See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/49399637

Lender Liability in the Consumer Credit Market

Article · January 2000

Source: OAI

CITATION READS

1 96

2 authors, including:

Elisabetta Iossa
University of Rome Tor Vergata
90 PUBLICATIONS   2,292 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Steering versus Crowding Out Effects: A Theory of Demand-side Innovation Policies View project

All content following this page was uploaded by Elisabetta Iossa on 28 May 2014.

The user has requested enhancement of the downloaded file.

 Lender Liability in the Consumer Credit Market∗

Elisabetta Iossa and Giuliana Palumbo†

10 January 2002

Abstract
In many countries consumer credit legislation provides for the ex-
tension of liability for product failure to the financial institution that
advances credit to the consumer. In particular, lender liability is im-
posed on those credit grantors who closely operate with the supplier
of the good.
This paper provides a rationale for lender-responsibility in the con-
sumer credit market. It shows that, when judicial enforcement is in-
efficient or there is risk of seller liquidation, lender-liability helps to
protect consumers who systematically underestimate the probability
of product failure and overestimate the extent to which they can ob-
tain compensation.

Keywords: consumer credit, lender liability, misperception, product

failure.
J.E.L. classification: D18, G28, K13.

∗
For helpful comments, we wish to thank Giuseppe Bertola, John Bennett, Philip Davis,
Marco Pagano, Bruno Parigi, Jean Charles Rochet, seminar participants at Birbeck Col-
lege, Brunel University, Ente Luigi Einaudi, European University Institute, London School
of Economics, University of Helsinki, University of Salerno and conference participants at
EEA 2000 and RES 2001.
Financial support from Findomestic S.p.A and Cetelem and Brunel University (Brief
Award) is gratefully acknowledged.
†
Respectively, Brunel University, Getulio Vargas Foundation and European Univer-
sity Institute (Finance and Consumption Chair). Corresponding author: Elisabetta
Iossa, Brunel University, Department of Economics and Finance, Uxbridge Middlesex
UB8 3PH, UK. Tel: ++44.1895.203161. Fax: ++44.1895.203384. E-mail: Elisa-
[email protected].

1
1 Introduction
When a consumer purchases a good on credit she enters two contractual re-
lationships: the sale or main contract with the supplier of the good and the
credit contract with the grantor of credit. An issue which has been highly
debated is whether the obligations assumed by the parties under each con-
tract should be independent of each other or rather be connected by virtue
of the link between the purchase of the good and its financing. In particular,
in the event of default or incomplete performance of the sale contract, should
the consumer keep fulfilling her obligations with the lender and address her
claim exclusively against the supplier or should she be entitled to a similar
claim against the lender?
A brief examination of regulations across the United States and the Eu-
ropean Union Member States shows that the legislator has been favorable to
the extension of the liability to the lender in all situations where the credit
is advanced under an agreement between the supplier of the good and the
credit grantor. This principle was first introduced in the United Kingdom
by the Consumer Credit Act 1974. The British example was then followed
by other countries and its principles appear in the Federal Trade Commis-
sion Holder Rule (196) of the United States and in the European Directive
EEC/102/87.1 Notice that an “agreement” arises whenever a supplier ar-
ranges a loan for a customer and this can also include the case where the
customer buys the good using a credit card.2 In these situations, the finance
company shares liability with the supplier. Instead, if a customer obtains
credit independently of the supplier - through her own bank for example -
the credit grantor does not bear any liability. Nor is a credit card company
liable if the customer uses her credit card to obtain cash and pay for her
purchases.
The main motivation for extending liability to the lender is consumer
protection. As emphasized by the British legislator, “where a transaction
involves a connected loan, it would be unfair and insufficient for the con-
sumer as debtor to have remedy only against the supplier, his obligation to
continue repaying the lender remaining unaffected”. The keys to consumer
protection are essentially two. One is the possibility of approaching the
1
The European Directive introduces subsidiary liability when the credit is provided on
the basis of a pre-existing agreement where credit is made available “exclusively” by that
grantor of credit to customers of that supplier. However, in most countries, it is common
jurisprudence to recognize lender-responsibility also for non exclusive agreements.
2
The regulation of credit cards agreements varies across countries. In UK for example,
most credit cards fall within the joint responsibility regimes.

