The Market for "Lemons": Quality Uncertainty and the Market Mechanism
Author(s): George A. Akerlof
Reviewed work(s):
Source: The Quarterly Journal of Economics, Vol. 84, No. 3 (Aug., 1970), pp. 488-500
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/1879431 .
Accessed: 13/10/2012 23:44
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The Quarterly
Journal of Economics.
http://www.jstor.org
THE MARKET FOR "LEMONS":
QUALITY UNCERTAINTY AND THE
MARKET MECHANISM *
GEORGE A. AKERLOF
I. Introduction, 488.-II. The model with automobiles as an example,
489.- III. Examples and applications, 492.- IV. Counteracting institutions,
499. -V. Conclusion, 500.
I. INTRODUCrION
This paper relates quality and uncertainty. The existence of
goods of many grades poses interesting and important problems for
the theory of markets. On the one hand, the interaction of quality
differences and uncertainty may explain important institutions of
the labor market. On the other hand, this paper presents a strug-
gling attempt to give structure to the statement: "Business in under-
developed countries is difficult"; in particular, a structure is given
for determining the economic costs of dishonesty. Additional appli-
cations of the theory include comments on the structure of money
markets, on the notion of "insurability," on the liquidity of dur-
ables, and on brand-name goods.
There are many markets in which buyers use some market
statistic to judge the quality of prospective purchases. In this case
there is incentive for sellers to market poor quality merchandise,
since the returns for good quality accrue mainly to the entire group
whose statistic is affected rather than to the individual seller. As
a result there tends to be a reduction in the average quality of goods
and also in the size of the market. It should also be perceived that
in these markets social and private returns differ, and therefore, in
some cases, governmental intervention may increase the welfare of
all parties. Or private institutions may arise to take advantage
of the potential increases in welfare which can accrue to all parties.
By nature, however, these institutions are nonatomistic, and there-
fore concentrations of power - with ill consequences of their own -
can develop.
*The author would especially like to thank Thomas Rothenberg for
invaluable comments and inspiration. In addition he is indebted to Roy
Radner, Albert Fishlow, Bernard Saffran, William D. Nordhaus, Giorgio La
Malfa, Charles C. Holt, John Letiche, and the referee for help and sugges-
tions. He would also like to thank the Indian Statistical Institute and the
Ford Foundation for financial support.
MARKET FOR "LEMONS": AND MARKET MECHANISM 489
The automobile market is used as a finger exercise to illustrate
and develop these thoughts. It should be emphasized that this mar-
ket is chosen for its concreteness and ease in understanding rather
than for its importance or realism.
II. THE MODEL WITH AUTOMOBILES AS AN EXAMPLE
A. The Automobiles Market
The example of used cars captures the essence of the problem.
From time to time one hears either mention of or surprise at the
large price difference between new cars and those which have just
left the showroom. The usual lunch table justification for this
phenomenon is the pure joy of owning a "new" car. We offer a
different explanation. Suppose (for the sake of clarity rather than
reality) that there are just four kinds of cars. There are new cars
and used cars. There are good cars and bad cars (which in America
are known as "lemons"). A new car may be a good car or a lemon,
and of course the same is true of used cars.
The individuals in this market buy a new automobile without
knowing whether the car they buy will be good or a lemon. But they
do know that with probability q it is a good car and with probability
(1-q) it is a lemon; by assumption, q is the proportion of good
cars produced and (1 - q) is the proportion of lemons.
After owning a specific car, however, for a length of time, the
car owner can form a good idea of the quality of this machine; i.e.,
the owner assigns a new probability to the event that his car is a
lemon. This estimate is more accurate than the original estimate.
