The Theory of Warranty Contracts: University
The Theory of Warranty Contracts: University
The Theory of Warranty Contracts: University
Winand Emons
University of Basel
Abstract. This paper surveys theories of the existence and design of warranty
contracts. Insurance, signalling, and incentive motives are used to explain the
existence of warranties. The incorporation of imperfections that prevent
warranties from serving their basic purposes in a first-best way describes actual
warranty practices.
1. Introduction
Consumer product warranties are prevalent in many markets. Consumer '
durables, as well as some necessities a n d services are frequently sold in
combination with some kind of warranty. In most countries the law implies a
warranty of merchantability in all sales contracts, i.e. goods should d o the job
they are supposed to do. In the US explicit warranties are mandated for all
consumer products sold for more than fifteen dollars by the Magnuson-Moss
Warranty Act of 1975. In spite of this prevalence, economists began only recently
t o study the economic roles of warranties. This paper surveys theories of the
existence and design of warranties.
Warranties constitute a claim for the buyer o n what a seller will d o in the event
of product failure. Accordingly, warranty contracts depend on the performance
of the good in question. The contract can only be conditioned on characteristics of
the good which are verifiable ex post by both parties to the contract as well as
by the courts. If one contracting party cannot observe the characteristics on
which the contract depends, the warranty would be vacuous. The characteristics
also must be verifiable by the courts to enforce the contract in case of disputes
between seller and buyer.
Typically, the buyer receives nothing from the warranty contract if the good
works satisfactorily. I f the good shows any defect, the buyer gets the coverage
which is promised in the contract. Warranties typically include replacement or
repair of defective products, price refunds a n d sometimes include reimbursement
for consequential damages. We will subsume under the term warranty the legal
obligations that derive from a contract of sale as well as commitments which a
seller undertakes in the form of a n additional warranty contract, because we are
only interested in the economic roles of warranties.
Warranties as defined above are thus distinct from service contracts and
money-back guarantees. A service contract specifies that the seller just does the
0950-0804/89/01 0043- 14 $02.50/0 0 1989 W. Emmons
J O U R N A L OF E C O N O M I C SURVEYS Vol. 3. No. 1
44 EMONS
regular maintenance work such as lubrication etc. but has no other obligations.
Accordingly, service contracts d o not depend o n the product’s performance.
Money-back guarantees extend to a consumer, who is merely dissatisfied with a
working product, the right t o return the item. Under a warranty a bride may have
defects in her wedding dress corrected. Yet she cannot return the dress simply
because the groom cleared off. T h e latter is possible under a money-back
guarantee. ’
T h e literature on warranties may be broadly classified as follows. O n e strand
analyses warranties that make buying goods a less risky venture by providing
either insurance o r information. T h e second strand analyses warranties as a
marketing strategy t o extract consumer surplus. The main thrust of the first
branch of the literature is to provide basic motives for the existence of warranties.
T h e insurance motive is based o n the assumption that buyers a r e more risk averse
than sellers. With different attitudes towards risk, sellers provide buyers with
insurance against the event of product failure in the form of warranty contracts.
The signalling motive rests o n the following idea. Several sellers offer a product
at different, exogenously given quality levels. Buyers cannot discern the respective
quality of a n individual offer. Sellers use warranties as signals of the qualitative
attributes of a product. Because a more reliable product incurs lower warranty
cost, a producer can signal a high quality of his product by an extensive warranty
coverage. T h e incentive motive for warranties is closely related to the signalling
motive. The decision of firms about which quality to produce is endogenous
rather than exogenously given. T h e incentive motive views warranties as an
incentive device for firms not t o cheat on quality. By lowering the quality level,
firms increase the probability of product failure and thereby incur higher
warranty cost. Accordingly, warranties provide sellers with the incentive to
supply high quality products.
In the following section we analyse the basic motives in detail. It will turn out
that warranties may act as a signal of product quality o r provide producers with
the incentive not t o cheat on quality a n d at the same time provide consumers with
full insurance. To understand the incomplete warranties that a r e prevalent in real
world economies, in Section 3 we describe theories which a d d additional
incomplete information about the buyer’s side t o the basic motives for
warranties.
In Section 4 we will look at the second branch of the literature which views
warranties as a marketing device. T h e main thrust of these theories is to study
warranties which serve t o insure consumers as a strategic variable of firms with
market power. The use of warranties by a monopolist t o segment consumers with
different tastes o r different incomes will be analysed. Further the role of
warranties as a surplus extracting device when consumers are uninformed about
the price-warranty combinations available o n the market will be described.
Economic theories of warranties have been developed in the last decade o r so.
Before that, their analysis was mainly confined t o law schools. In the concluding
section we will briefly look at a theory which takes the existence of warranties as
given and explains their design by t h e underlying market structure.