2
lender when the seller is unable or refuses to satisfy the consumer’s claim.
The other is the accordance to the consumer of the right to suspend the
repayment of the loan until the good object of the contract has been deliv-
ered. However, the effective “protection power” of this measure cannot be
assessed without simultaneously considering the behavior of all the parties
acting in the market. In this regard, one argument adduced against joint
liability is that the higher costs inflicted on creditors will be transferred to
consumers, in one of the following two forms. Either credit grantors will
refuse to sign agreements with the suppliers in order to avoid the liability
or they will charge higher interest rates to restore their profitability.
In this paper we investigate these issues and provide a rationale for
the existence of voluntary agreements where the credit grantor accepts co-
responsibility for inadequate performance of the seller. Further, we show
that the present legislation has positive effects on consumer welfare when
consumers are poorly informed about risk.
We consider a simple economy composed of two markets: a monopolistic
good market and a perfectly competitive credit market. Consumers are risk
neutral and decide whether to buy one unit of the good; the good is defective
(or more generally, there is lack of conformity with the sale contract) with
exogenous and positive probability and may cause damage. Depending on
their initial endowment of wealth, two classes of consumers are identified:
poor consumers, who must borrow money to finance their purchases and
rich consumers who can buy for cash if they wish. Following a strand of
literature dating back to Spence (1977) and related empirical evidence (see
e.g. Eisenberg 1995), we assume that, within each class, consumers are
affected by misperception, and in particular, they overestimate the expected
value of the good they are purchasing.3 In our setting optimism results
from the combination of two effects: consumers both underestimate the risk
of product failure and overestimate the extent to which they can obtain
compensation. The latter in turn may be due to misinformation about the
effectiveness of the judicial enforcement and/or the likelihood that the seller
is still in business at the time the consumer addresses her claim.4
We argue that joint liability helps mitigate the loss that consumers suffer
because of their misperceptions. Crucial to this result is that the protection
granted by joint liability cannot be fully transferred by the seller into higher
prices (or interest rates). This is because, as a direct consequence of their
3
Recent developments in marketing techniques such as internet sales, door-to-door
selling and distance selling often raise a concern of competition authorities.
4
It has been estimated that in England 95% by volume of the claims under Section 75
of the Consumer Credit Act arises from the seller going out of business.

3
being optimistic, consumers undervalue the benefits they obtain from joint
liability.
Joint liability protects consumers in two ways. First, they have the right
not to repay their debt to the lender whenever the seller does not fulfill his
obligations. This reduces the loss that consumers suffer from misperceiving
both the probability of product failure and the compensation they can obtain
through the judicial system. Second, they are entitled to obtain remedies
against the lender in case the seller goes out of business. This protects
consumers against their underestimating the risk of seller bankruptcy. In
both cases consumer welfare is enhanced because the price they are willing
to pay for the protection is lower than its effective value.
That joint liability reduces the rent that can be extracted out of the
misperception of consumers implies that joint liability per se would never
be voluntary undertaken. However, it is undertaken when it is a condition
sine qua non for coordination of price and interest decisions which allows the
seller to price discriminate between rich and poor consumers. This argument
provides a rationale for restricting joint liability to those situations involving
an agreement between the seller and the lender, as is the case under the
existing legislation. In this regard, joint-liability can be viewed as a means
to redistribute some of the gains from coordination towards the consumers.
To our knowledge the desirability of extending the liability for product
failure to the lender is an issue as yet still unexplored. On the one hand,
lender liability has mainly been analyzed in the field of environmental regu-
lation, where banks may be considered liable for the environmental damage
created by the firms they finance. (See for example, Pitchford 1995, and
Boyer and Laffont 1997). On the other hand, the literature on liability for
product failure mainly restricts attention to seller-only responsibility regula-
tion and focuses on the effects of different liability rules on risk distribution
and on the incentives of the buyer to exert product care and of the seller
to provide quality.5 In contrast, our paper studies the effects of voluntary
provision of lender and seller liability and it abstracts from risk and moral
hazard considerations. Moreover, none of these papers allows for any sort
of consumers’ misperception. In this respect, our paper is mostly related to
a strand of literature dating back to Spence (1977), who consider the loss
consumers may suffer when they underestimate the probability of product
failure.6
5
See for example, Green (1976), Spence (1977), Epple and Raviv (1978), Polinsky and
Rogerson (1983) and Daughety and Reinganum (1995).
6
See also Shapiro (1982) and Polinsky and Rogerson (1983).

4
 The rest of the paper is organized as follows. Section 2 outlines the
model. Section 3 presents two relevant benchmarks: the case of fully rational
consumers (Section 3.1) and the case where there is consumers’ mispercep-
tion, but the seller and lender are free to coordinate their decisions without
this implying the undertaking of joint responsibility (Section 3.2). Section
4 studies the effects of introducing joint liability, while Section 5 discusses
other means of protecting consumers from their misperceptions. It shows
that full protection from misperception requires liability rules that make the
utility in the bad state higher than that in the good state. It further shows
that joint liability may work as a device that induces the seller to give up
the rent he can make on the irrationality of consumers in order to be able
to serve also rational (and pessimistic) consumers. Section 6 concludes.

2 The model
The economy consists of a monopolistic good market and a perfectly com-
petitive financial market. Consumers may decide to buy a unit good from
the seller and have access to credit from the financial market. The good is
worth B to each of them. However, with exogenous probability d, where
0 ≤ d ≤ 1, the sale contract is inadequately honored: the good is defective
or not delivered or it is not in conformity with the standards specified in the
initial contract. In this case the value of the good to the consumers is zero
if there is no damage, and −D if damage occurs.
In most of the paper, we take a positive approach to liability legislation
and assume that the extent of the seller liability is equal to D + B. Further,
we assume that resorting to the legal system is costly (due to legal fees or to
non-monetary unrecoverable costs of the time wasted) or, analogously, that
the judicial enforcement is inefficient (lengthy trials, incompetent judges).
Finally, we consider the possibility that after the good is purchased and
before the consumer is able to go to court, the seller goes bankrupt. We
account for the inefficiency of the judicial system by assuming that if the
seller does not fulfil his obligations and the consumer goes to court, she can
recover only a fraction α of her claim, where 0 ≤ α ≤ 1.7 Further, we denote
by ν with 0 ≤ ν ≤ 1 the probability that the seller is still in business at the
time the consumer lays her claim.
There are two classes of consumers, rich and poor; all of them are risk
7
α could also be interpreted as the probability of winning the lawsuit. Note, that all
that is required for our analysis to hold is that the consumer would always prefer not to
have to go to court in order to have her claim satisfied.

