideration of consumer
and …rm behavior. The price a …rm charges will depend on how much consumers
are willing to buy from a non-polluting …rm. And consumers in turn will decide
what to buy on the basis of both pollution activity and the prices the …rms
charge.
Later in the paper we will consider behavior of a …rm when it is not subject
to regulation, but can instead choose whether to invest in pollution abatement.
Of interest is whether it would do so.
2
2.1
Assumptions
Consumers and ethical preferences
We cast the analysis in terms of a duopoly market where two …rms compete for
customers. The products (or services) are physically identical but the production processes can di¤er. There is, say, a polluting one and a non-polluting one.
People di¤er with respect to their morality in respect of pollution and some may
…nd it valuable to misrepresent their hidden preferences. There are two types
of consumers. An -type has moral preferences; a -type does not.6 The mass
of -type consumers is , while the mass of -type consumers is scaled to 1.7
Each consumer buys at most one unit of the good. In both groups, consumers
are indexed in decreasing order by [0 ] or [0 1] with respect to their basic
willingness to pay for the product. Consumer = 0 has the highest basic willingness to pay for the product, say in both groups; consumers = and = 1
have zero willingness to pay for it in the two groups. The willingness to pay
6 The origin of ethical preference lies beyond the scope of the current paper. A natural
source is that the preferences are created by evolutionary mechanisms within human beings
as part of a social contract, cf. Binmore (1998). It is appropriate to think that the ability to
commit to a social norm, the option to participate in a boycott, develops like a meme. The
idea of a social gene, “meme" in contrast to the biological gene, was introduced by Dawkins
(1976).
7 The society hence consists of individuals of homo moralis and homo oeconomicus.
3
by the remaining consumers is uniformly distributed on (0 ) in both groups.
To illustrate, and ignoring moral and reputational e¤ects for the moment, the
utility from consumption by consumer (in any group) is given by an indirect
utility function
= (1 ¡ ) ¡ where is the market price. In the absence
of ethical issues, the goods produced by di¤erent …rms are identical.
In our model, the products are not identical if one …rm pollutes while the
other does not. In the social context, a reputation for moral concerns has bene…ts for individuals. Absence of such reputation may lead to exclusion from
particular groups, loss of friendship, and even barriers in the marriage market.
We simplify matters by assuming that a consumer who participates in a boycott
gets a bene…t of
0 compared to a consumer who does not. This assumption
can approximate the situation when the bene…ts of having a reputation for ethical concerns is non-linear, with the bene…t increasing rapidly as the probability
that a person has moral concerns rises above zero. In our model, however, we
make the simpli…cation that only the less moral consumers enjoy this bene…t;
the moral ones are committed to boycott anyway.
We introduce a further e¤ect. We let
0 denote the cost imposed on a
consumer who does not join the boycott. This can be thought of as a social
pressure, commonly observed.
2.2
Firms
There are two periods. In period 1, each …rm decides on its code of ethical
conduct; it invests or not say, in pollution abatement. The cost of abatement can
di¤er across …rms. We consider the cases where mixed strategies or alternatively
pure strategies are optimal. In period 2, each …rm’s ethical conduct becomes
common knowledge, and each consumer decides at which …rm to buy.
3
Equilibria when the Ethical Codes Di¤er
We shall consider three di¤erent sets of investment: no …rm invests, both …rms
invest, or only one does. As we eventually have to …nd out the outcome of
the investment game under various strategies, it is most illuminating to start
with the case of two pure strategies. In this section therefore, we consider the
case where one of the …rms invests, thereby incurring a …xed cost
0, while
the other does not invest. In the market, the products of the …rms, though
perfect substitutes in consumption, are now di¤erent with di¤erent images of
their producers. Some consumers will then switch from buying the product of
the …rm with no ethical code, say , to buying from the …rm with the ethical
code, say . More speci…cally, the boycotting high-moral consumers abstain
from buying at …rm and buy only at …rm . Since in equilibrium not all may
buy, we denote the active high-moral buyers by . The number of low-moral
consumers who stay at …rm is denoted by
Some of the low types, however,
switch to the
…rm in order to (falsely) signal high morality. Their number is
denoted by
.