An asymmetry in available information has developed: for the
sellers now have more knowledge about the quality of a car than
the buyers. But good cars and bad cars must still sell at the same
price -since it is impossible for a buyer to tell the difference be-
tween a good car and a bad car. It is apparent that a used car can-
not have the same valuation as a new car - if it did have the same
valuation, it would clearly be advantageous to trade a lemon at
the price of new car, and buy another new car, at a higher prob-
ability q of being good and a lower probability of being bad. Thus
the owner of a good machine must be locked in. Not only is it
true that he cannot receive the true value of his car, but he cannot
even obtain the expected value of a new car.
Gresham's law has made a modified reappearance. For most
cars traded will be the "lemons," and good cars may not be traded
at all. The "bad" cars tend to drive out the good (in much the
490 QUARTERLY JOURNAL OF ECONOMICS
same way that bad money drives out the good). But the analogy
with Gresham's law is not quite complete: bad cars drive out the
good because they sell at the same price as good cars; similarly, bad
money drives out good because the exchange rate is even. But the
bad cars sell at the same price as good cars since it is impossible
for a buyer to tell the difference between a good and a bad car;
only the seller knows. In Gresham's law, however, presumably both
buyer and seller can tell the difference between good and bad
money. So the analogy is instructive, but not complete.
B. Asymmetrical Information
It has been seen that the good cars may be driven out of the
market by the lemons. But in a more continuous case with different
grades of goods, even worse pathologies can exist. For it is quite
possible to have the bad driving out the not-so-bad driving out the
medium driving out the not-so-good driving out the good in such
a sequence of events that no market exists at all.
One can assume that the demand for used automobiles depends
most strongly upon two variables - the price of the automobile p
and the average quality of used cars traded, a, or Qd = D (p, A). Both
the supply of used cars and also the average quality p will depend
upon the price, or p=j (p) and S=S(p). And in equilibrium the
supply must equal the demand for the given average quality, or
S(p) = D (p, p (p)). As the price falls, normally the quality will
also fall. And it is quite possible that no goods will be traded at
any price level.
Such an example can be derived from utility theory. Assume
that there are just two groups of traders: groups one and two. Give
group one a utility function
U1=M+ iXi
_.1
where M is the consumption of goods other than automobiles, x4
is the quality of the ith automobile, and n is the number of auto-
mobiles.
Similarly, let
U2 = M+ X 3/2x4
i.i
where M, xi, and n are defined as before.
Three comments should be made about these utility func-
tions: (1) without linear utility (say with logarithmic utility) one
gets needlessly mired in algebraic complication. (2) The use of
MARKET FOR "LEMONS": AND MARKET MECHANISM 491
linear utility allows a focus on the effects of asymmetry of informa-
tion; with a concave utility function we would have to deal jointly
with the usual risk-variance effects of uncertainty and the special
effects we wish to discuss here. (3) U1 and U2 have the odd char-
acteristic that the addition of a second car, or indeed a kth car,
adds the same amount of utility as the first. Again realism is sacri-
ficed to avoid a diversion from the proper focus.
To continue, it is assumed (1) that both type one traders and
type two traders are von Neumann-Morgenstern maximizers of
expected utility; (2) that group one has N cars with uniformly
distributed quality x, 0
l
Di=O =/p p
D2 =0 3u/2 < p
and
S2 =0.
Thus total demand D (p, u) is
D (p, u) = (Y2+ Y1)/P if p
D (p, ) = Y2/p if u< p <3u/2
D(p, y =0 if p>3u/2.
However, with price p, average quality is p/2 and therefore at no
price will any trade take place at all: in spite of the fact that at
any given price between 0 and 3 there are traders of type one who
are willing to sell their automobiles at a price which traders of type
two are willing to pay.
492 QUARTERLY JOURNAL OF ECONOMICS
C. Symmetric Information
The foregoing is contrasted with the case of symmetric infor-
mation. Suppose that the quality of all cars is uniformly distributed,
O1
S(p)=O p<1.
And the demand curves are
D(p) = (Y2+Yl)/P p<1
D(p) = (Y2/p) l 3/2.
