1
[ Journal of Political Economy, 2006, vol. 114, no. 1]
᭧ 2006 by The University of Chicago. All rights reserved. 0022-3808/2006/11401-0001$10.00
The Behavioralist Meets the Market: Measuring
Social Preferences and Reputation Effects in
Actual Transactions
John A. List
University of Chicago and National Bureau of Economic Research
The role of the market in mitigating and mediating various forms of
behavior is perhaps the central issue facing behavioral economics today.
This study designs a field experiment that is explicitly linked to
a controlled laboratory experiment to examine whether, and to what
extent, social preferences influence outcomes in actual market transactions.
While agents drawn from a well-functioning marketplace behave
in accord with social preference models in tightly controlled
laboratory experiments, when they are observed in their naturally
occurring settings, their behavior approaches what is predicted by selfinterest
theory. In the limit, much of the observed behavior in the
marketplace that is consistent with social preferences is due to reputational
concerns: suppliers who expect to have future interactions
with buyers provide higher product quality only when the buyer can
verify quality via a third-party certifier. The data also speak to theories
of how reputation effects enhance market performance. In particular,
reputation and the monitoring of quality are found to be complements,
and findings suggest that the private market can solve the lemons
problem through third-party verification.
Orley Ashenfelter, Raymond Battalio, Roland Benabou, Daniel Benjamin, Gary Charness,
Armin Falk, Edward Glaeser, Uri Gneezy, Glenn Harrison, Daniel Kahneman, Liesl
Koch, David Laibson, Steven Levitt, Derek Neal, Matthew Rabin, and Al Roth provided
remarks on an earlier version of this study that considerably improved the paper. Seminar
participants at the University of Chicago (Graduate School of Business and the Economics
Department), University of California at Berkeley, Cornell University, Harvard University,
Princeton University, Texas A&M, University of Texas at Austin, and the University of
Wisconsin provided comments that helped to shape the paper. Thanks to Michael Price
for research assistance.
2 journal of political economy
I. Introduction
More than two decades ago, Stigler (1981, 176) wrote that when “selfinterest
and ethical values with wide verbal allegiance are in conflict,
much of the time, most of the time in fact, self-interest theory . . . will
win.” While this is the conventional wisdom among economists, an influential
collection of laboratory experiments on “gift exchange” has
called into question the validity of Stigler’s position (see, e.g., Camerer
and Weigelt 1988; Fehr, Kirchsteiger, and Riedl 1993; Berg, Dickhaut,
and McCabe 1995). This literature is complemented by an entire body
of theoretical research exploring the economic consequences of “social
preferences,” wherein agents have preferences that are measured over
their own and others’ material payoffs (for models of reciprocity, see
Rabin [1993], Charness and Rabin [2002], Dufwenberg and Kirchsteiger
[2004], and Falk and Fischbacher [forthcoming]; for models of inequity
aversion, see Fehr and Schmidt [1999] and Bolton and Ockenfels
[2000]; on altruism, see Andreoni and Miller [2002]).1
A second generation
of experimental studies has subsequently emerged examining
the nature of social preferences and underscoring the robustness of the
gift exchange results (e.g., Charness 1996; Fehr, Ga¨chter, and Kirchsteiger
1997; Fehr and Falk 1999; Charness and Rabin 2002; Ga¨chter
and Falk 2002; Hannan, Kagel, and Moser 2002; Brown, Falk, and Fehr
2004; Fehr and List 2004).2
The gift exchange results, which are consistent with the notion that
people behave in a reciprocal manner even when the behavior is costly
and yields neither present nor future material rewards, have attracted
much attention, since many have argued that they are relevant beyond
the context inherent in the laboratory. For example, some view the
experimental results as providing key support for the labor market predictions
in Akerlof (1982) and Akerlof and Yellen (1988, 1990), whereby
higher than market-clearing wages and involuntary unemployment are
potential outcomes of fairness considerations in the workplace.3
Indeed,
Fehr et al. (1993, 437) note that their results “provide . . . experimental
1
In this study, I explore social preferences under this broad definition and am not
interested in pinpointing whether the behavior consistent with social preferences is due
to altruism, reciprocity, fairness, inequality aversion, or another motive. Yet within the gift
exchange literature, reciprocity motives have been highlighted; thus I shall continue this
spirit in the discussion below. For a parsing of trust and reciprocity in a laboratory experiment,
see Cox (2004).