5
neutral. Rich consumers have sufficient money to purchase the good for
cash if they wish; alternatively they may purchase on credit. Poor con-
sumers have no cash, and, therefore, if they are to purchase the good, they
must use credit. The number of rich and poor consumers is given by nr and
np , respectively. In most of the paper we focus on consumers who are inher-
ently optimistic for they systematically overestimate the value of the good.
The assumption of optimism has been a standard feature of the literature
on consumers’ behavior (see e.g. Spence 1977, Polinsky and Rogerson 1983,
Shapiro 1982) and is motivated by the existence of significant empirical evi-
dence in favor of consumer optimism (see e.g. Eisenberg 1995 and references
therein). In our setting, optimism results from consumers underestimating
both the risk of product failure and of seller bankruptcy and overestimating
the compensation they can obtain in court when the seller refuses or is un-
able to honor the contract. Formally, let d,b bν and αb denote the consumers’
imprecise estimates of the probability of product failure (d), the risk of seller
bankruptcy (1 − ν) and the inefficiency of the judicial enforcement (1 − α).
We assume that db ≤ d, b ν ≥ ν and α b ≥ α, with dα ≥ dbbα.8
The supplier of the good and the credit grantor may operate indepen-
dently or may decide to sign an agreement in order to coordinate their price
and interest rate decisions. Let R ∈ {r, i} denote the interest rate on credit,
where r and i are the interest rate under independence and under coordi-
nation, respectively. Thus the effective price, denote by P paid by the
consumers is equal to p if the consumer buys for cash and to (1 + R)p if the
consumer buys on credit.
When the seller and the lender act independently, a regime of seller
liability applies, where the discovery of a defective good does not affect the
consumers’ rights and duties with the lender. In this case, the real surplus
derived from the consumption of the good is
U S (P ) = (1 − d)(B − P ) + d(−D − P + να (D + B))
= (1 − d (1 − να))B − P − d (1 − να) D (1)
Notice that, under seller-only responsibility, when the good is defective, the
consumer can pursue her remedy only against the seller. Hence, she can
recoup the damage and a new unit of the good (or its money equivalent)
only if the latter is still in business.9 Moreover, due to the inefficiency of
8
We shall return on this assumption in Section 4.2 .
9
That consumers receive a new unit of the good (or its money equivalent) when the
seller is still active in the market involves no loss of generality. We could as well have
considered the case where the consumers only recoup the price paid and the compensation
for the damage.

6
the judicial system, only a portion α of her claim will be satisfied.
However, consumers are affected by misperception. Their perceived util-
ity is then given by

U b −b
b S (P ) = (1 − d(1 να b − νbα
b ))B − P − d(1 b )D (2)

When the seller and the lender coordinate their decisions, the latter be-
comes co-responsible with the seller towards those consumers to whom he
has supplied credit. This joint liability has a twofold effect. First, the con-
sumer pays P = (1 + i)p only if the good is not defective or her claim is
satisfied. Second, if the good is defective she can resort to the lender and
obtain a fraction α of the damage inflicted to herself, if the seller has gone
into liquidation.10 Thus, under joint liability, the real utility is

U J (P ) = (1 − d)(B − P ) + d(−D + να (D + B − P ) + (1 − ν)αD)

= (1 − d (1 − να)) (B − P ) − d (1 − α) D (3)

The perceived utility is

b J (P ) = (1 − db(1 − b
U ναb )) (B − P ) − db(1 − α
b) D (4)

The monopolistic seller aims to maximize expected profits, taking into

account the amount of responsibility he bears if the good is defective. De-
noting by n (P, ·) the number of consumers served, his expected profit when
he acts independently is given by

πSI (p) = [p − dναD] n − K (5)

where for simplicity we have assumed no production costs and positive fixed
cost K.
On the credit market, the supply of loans entails positive average trans-
action and management cost F if the seller and the lender operate indepen-
dently, and F c if they coordinate, with F c ≤ F . This latter assumption is
meant to capture the reduction in the lender’s management costs that result
from the possibility of using the facilities offered by the seller to manage the
supply of loans to his costumers. For simplicity we let F c = 0. The credit
grantor faces a perfectly elastic supply of funds at an exogenously deter-
mined interest rate, which we normalize to zero. Since perfect competition
forces the credit grantor to break even on his loans, under independence,
the interest rate he must charge is given by r = F .
10
Clearly, we are referring only to those consumers who seek credit from the connected
lender, those who pay cash still operate in a regime of seller responsibility.