In equilibrium
(3) p=1 if Y2 Y2, in which case the income of type
two traders is insufficient to buy all N automobiles, there is a gain
in utility of Y2/2 units.)
Finally, it should be mentioned that in this example, if traders
of groups one and two have the same probabilistic estimates about
the quality of individual automobiles - though these estimates may
vary from automobile to automobile - (3), (4), and (5) will still
describe equilibrium with one slight change: p will then represent
the expected price of one quality unit.
III. EXAMPLES AND APPLICATIONS
A. Insurance
It is a well-known fact that people over 65 have great difficulty
in buying medical insurance. The natural question arises: why
doesn't the price rise to match the risk?
Our answer is that as the price level rises the people who in-
sure themselves will be those who are increasingly certain that they
will need the insurance; for error in medical check-ups, doctors'
sympathy with older patients, and so on make it much easier for
the applicant to assess the risks involved than the insurance com-
pany. The result is that the average medical condition of insurance
applicants deteriorates as the price level rises -with the result
MARKET FOR "LEMONS": AND MARKET MECHANISM 493
that no insurance sales may take place at any price.' This is strictly
analogous to our automobiles case, where the average quality of
used cars supplied fell with a corresponding fall in the price level.
This agrees with the explanation in insurance textbooks:
Generally speaking policies are not available at ages materially greater
than sixty-five.... The term premiums are too high for any but the most
pessimistic (which is to say the least healthy) insureds to find attractive. Thus
there is a severe problem of adverse selection at these ages.2
The statistics do not contradict this conclusion. While de-
mands for health insurance rise with age, a 1956 national sample
survey of 2,809 families with 8,898 persons shows that hospital
insurance coverage drops from 63 per cent of those aged 45 to 54,
to 31 per cent for those over 65. And surprisingly, this survey also
finds average medical expenses for males aged 55 to 64 of $88,
while males over 65 pay an average of $77.3 While noninsured ex-
penditure rises from $66 to $80 in these age groups, insured expen-
diture declines from $105 to $70. The conclusion is tempting that
insurance companies are particularly wary of giving medical in-
surance to older people.
The principle of "adverse selection" is potentially present in
all lines of insurance. The following statement appears in an in-
surance textbook written at the Wharton School:
There is potential adverse selection in the fact that healthy term in-
surance policy holders may decide to terminate their coverage when they be-
come older and premiums mount. This action could leave an insurer with an
undue proportion of below average risks and claims might be higher than
anticipated. Adverse selection "appears (or at least is possible) whenever the
individual or group insured has freedom to buy or not to buy, to choose the
amount or plan of insurance, and to persist or to discontinue as a policy
holder." '
Group insurance, which is the most common form of medical
insurance in the United States, picks out the healthy, for generally
1. Arrow's fine article, "Uncertainty and Medical Care" (American Eco-
nomic Review, Vol. 53, 1963), does not make this point explicitly. He em-
phasizes "moral hazard" rather than "adverse selection." In its strict sense,
the presence of "moral hazard" is equally disadvantageous for both govern-
mental and private programs; in its broader sense, which includes "adverse
selection," "moral hazard" gives a decided advantage to government insurance
programs.
2. 0. D. Dickerson, Health Insurance (Homewood, Ill.: Irwin, 1959),
p. 333.
3. 0. W. Anderson (with J. J. Feldman), Family Medical Costs and In-
surance (New York: McGraw-Hill, 1956).
4. H. S. Denenberg, R. D. Eilers, G. W. Hoffman, C. A. Kline, J. J.
Melone, and H. W. Snider, Risk and Insurance (Englewood Cliffs, N. J.:
Prentice Hall, 1964), p. 446.
494 QUARTERLY JOURNAL OF ECONOMICS
adequate health is a precondition for employment. At the same
time this means that medical insurance is least available to those
who need it most, for the insurance companies do their own "ad-
verse selection."