WARRANTY CONTRACTS 45
Figure 1. Efficient risk-sharing requires the complete warranty w = I (depicted for a com-
pet i t ive market).
in the form of a warranty. Buyers gain from the transfer as long as their income
in the bad state is lower than their income in the good state. Accordingly, sellers
will insure buyers u p to the point where their income is the same in both states,
i.e. where w = 1.
Which side of the market appropriates the surplus from the efficient risk-
sharing depends upon the market structure. If the seller is a perfect competitor,
he has to offer a premium-benefit ratio which results in zero profits to have a
clientele at all. If the seller is a monopolist, he will charge a premium-benefit ratio
such that consumers are indifferent between buying and not buying the warranty
contract. See Heal (1977).
To summarize, the description of a warranty as a n insurance policy in a world
with symmetric information predicts that warranties provide buyers with
complete insurance w = 1, independent of the market s t r ~ c t u r e See. ~ Figure 1.
The insurance motive provides a good explanation of the various optional
extensions to tied-in warranty agreements. These optional warranties cover e.g.
transport charges o r repairs in foreign countries. The separate offer, for a
separate price, reduces the warranty premium tied to the sale of the basic product
and may optimize the risk allocation. A British car buyer who declines to drive
on the right may thus save o n the warranty premium for coverage of repairs in
continental Europe.
statements. Accordingly, the products are identical for consumers and will be
traded at an average price which reflects average quality. This average price may
not cover the production cost of high quality products. High quality producers
stop offering their products. The range of quality levels offered worsens; good
products are driven out of the-market by bad products. Rational consumers will
take this kind of producer behaviour into account. They will no longer base their
decision on the average quality of all potential producers, but on the average
quality of those firms which actually offer the product. This kind of consideration
lowers the price consumers are willing to pay. Those high quality producers who
still offered their product at the average price reflecting the average quality of all
potential firms may now opt to stop selling their product. Accordingly, even more
good products are driven out of the market. This informational asymmetry may
even result in market break-down. See Akerlof (1970).
A solution to this adverse selection problem consists in firms sending credible
signals of product quality. I f high quality producers send signals which are too
costly to be imitated by low quality producers, then rational consumers can infer
from the observation of these signals that the products must be high quality ones.
Assume that in our model some firms produce a high quality level q H while the
other firms produce a low quality level qi. Since q H > qi, a high quality producer
can offer an additional unit of warranty at a lower cost than a low quality
producer. See Figure 2.
I f high quality producers offer a warranty w > W at the fair-odds rate (1 - q H ) ,
a low quality producer mimicking this signal would incur losses, because the
failure probability of his product is higher. Therefore, low quality producers
prefer to stay out of the market. I f firms are perfect competitors and consumers
are risk neutral, i.e. U" = 0, in equilibrium only high quality producers offer their
product with any warranty w > W at the fair premium-benefit ratio (1 - q ~ ) .
Warranties serve only a signalling purpose. If consumers are risk averse, i.e.
U " < 0, firms offer the complete warranty w = 1. Warranties then serve both,
signalling and insurance motives. See Grossman (1981).'
P .14* P * I = O
v.IqhP,“ 1= v*
.W.P “1.0
V , I O . P “ I =v ,
‘ i I , . P . *I-0
, I
, I
,,
, I
,
I, ,,
,
*“ I
consumers purchase the warranty level w, < 1 at their fair odds rate (1 - q),
where the rationed warranty w, is constructed such that high-risk consumers have
no incentive to purchase this cheap contract for low-risk consumers. Since high-
risk consumers face a higher break-down probability than low-risk consumers,
they care more about insurance, a n d rationing the indemnity can be used as a self-
selection device.
Accordingly, under adverse selection circumstances we observe the following
two market outcomes. In a pooling situation, all consumers buy the same
warranty contract and the amount of warranty w,, is less than complete because
the premium-benefit ratio (1 - 4 ) is unfair for low-risk consumers. In a
separating situation, high-risk consumers purchase the complete “‘ai lcllity w = 1
at their fair odds rate (1 - q h ) . Low-risk consumers purchase a warranty level
w, < 1 at their fair odds rate (1 - q);ws is rationed so that high-risk consumers
prefer to purchase the contract designed for them.