7
 Suppose now that the seller and the lender coordinate their decisions
and as a result accept co-responsibility in case of product failure. Let nJ
be the number of consumers who buy on credit from the lender connected
to the seller, and n − nJ be the number of consumers who either buy
cash or buy on credit from an independent lender. Any agreement between
the seller and the lender will have the property that the price and interest
rate will be set so as to maximize the total expected profits of the two
agents. The contract will then specify a monetary transfer t that allocates
the gains from coordination among the two parties. Given that the seller
has all the bargaining power, in equilibrium the transfer will be equal to the
level that maintains the lender on his reservation profit of zero. This yields
t = il + (1 − ν)αDnJ where l = pnJ is the amount of loans and (1 − ν)αDnJ
is the amount of responsibility that the lender bears in case the seller goes
bankrupt and for which he will have to be compensated ex ante. Hence, the
expected profit of the seller is
¡ ¢
πJ = [(1 − d (1 − να)) p − dαD] nJ + [p − dναD] n − nJ + il − K (6)

Note that under joint responsibility, the seller receives the price p only if
the consumer receives a non-defective good, which occurs with probability
(1 − d (1 − να)) ; moreover, he bears the liability for damage with probabil-
ity 1, due to the compensation he needs to offer to the lender.
Finally, notice that if coordination did not involve any change in the
responsibility regime then the seller’s expected profit when he cooperates
with the lender would be given by

πSC = [p − dναD] n + il − K (7)

with t = il.

3 Benchmarks
3.1 No misperceptions
In this section we consider the benchmark where consumers are not af-
fected by misperception. For simplicity we shall refer to them as “rational”
consumers.
Rich consumers do not need to borrow to purchase the good. Therefore
their reservation price, from (1), is given by

pS = B − d (1 − να) (B + D) (8)

8
On the contrary, poor consumers must resort to credit to finance their pur-
chase. Since they have to borrow p, and r = F > 0, their reservation price
pS
is lower than their rich counterpart and equal to 1+F . Therefore, the profit
opportunities of a seller who relies on a independent lender to provide credit
to his customers are as follows. The seller can set p = pS and sell only to
pS
the rich consumers or he can set p = 1+F and supply the entire market at a
lower price. We assume that F is sufficiently high that it¡ is ¢optimal to sell
only to the rich consumers. Then, the seller’s profit, πSI pS , is given by
¡ ¢
πSI pS = Πnr − K (9)

where Π ≡ [(1 − d (1 − να))B − dD].

Now suppose that the seller can coordinate his decisions with the lender
without this implying any changes in the responsibility regime (i.e. there is
seller-only responsibility). Coordination increases the profit opportunities
for the seller since he can choose the price and interest rate that maximize
his profits (net of a transfer to the lender). In particular, by charging pS
and setting i = 0, the seller can attract the poor consumers without having
to reduce the price below the willingness to pay of the rich ones. This is
clearly the optimal strategy since it allows the seller to extract the entire
consumer surplus. Substituting for pS , i = 0, n = (nr + np ) and l = np into
(7), yields
¡ ¢
πSC pS , i = 0 = Π (nr + np ) − K (10)
¡ ¢ ¡ ¢
Notice that πSC pS , i = 0 > πSI pS , that is, it is always in the interest
of the seller to coordinate his decisions with the lender, since he can attract
the poor consumers without reducing the profits he can make on the rich
ones.11
Let us assume now that a regime of joint liability applies to any agree-
ment between the seller and the lender.

Proposition 1 When all consumers are rational, joint liability is ineffec-

tive.

This result is not surprising. The rationality of consumers, coupled with

their neutrality towards risk implies that additional protection is irrelevant.
This is because, when consumers are rational the seller can always repli-
cate the same situation as with coordination and seller-only liability by
11
More generally, coordination is optimal for F c sufficiently low.

9
transferring the additional liability into higher prices. In particular, the
seller charges a price equal to the willingness to pay ¡of the rich¢ consumers,
¡ ¢
d(1−να)
pJ = B − (1−d(1−α)) D, and sets i = 0. This yields: π J pJ , i = 0 = πSC pS
¡ ¢ ¡ ¢
and U J pJ , i = 0 = U S pS = 0.

3.2 Misperceptions and seller-only responsibility

We now allow for consumers’ misperception and analyze the seller’s choice
when a regime of seller-only responsibility applies. The effects of introducing
joint liability legislation will be investigated in Section 4.
The willingness to pay for the good of a rich and optimistic consumer,
from (2), is given by
pbS = B − P − db(1 − b
ναb ) (B + D) (11)
At this price the real utility of the consumer is
¡ ¢ h i
U S pbS = db(1 − b ναb ) − d (1 − να) (B + D) < 0 (12)
That is, when consumers are affected by misperception and the seller sets
the price equal to their willingness to pay, they suffer a loss in real terms.
The reason for this is that pbS is greater than the real expected value of the
good, which is given by what rational consumers would be willing to pay,
i.e. pS . Assuming again that under no coordination it is never optimal to
sell to poor consumers (F sufficiently high), the seller will charge pbS and his
expected profit will be
¡ ¢ £ ¤
π SI pbS = Π − U S (b ps ) nr − K (13)
Note that the loss that consumers make due to their misperception, U S (bps ) ,
becomes additional profits to the seller. Conversely, when the seller and the
lender coordinate their decisions, the following results obtains.
Lemma 2 When seller-only responsibility applies to coordination, the seller
always prefers to coordinate and i = 0. Moreover, he sells to all consumers
and his profit is given by
¡ ¢ £ ¡ ¢¤
πSC pbS , i = 0 = Π − U S pbS (nr + np ) − K
¡ ¢
Consumers are fooled by their misperception: U S pbS < 0 from (12).
As in Section 3.1, the optimal strategy under coordination is to fully sub-
sidize credit, i.e., to set i = 0 and sell to both the rich and poor consumers.
Moreover, coordination is always desirable for it allows full extraction of
consumers’ surplus by discrimination between the two classes.