4
Equilibrium We denote the resulting price at the ethical …rm by
and the
price at the non-ethical …rm by . The resulting market equilibrium has the
following structure. From the de…nition of the marginal moral consumer , we
know that the equilibrium price at …rm
satis…es
(1 ¡
8
)=
(1)
We can of course have a market equilibrium where no low-moral consumer
switches to …rm . However, to make the analysis interesting, we assume that
the bene…t from signalling is su¢cient so that some, i.e.
do. The marginal
low-moral consumer must be indi¤erent between the two markets. Thus, prices
must satisfy
(1 ¡
)+ ¡
= (1 ¡
)¡ ¡
(2)
where, to recall is the signalling bene…t when the consumer buys at …rm
and is the social pressure when he does not. Therefore,
Lemma 1. The price di¤erence arises from the signalling bene…t and from
the cost of social pressure,
¡
9
= +
(3)
The marginal low-moral consumer (with an index
+ ) is indi¤erent
between buying at …rm or buying nothing. Thus, his net utility is
(1 ¡
¡
(4)
)¡ =
In the Cournot model, …rms decide on their outputs allowing the prices to
adjust. We denote the outputs by
and
Then,
=
+
(5)
=
Thus,
¡
=
To solve for the prices, we …rst …nd the number of signalling consumers.
Using
¡
= +
(1 ¡
) ¡ (1 ¡
+
This gives
=
¡
1+
¡
+
8 It
)+ = +
(1 + )
(6)
will thus not be the case that all moral consumers buy. They buy only if
= making
= 0 Otherwise,
9 We notice that actually all low-moral types are indi¤erent between the two markets as
the signalling bene…t and the social cost of pressure just match the price di¤erence. The …rms
plan their output knowing the consumers’ behavior.
5
Then the pro…ts of the …rms are
=
¡
=
( ¡
+
1+
·
+
1+
=
=
¡
+
¡
=
¡
1+
¡
1+
¡
1+
From the …rst-order conditions,
µ
¶
=
1¡
+
1+
1+
(7)
)¡
1+
¡2
1+
¡ ¡2
1+
¸
(8)
Solving for the Nash-Cournot equilibrium, the …rst FOC gives:
2
1+
=
=
µ
1¡
1+
2
1+
¡
2
¶
+
+
1+
2
We obtain:
=
=
Proposition 1.
shares
(1 + ) (1 + 2 ) + 2(1 + )
¡
3
3
(1 + ) (2 + ) + (1 + )
+
3
3
(9)
(10)
The Cournot-Nash equilibrium is given by the market
=
=
(1 + ) (1 + 2 ) + 2(1 + )
¡
3
3
(1 + ) (2 + ) + (1 + )
+
3
3
The ethical …rm bene…ts from the incentive of the less moral consumers
who mimic the moral ones. Similarly, social pressure on the amoral consumers
bene…ts this …rm; the unethical …rm su¤ers from a loss of customers.
Noticing that the signalling bene…t or the social pressure may di¤er between
products, we state
6
Corollary 1. The economic e¤ect of a boycott is smaller when the boycott
is of products which are less useful for signalling reasons, or where the social
punishment is low.
Solving next for the number of mimicking customers,
=
=
¡
+
1+
(1 + )
1¡
(2 + 5 + 2 2 )
(1 + 2 )
+
+
3
3 (1 + )
3
(11)
Lemma 2 The number of mimicking customers is determined by the signalling bene…t and the cost of social pressure relative to the basic willingness to
pay by the consumers.
Thus, a consumer boycott by moral (non-opportunistic) people induces some
opportunistic amoral people to take advantage of the signalling bene…t, making
the consumer market in the aggregate behave more morally. The smaller is the
signalling bene…t relative to the basic willingness to pay, the less e¤ective is the
boycott.