2
Fehr and Ga¨chter (2000) provide an overview. The interested reader should also see
the related literature on “lemons” markets (e.g., Miller and Plott 1985; Holt and Sherman
1990).
3
This conjecture is typically termed the “fair wage–effort” hypothesis. Alternatively, note
that the “efficiency wage theory” surmises that wages above market-clearing levels occur
because these wage profiles induce workers to be motivated in an effort to avoid being
fired, which economizes on firm-level monitoring (see, e.g., Katz 1986).
social preferences 3
support for the fair wage–effort theory of involuntary unemployment.”
Of course, social preferences might be important in many other strategic
situations as well (for overviews, see, e.g., Sobel [2002] and Camerer
[2003]), and therefore such results have broad implications for economists
and noneconomists alike. Despite these advances and the topic’s
importance, it is fair to say that little is known about whether, and to
what extent, social preferences influence economic outcomes in naturally
occurring markets.4
The major goals of this study are to explore the nature of such preferences
among real-market players in both the laboratory and naturally
occurring environments. In doing so, the study provides a framework
with which to disentangle social preferences and reputation effects. Measuring
and disentangling social preferences and reputation effects are
important in both a positive and a normative sense, since optimal contracting
and proposed government intervention in principal-agent settings,
appropriate design of collective choice mechanisms, and theory
testing all depend critically on proper measurement of these effects.
Equally important, the experimental design permits an examination of
whether individual behavior in laboratory experiments provides a reliable
indicator of behavior in the field—an issue fundamental to experimental
economics.
To complete these tasks, I use several distinct experimental treatments
to create a bridge between the laboratory and the field. A major attraction
of this approach is that if behavioral differences are observed,
I can pinpoint the important factors driving the disparities. I begin with
a gift exchange laboratory treatment (which is in effect a sequential
prisoner’s dilemma game) that closely follows the received literature
(e.g., Fehr et al. 1993). Rather than using student subjects, however, I
make use of subjects drawn from a well-functioning marketplace—the
sports card market. In this setting, I place the experimental participants
in their typical roles: consumers are placed in the role of buyers and
dealers are placed in the role of sellers. This treatment is potentially
important in that a fundamental feature of markets is that they sort
agents into roles, whereas laboratory experiments randomly assign positions
to agents. Experimental results are nevertheless consistent with
those in the literature that uses students randomly allocated to roles:
gift exchange is observed, and such behavior has an important influence
4
There is some survey evidence reported from interviews with managers that social
preference considerations are important in the workplace (Blinder and Choi 1990; Bewley
1995). Furthermore, in a novel paper exploring the role of fairness in the marketplace,
Kahneman, Knetsch, and Thaler (1986) report results from telephone surveys of residents
of two Canadian metropolitan areas (Toronto and Vancouver). They use a “dual entitlement”
theory to explain their data: previous transactions establish a reference level of
consumer and producer surplus, and fairness considerations arise from outcomes relative
to these “entitlements.”
4 journal of political economy
on economic outcomes. This finding provides a validity check of the
extant laboratory results on gift exchange, since it suggests that the
major results can be replicated with real economic players from a much
different population.
I proceed to explore several further treatments in the laboratory by
varying the “context” in the experimental instructions; previous experimental
studies typically use “context-free” instructions, such as neutral
wording and the avoidance of words that might provide familiar contextual
cues. Of course, this traditional approach potentially attenuates
important elements of the exchange process and therefore may suppress
important psychological effects. A final laboratory treatment in this spirit
moves a step toward the naturally occurring marketplace by creating an
experimental lab market in which buyers and sellers play a sequential
prisoner’s dilemma game by exchanging cash for goods of uncertain
quality in face-to-face transactions. If one ignores the artificiality invoked
by the laboratory experimental setting, this particular treatment provides
an environment that mirrors the actual decision-making process
in the marketplace from which these subjects are drawn. As a whole,
these design changes yield some behavioral differences, but gift exchange
in these settings remains alive and well, both statistically and
economically.