10
4 Joint Liability
In this section we consider the case of a legislation that imposes a joint
liability regime to those sale contracts where the consumer obtains credit
from a lender connected to the seller.
A measure of the extent by which consumers can be harmed by their
misperception is given by the difference between their perceived utility and
their real utility, for the greater this difference the more a consumer will
be incorrectly evaluating the benefits from consumption. Under seller-only
responsibility, Ub s (·) − U s (·) , as given respectively by (2) and (1), amounts
to
h i
b S (·) − U S (·) = d (1 − να) − db(1 − b
U b ) (B + D) > 0.
να

whereas, under joint liability, from (4) and (3), we obtain

h i
b J (·) − U J (·) = d (1 − να) − db(1 − b
U b ) (B − (1 + i)p)
να
h i
+ d (1 − α) − db(1 − αb) D ≥ 0

Straightforward calculations show that, for any given p and i such that
b J (·) ≥ 0
U

U b J (·) − U J (·)
b s (·) − U s (·) > U (14)

Two main implications follow from expression (14). First, the seller can-
not fully transfer the additional liability into higher prices, that is, he can
never replicate the same situation as with coordination and seller-only lia-
bility. This is because optimistic consumers underestimate the value of the
additional protection given by joint liability and therefore are not willing
to pay the price that reflects the true value of the benefits they obtain.12
Thus, if the seller increases the price and/or the interest rate so as to leave
them indifferent between the two liability regimes (U bs = U b J = 0), their
J S
real utility increases (U > U ). This is a crucial point for it suggests that
joint liability may help protect consumers from their misperception and that
Proposition 1 does not extend to the case where consumers are irrational.
In fact, since d > db consumers underestimate the probability that the bad state occurs.
12

Other things equal, this implies that they underestimate the value of protection. However,
the assumption that α b > α works in the opposite direction, for consumers overestimate
the amount they can recoup in court. Our assumption that dα − db bα > 0 implies that the
first effect prevails and overall consumers underestimate the value of additional protection.
As we discuss in Section 4.2, if this were not the case, then seller liability should be zero.

11
A second implication of expression (14), which is strictly related to the pre-
vious point, is that it is in the interest of the seller to minimize the number
of consumers who borrow from a connected lender. This means that the
beneficial effects of joint liability to the poor consumers may not extend to
the rich ones. Indeed, as we shall see, the seller will use joint liability as
a market-segmentation technique: by appropriately choosing the price and
the interest rate, he will continue to extract a rent from the rich consumers
by giving them incentives not to switch to credit.

To illustrate both points, as a first step, take the case where consumers
incur no loss for damage when they purchase from a connected seller under
a regime of joint liability. This can happen either when D = 0 or when the
judicial system works efficiently and the consumers know it: α b = α = 1.
Indeed, from (3), when α b = α = 1 the misperception on the risk of seller
bankruptcy plays no role under joint liability, since consumers can resort to
the lender. As an example of a situation where D = 0 one can think of the
case where the seller does not deliver the good or delivers a good which is of
no use to the consumer. The assumption α b = α = 1 works well for countries
with a well established tradition for protecting consumers’ interests or where
consumers associations are strong enough to ensure that consumers are fully
compensated for the damages they suffer.
> From expressions (3) and (4), when D = 0 or α b = α = 1, the real and
perceived utility of a consumer who purchases the good seeking credit from
a connected lender are respectively given by

U J (·) = (1 − d (1 − να)) (B − (1 + i)p)

b J (·) = (1 − db(1 − b
U ναb )) (B − (1 + i)p)

These expressions suggest that when consumers suffers no loss for damage
under joint liability, there does not exist a couple (p, i) such that they receive
a negative real utility, since for any p, i such that U J ≥ 0, we have U b J ≥ 0.
The intuition lies in the fact that consumers pay for the good if and only if
they receive B, and enjoy zero utility otherwise. Since consumers will never
buy if the effective price is greater than the value of the good, B, the real
utility can never be negative. This leads us to the following result.

Proposition 3 When product failure entails no damage (D = 0) or judicial

enforcement is efficient (b α = α = 1), joint liability fully protects the poor
consumers from their misperceptions, while the rich ones are unaffected. In
particular, the seller signs the agreement with a lender; all consumers are

12
served; the poor ask for credit to the connected lender and obtain zero utility;
the rich buy for cash and are fooled.

Proposition 2 can be understood as follows. The optimal policy for a

seller who signs the agreement with a lender is to charge pbS as defined in
b αb
(11) and iJ = d(1−b ν)
. This ensures Ub S (b b J (b
pS ) = U pS , iJ ) = 0. Thus, all
b
1−d(1−bαbν)
consumers are willing to enter the market: poor consumers ask for credit
to the connected lender while the rich ones buy for cash. Moreover, since
joint liability enables the seller to price discriminate among poor and rich
consumers, it is always in his interest to sign the agreement. The seller’s
profit under this policy is given by
¡ ¢ ¡ ¡ ¢¢
πJ pbS , iJ = Π − U S pbS nr + Πnp − K

The effect of joint liability on consumer welfare is positive. More precisely,

joint liability fully protects poor consumers (U J (b
pS , iJ ) ¡= 0)
¢ and has no
effect on the welfare of the rich ones who still receive U S pbS < 0.
¡ ¢ ¡ ¢
Comparing πSC pbS and πJ pbS , iJ , an immediate consequence of Propo-
sition 2 is as follows.