We solve the prices starting from non-moral consumers. It holds that
=
=
(1 ¡
¡ )¡
(1 + 2 )
2
¡
¡
3
3(1 + )
3
and
=
=
+ +
( + )(1 + ) + (2 + )
3(1 + )
(12)
and rewriting
=
( ¡ 2 )(1 + ) ¡ (1 + 2 )
3(1 + )
The equilibrium studied above is characterized by pro…ts
we calculate next
=
=
¡
(( + )(1 + ) + (2 + ) )2
¡
9(1 + )
(13)
¡
¢
which
(14)
=
=
(( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
7
(15)
4
Firm’s Ethical Decisions: To Be or Not To
Be?
4.1
Mixed strategy equilibria
To examine the conditions for the mixed strategy, call the …rms
and
Assume that the investment is not observable when undertaken but that in
the production stage, consumers observe whether a …rm invested in the ethical
action. Now, the …rms can for competitive reasons randomize their investments
and we start by studying the mixed strategy equilibrium. We also work out
whether and when a pure strategy can be optimal.
We let
denote the probabilities of undertaking the investment. The
expected pro…ts then are
£
¤
£
¤
[ ]=
+ (1 ¡ )
+ (1 ¡ )
+ (1 ¡ )
(16)
[
]=
£
+ (1 ¡
¤
)
+ (1 ¡
)
£
+ (1 ¡
)
¤
(17)
We notice that the expected pro…ts are linear in the probabilities. A (mixed)
Nash equilibrium in terms of the optimal probabilities satis…es
[
]
[
=0
The …rst-order conditions are
=
=
¡
1¡
=
=
¡
1¡
¡
¡
1
(
(
¡
¡
¡
¡
1
(
(
¡
¡
¡
¡
]
¡
¢ ¡
¡
=0
¢
¡
)
)
¡
¢ ¡
¡
)
)
¢
¡
¢
(18)
¢
(19)
These conditions represent a Nash equilibrium without a dominating strategy. An interior solution, 0
1 0
1 it is required that
¡
¢
¡
¢
¡
¡
¡
¡
¢
¢
0
0
¡
¡
Therefore, before addressing these conditions, we need to study the pure
strategies to …nd out the pro…t levels.
8
4.2
Pure Strategies
In this section, we allow for the costs of investment,
4.2.1
to di¤er.
Neither Firm Acts Ethically
It is also possible that one of strategies is a dominating one for both …rms.
When it is optimal to choose
=
= 0 no investing represents the dominant
strategy. Now one group of customers - boycotters - leaves the market, i.e. the
market size is squeezed to 1, there is one price and the …rms share the customers
on an equal basis.
The pro…ts are then
=
=
10
Necessary conditions
for no investing representing a dominant strategy are
To solve, the marginal customer has zero utility
(1 ¡
¡
)¡ ¡
=0
Pro…ts are
= [ (1 ¡
= [ (1 ¡
¡
¡
)¡ ]
)¡ ]
Solving for the Nash-Cournot equilibrium,
=
1 ¡
(
)
3
1 ¡
= (
)
3
(20)
The price can be solved as
1
= ( ¡ )
3
Pro…ts in the no investment equilibrium are
=
=
1 ( ¡ )2
9
(21)
1 0 These conditions are not, however, su¢cient, as a prisoners’ dilemma to be studied below arises under these same conditions (strengthened by some others) with
=
= 1
representing the dominating strategy.
9
4.2.2
Both Firms Invest
When it is optimal to choose
strategy. Pro…ts are then
= 1 ethical behavior is a dominant
=
¡
=
¡
=
The outcome is a symmetric Cournot equilibrium. This equilibrium can
appear if the cost saving is small.11 Necessary conditions for investment
representing a dominant strategy are
To solve for the market shares, we notice …rst that the total mass of potential
customers in the market now is 1 + and the two …rms share them. The …rms
now face a less steep market demand as the mass of potential customers is
increased. Denoting the total amount of buying customer
which also is
the index of the last buyer, it must hold that he is indi¤erent between buying
and not buying. As his willingness to pay has to match with the market price,
it must hold,
(1 ¡
)=
(22)
1+
As the market is shared, we have that
=
+
The pro…ts are
=
(1 ¡
=
(1 ¡
+
1+
+
1+
)
¡
)
¡
The …rst-order conditions are
= (1 ¡
=
+
1+
)¡
1+
3
=
1
1+
=0
(23)
The pro…ts are
=
(1 ¡
=
(1 + )
9
+
1+
¡
)
µ
1+
3
¶
¡
(24)
1 1 We show below that this is not the only case where investing is the optimal strategy. It
can arise as a prisoners’ dilemma.