This adds one major argument in favor of medicare.5 On a
cost benefit basis medicare may pay off: for it is quite possible that
every individual in the market would be willing to pay the ex-
pected cost of his medicare and buy insurance, yet no insurance
company can afford to sell him a policy - for at any price it will
attract too many "lemons." The welfare economics of medicare, in
this view, is exactly analogous to the usual classroom argument
for public expenditure on roads.
B. The Employment of Minorities
The Lemons Principle also casts light on the employment of
minorities. Employers may refuse to hire members of minority
groups for certain types of jobs. This decision may not reflect ir-
rationality or prejudice -but profit maximization. For race may
serve as a good statistic for the applicant's social background,
quality of schooling, and general job capabilities.
Good quality schooling could serve as a substitute for this
statistic; by grading students the schooling system can give a
better indicator of quality than other more superficial character-
istics. As T. W. Schultz writes, "The educational establishment
discovers and cultivates potential talent. The capabilities of chil-
dren and mature students can never be known until found and culti-
vated." 6 (Italics added.) An untrained worker may have valuable
natural talents, but these talents must be certified by "the educa-
tional establishment" before a company can afford to use them. The
certifying establishment, however, must be credible; the unrelia-
bility of slum schools decreases the economic possibilities of their
students.
This lack may be particularly disadvantageous to members of
5. The following quote, again taken from an insurance textbook, shows
how far the medical insurance market is from perfect competition:
is
. .insurance companies must screen their applicants. Naturally it
is true that many people will voluntarily seek adequate insurance on
their own initiative. But in such lines as accident and health insurance,
companies are likely to give a second look to persons who voluntarily seek
insurance without being approached by an agent." (F. J. Angell, Insur-
ance, Principles and Practices, New York: The Ronald Press, 1957, pp.
8-9.)
This shows that insurance is not a commodity for sale on the open market.
6. T. W. Schultz, The Economic Value of Education (New York: Colum-
bia University Press, 1964), p. 42.
MARKET FOR "LEMONS": AND MARKET MECHANISM 495
already disadvantaged minority groups. For an employer may
make a rational decision not to hire any members of these groups
in responsible positions - because it is difficult to distinguish those
with good job qualifications from those with bad qualifications.
This type of decision is clearly what George Stigler had in mind
when he wrote, "in a regime of ignorance Enrico Fermi would have
been a gardener, Von Neumann a checkout clerk at a drugstore." 7
As a result, however, the rewards for work in slum schools
tend to accrue to the group as a whole -in raising its average
quality -rather than to the individual. Only insofar as informa-
tion in addition to race is used is there any incentive for training.
An additional worry is that the Office of Economic Opportunity
is going to use cost-benefit analysis to evaluate its programs. For
many benefits may be external. The benefit from training minority
groups may arise as much from raising the average quality of the
group as from raising the quality of the individual trainee; and,
likewise, the returns may be distributed over the whole group rather
than to the individual.
C. The Costs of Dishonesty
The Lemons model can be used to make some comments on
the costs of dishonesty. Consider a market in which goods are
sold honestly or dishonestly; quality may be represented, or it may
be misrepresented. The purchaser's problem, of course, is to identify
quality. The presence of people in the market who are willing to
offer inferior goods tends to drive the market out of existence - as
in the case of our automobile "lemons." It is this possibility that
represents the major costs of dishonesty - for dishonest dealings
tend to drive honest dealings out of the market. There may be
potential buyers of good quality products and there may be po-
tential sellers of such products in the appropriate price range;
however, the presence of people who wish to pawn bad wares as
good wares tends to drive out the legitimate business. The cost of
dishonesty, therefore, lies not only in the amount by which the
purchaser is cheated; the cost also must include the loss incurred
from driving legitimate business out of existence.
Dishonesty in business is a serious problem in underdeveloped
countries. Our model gives a possible structure to this statement
and delineates the nature of the "external" economies involved.
In particular, in the model economy described, dishonesty, or the
7. G. J. Stigler, "Information and