Corollary 4 When product failure entails no damage (D = 0) or judicial

enforcement is efficient (b α = α = 1), joint liability hurts the seller, for it
redistributes some of his profits to the consumers.

The above corollary suggests that sellers and lenders would never volun-
tarily
¡ offer¢ joint liability,
¡ ¢ because it reduces their gains from coordination
(πJ pbS , iJ < πSC pbS ). However they do undertake joint responsibility
when this is a condition sine qua non for coordination,
¡ i.e. ¢to price ¡discrim-
¢
inate between rich and poor consumers. In fact πJ pbS , iJ > πSI pbS . In
this respect, joint liability can be viewed as a means to redistribute some of
the gains from coordination to the consumers.

Now, let us turn to the case where product failure leads to damage,
D > 0 and α b > α. In this setting, while the inequality in (14) still holds,
it is no longer true that consumers under joint liability can never suffer a
negative real utility. Indeed, from (4), the willingness to pay for the good,
inclusive of interest (i.e. (1 + i)p), under joint responsibility is

db(1 − α
b)
PbJ = B − ³ ´D (15)
1 − db(1 − α
bbν)

13
with associated real utility equal to (from (3))
h i
db(1 − α
b ) − d (1 − α)
U J (PbJ ) = ³ ´ D<0 (16)
1 − db(1 − αbb
ν)
where
pS ) < U J (PbJ ) < 0
U S (b (17)
Hence, consumers can still be fooled by their misperception. This is because
when D > 0, and α b > α consumers overestimate the level of protection given
by the judicial system and their utility in the bad state will be negative.
pS ) < U J (PbJ ) implies that the extra rent that the seller can
However, U S (b
make on irrational consumers is lower than under seller-only responsibility.
This underlies the following Proposition.
Proposition 5 When product failure entails damage (D > 0) and judicial
enforcement is inefficient (b α > α), joint liability redistributes the gains of
coordination from the seller to the poor irrational consumers. However, poor
consumers are not fully protected from their misperceptions.
Notice that the seller can still price discriminate between the two classes
of consumers. It suffices that he chooses pbS and adjusts the interest rate
so as to obtain PbJ = (1 + iJ )b pS as given by (15). This yields U b S (b
pS ) =
Ub (Pb ) = 0. Thus rich consumers buy for cash while poor consumers buy
J J

on credit from the connected lender. Consequently, as in Proposition 2, rich

consumers receive no benefit from joint liability. Poor consumers are better
off since the extra rent that the seller can make on irrational consumers is
lower under joint liability (from (17)). However, contrary to Proposition 2,
they are not fully protected (from (16)). More precisely, the seller obtains
πJ (b ps )) + np (Π − U J (PbJ ) − K
pS , iJ ) = nr (Π − U S (b (18)
b αb
where iJ = d(1−b
b
ν)
.
1−d(1−bαb
ν)
In the next section we shall discuss other factors that may help to deal
with consumers’ misperceptions.

5 Other means to deal with misperception

5.1 The extent of liability
So far we have taken a positive approach and assumed that when the good is
defective the consumer is allowed to claim B +D. This was proven to always

14
increase consumer welfare, but not to ensure full protection when there is
damage and the judicial system does not work perfectly. This conclusion
raises the question of whether consumers could be made better-off by a
different choice of the extent of liability. To answer this question, consider
the perceived and real utility of a consumer for a given liability L, under
seller-responsibility (similar results are obtained under joint responsibility)
b S (p, L) = B − P − db(B + D − b
U ναb L)
S
U (p, L) = B − P − d (B + D − ναL)

It follows that at the price Pb such that U(

b Pb, L) = 0, the real utility of the
consumers is given by
b
U S (Pb, L) = −(d − d)(B bν α
+ D) + (dνα − db b )L

bα > 0 (which implies dνα − db

which is increasing in L since dα − db bν α
b > 0),
and is equal to zero for13

b
(d − d)(B + D)
L∗ = .
b
dνα − db b
να
Thus, the optimal extent of liability is larger than B + D. The intuition
is that optimistic consumers always benefit from an increase in the level of
protection since the price they are willing to pay for the protection is lower
than its effective value. Notice also that L∗ is greater than the value of L that
equates utilities across states, which is given by (B+D)
να . Equal real utilities
across states results in the consumer overestimating the value of the good
since the perceived utility in the bad state is greater than the real utility.
Therefore, in order to fully protect consumers from their misperceptions,
the utility in the bad state needs to be higher than that in the good state.14
Unfortunately, such a solution would be difficult to implement as it would
require the regulator to possess information on the extent of consumers’
misperception which may be rather expensive to acquire.
13
Notice that if dα − dbbα (respectively dνα − db
bν α
b ) were negative, then under seller-only
responsibility (respectively joint responsibility) consumer protection would be ensured by
setting the seller liability equal to zero.
14
This result differs from that obtained by Spence (1977) where a liability rule that
equalizes the real utilities across states is sufficient to ensure full protection when con-
sumers are irrational and risk neutral. This is because in Spence’s model consumers un-
derestimate the probability of product failure but not the probability of seeing their claims
satisfied. Hence the consumers’ perceived and real utilities in the bad state coincide.