10
5
When is the mixed strategy optimal?
We …rst collect the above …ndings under pure strategies
=
(( + )(1 + ) + (2 + ) )2
¡
9(1 + )
=
(( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
=
(1 + )
¡
9
=
1 ( ¡ )2
9
Repeating, to have the mixed strategy appear optimal, we must have
[
]
[
=0
1, we must have that12
¡
¡
¡
¡
Now, for
¢
¢
]
=0
0
This condition holds in two exclusive cases.
(i) Case 1
0 &
¡
Evaluating the …rst condition,
¡
=
(1 + )
¡
9
(1 + )2
2
¡
¡
0
(( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
0
¡ (( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
Evaluating the second condition,
¡
1 2 We
=
(( + )(1 + ) + (2 + ) )2
1 ( ¡ )2
¡
¡
9(1 + )
9
2
2
(( + )(1 + ) + (2 + ) )
1( ¡ )
¡
9(1 + )
9
=
notice that when thse conditions are satis…ed, it also always holds
11
0
Combining,
(( + )(1 + ) + (2 + ) )2 1 ( ¡ )2
¡
9(1 + )
9
(1 + )2
2
¡ (( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
This can never hold because the value of the left-hand side exceeds the value
of the right-hand side.13
(ii) Case 2
0 &
¡
¡
0
These conditions amount to stating
(1 + )2
2
¡ (( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
(( + )(1 + ) + (2 + ) )2 1 ( ¡ )2
¡
9(1 + )
9
We have proved:
Proposition 2. The necessary and su¢cient condition for the existence of
a mixed strategy as an optimal choice for a …rm is that the cost of investment
of its rival ( ) satis…es the above conditions.
Corollary 2: Pure strategies can be played only if
2
(1+ )2 2 ¡(( ¡2 )(1+ )¡(2+ ) )2
(( + )(1+ )+(2+ ) )2
or if
¡ 19 ( ¡ )
9(1+ )
9(1+ )
As the mixed strategies are not always optimal, it does make sense to study
also the pure strategies i.e. when
does not satisfy any of the above two
conditions.
Pure strategies in the limit Consider the optimal strategies when
ap2 2
¡2 )(1+ )¡(2+ ) )2
proaches its limits. In the limiting cases, when
! (1+ ) ¡(( 9(1+
)
¡
¢
)+(2+ ) )2
we have
¡
= 0 Similarly, when
!¡ (( + )(1+
¡
9(1+ )
¡
¢
2
(
¡
)
1
= 0 Linking with the expression for
we
we have
¡
9
have
lim
!
= 1 lim
!
(25)
=0
Similarly for
We conclude that the …rms do not always choose to play a mixed strategy.
For a mixed strategy for being optimal it is necessary that the conditions (24)
1 3 Strictly
speaking, this conclusion is based on the case with
12
= 1.
are satis…ed. These conditions link the value of consumer signalling with the
…rms’ cost of being ethical.
Proposition 3. Ethical behavior can arise an optimal strategy but only in
conditions where the cost of investment is not high.
6
Ethical behavior arising as the prisoners’ dilemma:
low pro…t equilibrium.
An argument put forward by Shleifer (2001) is that competition is detrimental to
high corporate ethics. Our analysis challenges his view. We have arrived at this
view by examining the case whether there exists a combination of pure strategies
which satis…es the conditions for the prisoners’ dilemma. This amounts to asking
whether, in the absence of commitment, a low-pro…t equilibrium with both …rms
investing replaces a joint pro…t maximization where neither invests. It amounts
thus to asking whether both …rms would generate more pro…ts by not investing
but in the face of a rival’s move of investment, it is optimal to follow the lead.