When one moves from the lab to the field, an important consideration
is to remain parallel to the important lab features while ensuring that
the transaction is a natural one in the field. The field experimental
treatments mirror the laboratory gift exchange treatments and resemble
many types of markets for goods or services: after receiving a price offer,
sellers determine the good’s quality, which cannot be perfectly measured
by buyers. In the first field treatment, subjects approach dealers (who
are unaware that they are taking part in an experiment) and offer either
$20 or $65 for a sports card of certain quality. Since quality is difficult
for untrained consumers to detect in this market and the approached
dealers have a sufficient stock of cards on hand to provide the requested
quality levels, if social preferences play a role in this case, the card’s
grade and the price offer should be positively correlated. Once the
buying agents had purchased each of the cards from the dealers, I had
every card professionally graded. I do find a positive correlation between
the prices and grades received, but only among dealers who are “locals”;
among dealers who are likely to have little future interaction with the
buying agents (“nonlocals”), no such relationship emerges.
This result is interesting, but the data do not allow an unequivocal
insight into the underlying mechanism at work. For instance, such data
patterns might be due to several factors, including two competing alternatives
critical to the issue at hand: selection effects—local dealers
have social preferences and nonlocal dealers do not—or reputation
social preferences 5
effects—local dealers are concerned with their reputations whereas nonlocal
dealers are not. A final set of three treatments in the marketplace
provide insights into what is driving these behavioral differences by
examining outcomes in an identical experiment for collector tickets
and ticket stubs. Tickets and ticket stubs provide a unique test because
no third-party verification service existed to grade tickets until June 2003,
though the major grading company announced in April 2003 that it
would soon begin grading tickets. By comparing temporal outcomes,
not only am I permitted a unique opportunity to examine the nature
of market exchanges with and without third-party enforcement, but I
am also able to explore the role of social preferences in such settings.5
In stark contrast to the results obtained from the sports card data,
the empirical results in the time period in which no grading service was
available and the public was unaware that a service was imminent (pre–
April 2003) provide little evidence consistent with social preferences:
ticket quality is not correlated with price for either dealer type, and
local and nonlocal dealers provide similar quality levels. One could
reason that dealers had little idea how to grade tickets since they had
never been professionally graded (though many dealers made quality
claims), and therefore the inability of this treatment to reject the homogeneity
null is consistent with informational problems.
This potential drawback is rectified in an experimental treatment
conducted after the announcement of grading (April 2003) but before
the grading company released its grading criteria (June 2003). Purchasing
identical tickets and using analogous protocol, I find that during
this time period quality and price are correlated for tickets sold by local
dealers, but no correlation is present in ticket sales among nonlocal
dealers. Completing the experimental design is an identical treatment
conducted after grading services commenced (post–June 2003). Insights
gained from this treatment are quite similar to those obtained from the
treatment conducted between April and June: gift exchange is evident
among local dealers but not among nonlocals. This result stands to
reason because Professional Sports Authenticators’ (PSA’s) ticket grading
criteria are similar to its scheme for grading sports cards.
In summary, several insights follow. First, even though the data collected
from one-shot laboratory experiments suggest that social preferences
are quite important among these agents, parallel treatments in
5
Brown et al. (2004, 751) summarize the attractiveness of such treatments in motivating
their laboratory experiments by noting that “the ideal data set for studying the effects of
the absence of third party enforceability on market interactions . . . is based on a truly
exogenous ceteris paribus variation in the degree of third party enforceability. . . . The
problem is, however, that it seems almost impossible to find or generate field data that
approximates this ideal data set.” This is exactly what these three treatments offer, and
to the best of my knowledge such exogeneity has not heretofore been achieved in the
literature.
6 journal of political economy
the field suggest that such effects have minimal influence in naturally
occurring transactions. In this sense, dealer behavior in the marketplace
approaches what is predicted by self-interest theory. From a methodological
viewpoint, it is important to note that several changes to the
laboratory environment had little influence on behavior, whereas moving
from the lab to the field had striking effects. Second, empirical
results provide insights into how reputation effects and professional
certification influence market performance (see, e.g., Akerlof 1970;
Klein and Leffler 1981). For example, I find that (i) reputation effects
enhance the quality of goods, and (ii) reputation and the monitoring
of quality are complements. In this spirit, the data suggest that the private
market can solve the lemons problem through third-party verification.