15
5.2 The presence of well informed consumers
Contrary to the previous sections, let us assume that, within each class (the
rich and the poor) a fraction µ of consumers is not affected by misperception.
For simplicity we shall refer to these consumers as ‘rational’ and to those
affected by misperception as ‘irrational’. Rational consumers value the good
correctly and their willingness to pay under seller-only responsibility is given
by pS from (8) with pS < pbS . The profit opportunities for the seller under
seller-only responsibility are the following. Either he sets pbS and serves
only the irrational consumers, or he sets pS and attracts both ¡rational
¢ and
irrational ones. In the former case, he gains the extra rent U S pbS on each
of them. In the latter,¡ he¢ serves a higher segment of the market but cannot
extract the rent U S pbS from the irrational consumers. This reasoning
suggests that the presence of rational consumers may work as a protection
device that gives incentives to the seller to give up the rent on misperception
in order to increase the market demand. Clearly, whether the seller finds
this profitable depends on how large the proportion of rational consumers
is.
Let µS and µJ be the cut-off value such that the seller prefers to keep
rational consumers out if and only if µ ≤ µS under seller-only responsibility
and µ ≤ µJ under joint liability when D > 0 and α b > α.

Proposition 6 When product failure entails damage (D > 0) and judicial

enforcement is inefficient (b α > α), µJ < µS : joint liability increases the
welfare of all irrational consumers, where the poor ones are fully protected
by their misperception.

P roof. See Appendix.

Proposition 4 suggests that under joint liability the number of rational
consumers that is needed in order to induce the seller to supply the entire
market is lower than under seller-only responsibility. This is because joint
liability reduces the rent that can be extracted from the irrational consumers
(Proposition 3) and thus increases the seller’s incentives to serve the rational
ones. Note that serving the (poor) rational consumers requires U J (.) = 0.
Hence, once the rational consumers are served, the poor irrational ones are
fully protected. Moreover, price discrimination between rich and poor con-
sumers becomes more difficult. Indeed, deterring irrational rich consumers
from switching to credit implies U b J (.) , which is always positive
b S (.) = U
J
given that U (.) = 0. Hence, the perceived utility of the rich irrational con-
sumers is higher than under seller-only responsibility (U b S (b
pS ) = 0). The

16
seller charges pS < pR < pbS which still enables him to extract some rent but
leaves the rich irrational consumers better off.

Remark 1 For the presence of rational consumers to act as a protection de-

vice, it is necessary to monitor the behavior of the seller and impede practices
that may allow the seller to price discriminate among the different degrees
of rationality. Indeed, the seller can try to offer different types of contracts
in order to induce the consumers to self-select. One way to achieve this is
through optional warranties. Suppose that the seller charges a price for the
good equal to the reservation price of the irrational consumers and offers an
optional warranty at a price that leaves the rational consumers with zero util-
ity. Contrary to rational consumers, irrational consumers have no incentive
to buy the warranty, since their willingness to pay for³the protection is
´ lower
than the willingness to pay of the rational consumers since dα > db bα . This
self-selection mechanism allows the seller to attract rational consumers with-
out losing the rent that can be extracted from the irrational ones.

The above considerations can easily extend to the case where some con-
sumers are pessimistic. Indeed, pessimistic consumers underestimate the
value of the good and, in the absence of price discrimination devices, their
entry can be beneficial to the optimistic consumers. Further, they suggest
that government policy devoted to increase the proportion of informed con-
sumers can be effective even if not all consumers are made informed.

6 Conclusion
That consumers’ misperception my result in wrong purchasing decisions is
now well known and underlies widespread government policies devoted to in-
crease consumers’ awareness of the risk of product failure. In UK the Office
of Fair Trading constantly publishes statistics as to the extent of consumers’
complaints for the quality of the good or service received. Clearly, if con-
sumers could be made perfectly informed, the risk of product failure would
be a problem of a minor concern, since prices would adjust and consumers
could correctly estimate the value of warranties or insurance packages. Sim-
ilarly, if regulators were aware of the extent of consumers’ misperception,
they could be in a position to impose the right level of seller liability even
case by case. In practice this is difficult and costly to implement and it
seems reasonable to think of some more indirect ways to protect consumers.
In this paper we have studied a widespread legislation in the consumer
credit market that extends the liability for product failure to financial in-

17
stitutions connected with suppliers of goods and services. We have shown
that lender liability is an effective, though not perfect, device to increase
consumer protection and to redistribute some of the gains that result from
the coordination of price and interest rate decisions away from the seller
(lender) toward the consumers. Specifically, joint liability yields the full
protection of those consumers who buy on credit in each of the following sit-
uations: (i) product failure entails no damage (ii) the judicial enforcement
is efficient, (iii) the proportion of informed consumers is sufficiently high. In
the remaining cases, it helps but is not sufficient; neither does joint liability
fully protect those consumers who can afford to buy for cash and do not
understand the benefit of buying on credit from a connected lender.
We have considered a perfectly competitive financial market and a mo-
nopolistic product market. Our results extend to more general settings pro-
vided that on either market there exists some degree of monopoly power that
makes advantageous the coordination of price and interest rate decisions .15
This paper has taken a positive approach to joint liability rules, rather
than a normative one, although some suggestions have been provided as to
how to design the optimal liability rule. In particular, we have shown that
when consumers not only underestimate the probability of product failure
but also overestimate the probability of seeing their claim satisfied, then
liability rules need to ensure a utility in the bad state that is higher than
that in the good state. A more detailed normative analysis could constitute
an interesting objective for further research. Similarly, new insights could
be gained by endogenizing the probability of product failure.
15
Spence (1977) shows that misperception may raise an issue of consumer protection
even when the market is perfectly competitive.