In the current model, such a harsh requirement appears reasonable as otherwise
the …rms lose all the boycotting customers which leave the market.
Intuitively, the equilibrium depicting a prisoners’ dilemma can be characterized by the conditions
(26)
The condition
indicates that if invests it can increase its pro…t
subject to the condition that
does not invest. The condition
indicates that in such a case can lose a lot.
To state in terms of the optimal strategies, the strategy pair
arising as
prisoners’ dilemma requires that
=
= 1 I.e. it should be optimal to have
the
-equilibrium when the …rms optimize individually,
[
Evaluating, say
¡
¡
[
]
+
]
[
0
]
(27)
0
0 gives
+
¡
¡
¢¢
¡
Given that the incentive to deviate from the
¡outcome exists for both
…rms,
the right-hand side is positive. It is then su¢cient for this
inequality to hold that ¡
+
0 which amounts to having
representing one of the characterizations of the prisoners’ dilemma. We now
examine what conditions are required for the equilibrium to be characterized as
prisoners’ dilemma given the structure of our model.
13
First, when a …rm, say switches individually from the strategy pair
, the other playing
the pro…t of …rm
increases,
(( + )(1 + ) + (2 + ) )2
¡
9(1 + )
Second, the pro…t of the …rm, say
to
1 ( ¡ )2
9
is reduced if it does not follow, i.e.
(( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
1 ( ¡ )2
9
Third, the pro…ts are lower under the strategy pair
than under the
strategy pair
(1 + )
1 ( ¡ )2
¡
9
9
Fourth, the pro…ts cannot be negative when the strategy pair
is chosen,
(1 + )
9
Multiplying the second condition by (1 + )
(( ¡ 2 )(1 + ) ¡ (2 + ) )2
( ¡ )2 (1 + )
one can judge that this (quite surely) is satis…ed. Consider the rest of the
conditions to be combined as
min
µ
(1 + )
9
¡
1 ( ¡ )2
9
(( + )(1 + ) + (2 + ) )2 1 ( ¡ )2 (1 + )
¡
9(1 + )
9
9
¶
It remains to produce a numerical example as to when these conditions can
hold. Take = 1 Moreover, it makes sense to examine the limiting case ! 0.
The lower limit for then becomes
1
(2 ¡ )
9
For example, having then
= 0 25 one obtains
0 0486
Now the upper limit.
µ
(1 + + 2 )2 1
1
min
¡ (1 ¡ )2
9
9
9
With
= 0 25
= 0 25 one has
min(2 50 0 11)
= 0 11
14
¶
Thus, when the investment satis…es
0 0486
0 11
we can have that the
strategies arises from the prisoners’ dilemma - provided that it is not optimal to play a mixed strategy studied earlier. From there
we know that a mixed strategy cannot be optimal if
(1 + )2
2
¡ (( ¡ 2 )(1 + ) ¡ (2 + ) )2
9(1 + )
With the numerical values above,
1 ¡ ((1 ¡ 0 5) ¡ 2 )2
1
=
9
9
= 0 11
Indeed, when 0 0486
0 11 the equilibrium arises from a prisoners’
dilemma. This conclusion was strictly obtained by having the number of moral
consumers in the market
= 0 By continuity, it must hold also when is
positive but su¢ciently small.
Would it hold that
0 0486 the
equilibrium would arise as a dominating pure strategy. We can state the conclusion as a proposition:
Proposition 4: When the number of moral consumers is small and when
the investment cost is small but strictly positive, …rms may end up undertaking
the ethical investment even if their pro…ts are reduced. The equilibrium then
arises from a prisoners’ dilemma.
For completeness, it is worth stating that when the investment cost is high,
the
becomes the dominating strategy.
7
Final Remarks
Modern communication media, including the internet, enhance the opportunities for consumers to in‡uence the ethical behavior of producers. This increased in‡uence has induced …rms to devote substantial e¤ort on their image
among consumers. Firms often highlight their contributions to economic development or to environmental quality. Consumer power apparently has been
increasing and will continue to increase though we only have indirect evidence
on that. Consequently, with increasing concern about environmental issues like
the greenhouse e¤ect, one can expect that we will see an increased incentive for
individuals to organize boycotts in the future. In the end, there may be fewer,
however, if such a threat leads …rms to behave better.