This adverse selection phenomenon has also been used to explain limited
warranty duration rather than limited warranty coverage. Consider the situation
where firms offer a product with a stock of services which consumers exhaust over
time. High-risk consumers use the product at a higher intensity than low-risk
consumers. Therefore, high-intensity users have exhausted total capacity at some
point in time at which low-intensity users are still utilizing the product. Firms
cannot distinguish between the two groups nor can they determine whether a
product failed because it was worn out by a high-intensity user or because a n
unlucky low-intensity user experienced a random failure. Although in a pooling
and a separating situation low-intensity users can obtain a warranty duration
equal to their product’s lifetime, they may prefer a limited warranty duration
under both circumstances. Limited warranty duration means that low-intensity
users bear the risk of product failure alone for the last period of product life. See
Emons (1989). Adverse selected phenomena a r e thus able to explain not only
52 EMONS
limited warranty coverage but also the limited warranty duration which is
generally observed.
incomes such that rich consumers have a higher preference for insurance than
poor consumers, then the following strategy is superior to pure monopoly
pricing. The monopolist offers the product at a low price without any warranty
and at a high price with a warranty which replaces defective units free of charge.
The warranted product is bought by rich consumers whereas the unwarranted
product is bought by poor consumers. With this optional pricing strategy, the
monopolist extracts more consumer surplus than with pure monopoly pricing.
See Kubo (1986).
Warranties as a marketing device for a monopolist to segment consumers thus
explain the phenomenon that t o p of the line products come with a more extensive
warranty than economy versions. The different warranty terms serve the purpose
of consumer self-selection so that the seller extracts more surplus than with pure
monopoly pricing. An obvious example is the automobile market in which new
cars typically come with a warranty whereas it is often difficult to get a warranty
on used cars. The airline passenger market provides another example. When first-
class passengers are bumped, they receive a ticket for a later flight plus additional
compensation. When standby passengers are denied boarding, they receive n o
compensation. A n alternative explanation for distributions of warranty contracts
is provided by the search approach.
market may result in rhe product being offered both with and without warranty.
Firms which exploit nonshoppers d o this by supplying the product without a
warranty. Search models thus provide an alternative explanation for the
distributions of warranties.
5 . Conclusions
This paper has summarized recent work on warranty contracts. At this point, it
is appropriate to mention the first approach to explain the design of warranties,
dating back to Kessler (1943). The so called exploitation theory takes the
existence of warranties as given and explains their design by the market power
of manufacturers. According to this theory standardized warranty contracts are
drafted unilaterally by the seller and only involuntarily adhered to by the
Co&mer. Following Kessler, standardized contracts are typically used by sellers
with strong market power. The consumer, in need of the goods, is frequently not
in a position to shop around for better terms, either because the seller has a
monopoly, or because all competitors use the same clauses. The seller possesses
‘unfettered discretion’ to incorporate terms that serve his interests because his
bargaining position is superior to that of the consumer.
I t follows from the exploitation theory that sellers will limit their obligations
to consumers as far as possible. If collusion is widespread, warranty contracts
within individual industries are likely to be similar. More recent statements of the
theory emphasize the marketing power gained from combining advertising that
makes extraordinary promises with warranties that disclaim responsibility for the
promises.
The exploitation theory found wide acceptance because it was the only
approach to standardized warranty contracts until the 1970s. The theory seems
to be consistent with case histories of warranty practices, see Priest (1981). The
theory substantially influenced courts which henceforth refused to enforce
exploitative elements of standardized warranty contracts. Nevertheless, the
theory does not provide any explanation of the existence of warranties.
Therefore, it is unclear why warranties can serve exploitative ends. If the seller
disclaims promises in the warranty contract, rational consumers will ignore these
promises beforehand. As a marketing device the warranty contract is ineffective
and cannot actually serve to exploit consumers. One has to assume irrational
consumer behaviour to explain exploitation with warranties that disclaim
responsibility. We have described several ways that binding warranties can be
used as a strategic variable of firms with market power. Accordingly, if the seller
can sign binding warranties, he extracts more surplus from rational consumers
than with contracts that disclaim responsibility.
In summary, during the last decade the economic theory of warranties has
profited greatly from the development of models with incomplete information
and moral hazard. The insurance, signalling, and incentive motives all provide
good reasons for the existence of warranties. The incorporation of countervailing
effects which are due to incomplete information about the buyers’ side provides
WARRANTY CONTRACTS 55
Acknowledgements
I wish to thank Helmut Bester, Martin Hellwig, John Horowitz, and an anonymous
referee for helpful comments. The usual disclaimer applies. Financial support by the
Deutsche Forshungsgemeinschaft through SFB 303 and Grant Em 39/1-1, and the
hospitality of the Dept. of Economics at U.C. San Diego is gratefully acknowledged.
Notes
I . Priest (1981) even claims that warranties are the most common o f written contracts.
2. Lutz (1985) defines any contract that is optional as a service contract. An optional
contract specifying that the labour costs necessary to repair the good are borne by the
manufacturer is an optional warranty according to our definition and a service
contract according to Lutz's definition. See Mann and Wissink (1986) for an analysis
of money-back guaranties.