18
7 Appendix
¡ ¢
Proof of Proposition 4. Suppose that µΠ ≤ −(1 − µ)U S pbS , which
holds for µ ≤ µs where
¡ ¢
S −U S pbS
µ =
Π − U S (b
pS )

This implies that under seller-only responsibility, the seller prefers to set
pbS > pS ; his profits, under independence and coordination, will be respec-
tively
¡ ¢ £ ¡ ¢¤
πSI pbS = Π − U S pbS (1 − µ)nr − K
¡ ¢ £ ¡ ¢¤
πSC pbS , i = 0 = Π − U S pbS (1 − µ) (nr + np ) − K
¡ ¢
where U S pbS < 0 is given by (12).
Now consider the case where joint liability is imposed on seller-lender agree-
ments. If the seller chooses to serve all consumers, profit maximization is
achieved when the cash price and interest rate charged are such that U b J (·) =
Ub (·) and U (·) = 0. Notice that U (.) = 0 is necessary to induce the poor
S J J

and the rich rational consumers to enter the market, whereas U b J (.) = Ub S (.)
is required to deter rich irrational consumers³ from switching
´ ³ to credit. Sim- ´
b
ple algebra shows that this yields: pR = 1 − d (1 − α d(1−α)
b ) B − 1−d(1−α) D
b ¡ ¢
and iJ = d(1−b b
α)
where 0 < pS < pR < pbS and 0 > U S (pR ) > U S pbS : the
1−d(1−b α)
rich and irrational consumers are also better off than under no coordination.
In this case, the seller’s profit is given by
¡ ¢
πJ pJR , iJ = (1 − µ)nr (Π − U S (pR )) + (µnr + np )Π − K (19)

By contrast, if the seller decides to supply only irrational consumers, profit

b ¡ ¢
maximization entails pbS and iJ = d(1−b α)
such that U b S pbS = 0 and
b
1−d(1−b
α)
Ub J (b pS = PbJ . Rich consumers buy for cash while
pS , iJ ) = 0, with (1 + iJ )b
poor consumers buy on credit from the connected lender. Both groups are
fooled. Under this policy, the seller only supplies a fraction of the market,
but extracts an extra rent from both the rich and poor irrational consumers.
This yields
¡ ¢ ¡ ¢
πJ pbS , iJ = (1 − µ)nr (Π − U S pbS ) + (1 − µ)np (Π − U J (b
pS , iJ ) − K
(20)

19
 Comparing (20) with (19) the seller will supply the entire market iff µ < µJ
where
¡ ¢
J −nr (U S pbS − U S (pR )) − np U J (b pS , iJ )
µ = (21)
−nr (U S (b
pS ) − U S (pR )) − np U J (b
pS , iJ ) + (nr + np ) Π
s s
Note that µJ from (21) can be rewritten as µj (A) = (n−(n r +np )U (b
p )+A
s ps ))+A ,
r +np )(Π−U (b
¡ ¡ ¢ ¢ J
with A = nr U S (pR ) + np U s pbS − U J (b pS , iJ ) < 0. Since − ∂µ∂A(A) < 0
and µJ (0) = µS , it follows that µJ < µS .

References
[1] Boyer, M. and J.J. Laffont (1997). Environmental Risk and Bank Lia-
bility. European Economic Review, 41(8): 1427-1459.
[2] Daughety, A and Reinganum, J (1995). Product Safety: Liability, R&D,
and Signalling. American Economic Review, 85(5):1187-1995.
[3] Eisenberg, M. A. (1995). The Limits of Cognition and the Limits of
Contracts. Stanford Law Review, 47: 211-259.
[4] Epple, D. and A. Raviv (1978). Product Safety: Liability Rules, Market
Structure And Imperfect Information. American Economic Review, 68:
80-95.
[5] Green, J. (1976). On the Optimal Structure of Liability Laws. Bell Jour-
nal of Economics, 7: 553-574.

[6] Pitchford, R (1995). How Liable Should a Lender Be? The Case of
Judgement-Proof Firms and Environmental Risk. American Economic
Review, 85(5): 1171-1186.

[7] Polinsky, A M. and W.P. Rogerson (1983). Products Liability, Consumer

Misperceptions, and Market Power. Bell Journal of Economics, 14(2):
581-589.

[8] Shapiro, C. (1982) Consumer Information, Product Quality, and Seller

Reputation. Bell Journal of Economics 13: 20-35.
[9] Spence, A. (1977). Consumer Misperceptions, Product Failure and Prod-
uct Liability. Review of Economics and Statistics, 44: 561-572.

20

View publication stats