Are there two few boycotts from the social point of view? Do markets produce the right amount of boycotts? This is a challenging welfare issue. In
an individualistic society with utilitarian preferences, the well-being of (all) citizens are often taken as the starting point for the evaluation of the social welfare.
Boycotts enhance the market position of well-behaving …rms and the consumer
15
surplus of their customers. There are, however, quite a few welfare aspects. The
negative externalities, say pollution, is reduced - we have not modelled such an
externality explicitly. The moral individuals who abstain from buying from the
unethical …rm do bene…t, though we abstracted from introducing this mechanism in our formal model if only to simplify wherever possible. By implication,
the immoral consumers who for the reasons of opportunism switch the …rm tend
to cause a positive externality on the moral ones. Such a positive externality
tends, however, to be diluted if the private return on signalling a moral characteristic su¤ers from the number of boycotters. Organizing a boycott, however,
can be socially costly as it represents a reduction in use of resources in a productive activity. People who bear these costs may not be the same people who
bene…t from boycotts as some are free-riders and can take opportunistically the
advantage of lower prices of the products boycotted. Those less moral individuals who buy from the unethical …rm su¤er a cost as they de…nitively are now
separated from the moral ones.
References
[1] Andersen, T., (1999) "Consumer Power via the Internet".
[2] Antheon, C., Camarero, C. and Carrero, M., (2007), "Analyzing …rms’
failures as determinants of consumer switching intentions: The e¤ect of
moderating factors", European Journal of Marketing, 41, 135-158.
[3] Baron, D.P., (2002) “Private Politics and Private Policy: A Theory of
Boycotts." Stanford Graduate School of Business.
[4] Binmore, K., (1998), Game Theory and the Social Contract II. Just playing,
MIT Press.
[5] Brennan, T.J., (1992) "Refusing to Cooperate with Competitors: A Theory
of Boycotts", Journal of Law and Economics, 35, 247-264.
[6] Cronberg, T., (1986) "Consumer in‡uence on new technology - Real potential or false hopes?", Journal of Consumer Policy, 9, 335-343.
[7] Dawkins, R., (1976) The Sel…sh Gene.
[8] Diamond, J., (2005), Collapse. How Societies Choose to Fail or Survive,
Penguin Books.
[9] Florida, Richard (1996) “Lean and green: The move to environmentally
conscious manufacturing." California Management Review, 39: 80-105.
[10] Folkes, V.S., (1999) "E¤ects of Information About Firm’s Ethical and Unethical Actions on Consumers’ Attitudes", Journal of Consumer Psychology, 8, 243-259.
16
[11] Innes, R., A. “A Theory of Consumer Boycotts under Symmetric Information and Imperfect Competition." University of Arizona.
[12] Kanniainen, V., and Pietarila, E., (2006) “Corporate Social Responsibility:
Can Markets Control?" Homo Oeconomicus, 23: 153-179.
[13] Klein, J.K., Smith, N.C., and John, A., (2001) "Exploring Motivations for
Participation in a Consumer Boycott", London Business School, Centre for
Marketing Working Paper No. 01-701.
[14] Klein, J.K., Smith, N.C., and John, A., (2004) "Why We Boycott: Consumer Motivations for Boycott Participation", Journal of Marketing, 68,
92-109.
[15] Klemperer, P., (1995), "Competition when Consumers have Switching
Costs: An Overview with Applications to Industrial Organization, Macroeconomics, and International Trade." Review of Economic Studies 62, 515539.
[16] Morales, A.C., (2005), "Giving Firms and "E" for E¤ort: Consumer Responses to High-E¤ort Firms." Journal of Consumer Research, 312: 806812.
[17] Shleifer, Andrei, (2001), "Does Competition Destroy Ethical behavior?",
NBER Working Paper No: 10269.
17