3 . In our model the product's contingencies are measured in monetary units. Therefore,
the different warranty coverages that are mentioned in the Introduction can be
expressed in terms of the monetary warranty w . Throughout this paper we will assume
that no disputes arise about whether a product is defective or not. See Palfrey and
Romer (1983) for a discussion of possible dispute resolution mechanisms.
4. In a multiperiod framework, Brown (1974) shows that the warranty duration equals
the lifetime of the product.
5 . Gerner and Bryant (1981) incorporate an information processing cost to consumers
in a framework where warranties act as signals. Consumers purchase durables only
occasionally. The cost o f understanding a specific warranty contract is high.
Therefore, warranty terms across different markets should be similar, each taking
advantage of the greater benefit to consumers of general rather than specific
56 EMONS
References
Akerlof, G. (1970) The market for ‘Lemons’: quality uncertainty and the market
mechanism, Quarterly Journal of Economics 84, 488-500.
Braverman, A , , Guasch, J. L. and Salop, S. (1983) Defects in Disneyland: quality control
as a two-part tariff, Review of Economic Sfudies 50, 121-31.
Brown, J . P. (1974) Product liability: the case of an asset with random life, American
Economic Review 64, 149-61.
Cooper, R. and Ross, T. W. (1985) Product warranties and double moral hazard, Rand
Journal of Economics 16, 103-13.
Emons, W. (1989) On the limitation of warranty duration, Journal of lndusrrial
Economics 37, 287-301.
- (1988) Warranties, moral hazard, and the lemons problem, Journal of Economic
Theory 46, 16-33.
Gerner, J . and Bryant, K. (1981) Appliance warranties as a market signal, Journal of
Consumer Affairs 5 1, 75-86.
Grossman, S . J . (1981) The informational role of warranties and private disclosure about
product quality, Journal of Law and Economics 24, 461-83.
Heal, G. (1977) Guaranties and risk-sharing, Review of Economic Studies 44, 549-60.
WARRANTY CONTRACTS 57
Holmes, T. J . (1984) Monopoly bundling of warranty and product quality when quality
is unobservable, CMSEMS Discussion Paper No. 612s, Northwestern University,
Illinois.
Kessler, F. (1943) Contracts of adhesion - some thoughts about freedom of contract,
Columbia Law Review, 43, 629-43.
Kubo, Y. (1986) Quality uncertainty and guarantee - a case of strategic market
segmentation by a monopolist, European Economic Review 30, 1063-79.
Lutz, N . (1985) Discrete warranties, signalling and consumer moral hazard, mimeo,
Graduate School of Business, Stanford University.
Mann, D. P . and Wissink, J . P. (1986) Money back guaranties, mimeo, Cornell
University, Ithaca, New York.
Matthews, S. and Moore, J . (1987) Monopoly provision of quality and warranties: an
exploration in the theory of multidimensional screening, Economerrica 55, 441 -67.
McKean, R. (1970) Products liability: implications of some changing property rights,
Quarlerly Journal of Econotnics 84, 61 1-26.
Mussa, M. and Rosen, S. (1978) Monopoly and product quality, Journal of Economic
Theory 18, 301-17.
Oi, W. (1973) The economics of product safety, Bell Journal of Economics 4, 3-28.
Palfrey, T. and Romer, T. (1983) Warranties, performance, and the resolution of buyer-
seller disputes, Bell Journal of Econotnics 14, 97-1 17.
Priest, G. (1981) A theory of the consumer product warranty, Yale Law Journal 90,
1297-352.
Prosser, W . L. (1943) The implied warranty of merchantable quality, Minnesora Law
Review 27, 117-68.
Rothschild, M. and Stiglitz, J . E. (1976) Equilibrium in competitive insurance markets:
an essay on the economics of imperfect information, Quarrerly Journal of Economics
90, 659-79.
Salop, S . and Stiglitz, J . E. (1977) Bargains and ripoffs: a model of monopolistically
competitive price dispersion, Review of Economic Sludies 44, 493-5 10.
Schwartz, A. and Wilde, L. L. (1983) Imperfect information in markets for contract
terms: the examples of warranties and security interests, Virginia Law Review 69,
1387-485.
Sharpiro, C. (1983) Premiums for high quality products as returns to reputation,
Quarterly Journal of Economics 97, 659-79.
Spence, M. (1977) Consumer misperceptions, product failure and producer liability,
Review of Economic Sludies 44, 561-72.
Varian, H . R . (1980) A model of sales, American Economic Review 70, 651-9.
Wilde, L. L. and Schwartz, A. (1979) Equilibrium comparison shopping, Review of
Economic Studies 46, 543-54.
Wilson, C. (1977) A model of insurance markets with incomplete information, Journal
of Economic Theory 16, 167-207.