The remainder of this study is organized as follows. Section II describes
the experimental design and summarizes the institutional details
of the market. Section III provides a summary of the empirical findings,
highlighting differences in results across the various treatments and
describing the effects of reputation and social preferences on market
outcomes across both local and nonlocal dealers. Section IV concludes
with a more general discussion of the empirical results.
II. Experimental Design and Hypotheses
The experimental investigation begins with an examination of behavior
in standard laboratory gift exchange games. Treatment Lab-R (R denotes
laboratory replication; see table 1 for a summary of the experimental
design) makes use of the typical gift exchange experimental design.6
One session was run in this treatment. In this session, each participant’s
experience typically followed four steps: (1) consideration of the invitation
to participate in an experiment, (2) a session to learn the experimental
rules, (3) actual participation, and (4) conclusion of the
experiment and exit interview. In step 1, the monitor approached dealers
on the floor of a sports card show and inquired about their interest
in participating in an economics experiment that would take about an
hour. If the dealer agreed, the monitor summarized the meeting time
and place. A similar approach was used to recruit consumers (non-
dealers).
Subjects met in a large room adjacent to the floor of the sports card
show: dealers entered on one side of the room and nondealers on the
other side, and a divider was in place to ensure that identities were not
revealed. The session consisted of five periods, with five dealers acting
as sellers and five nondealers acting as buyers. Each participant received
6
Appendix A in List (2005) contains a copy of the experimental instructions, which
are closely related to those in Fehr et al. (1993, 1997) and Ga¨chter and Falk (2002).
social preferences 7
TABLE 1
Experimental Design
(1) (2) (3)
Treatment Lab Treatment Lab-R:
Replicate lab studies
N p 25
Treatment Lab-RF:
Extend to field
values
N p 25
Treatment Lab-RF1:
Extend to one-shot
environment
N p 27
Treatment Lab-
Context
Treatment Lab-Con-
text:
Adds market
context
N p 32
Treatment Lab-Mar-
ket($20):
Adds market
interaction
N p 30
Treatment Lab-Mar-
ket($65):
Adds market
interaction
N p 30
Treatment Floor
(Cards)
Treatment Floor-
$20:
Naturally occurring
sports card
market
N p 50
Treatment Floor-
$65:
Naturally occurring
sports card
market
N p 50
Treatment Floor
(Tickets)
Treatment Floor-
NoGrading:
Naturally occurring
ticket market before
grading was
available
N p 60
Treatment Floor-
AnnounceGrading:
Naturally occurring
ticket market after
grading
announcement
N p 54
Treatment Floor-
Grading:
Naturally occurring
ticket market
when grading service
was available
N p 36
Note.—Each cell represents one (or two, in the case of Treatment Floor [Tickets]) unique treatment. For example,
Treatment Lab-R in row 1, col. 1, denotes that 25 dealer and 25 nondealer observations were gathered to replicate the
laboratory gift exchange studies in the literature.
a copy of the instructions, and to ensure common information, the
monitor read the instructions aloud as the subjects followed along. The
instructions noted that in each of the five periods each buyer would be
paired with a different seller. In every period, the buyer determines an
integer value (denoted p for price) to send to the seller and requests
a specific quality of the good (denoted for quality request). Only theqr
seller who is paired with the buyer is aware of these two choices. After
the buyer makes these private decisions on the decision sheet, the monitor
collects the sheets and walks them to the seller partners. Sellers
then choose a quality level (denoted q for quality chosen), with an
associated cost of quality that is increasing monotonically with product
quality (denoted ; see List [2005, app. A] for the cost of productc(q)
quality parameters, which closely follow the literature). The product
quality choice is revealed only to the buyer partner, and as in the literature,
all choices are revealed to the monitor.
Individual p and q choices combine to determine monetary payoffs
for the pair according to the following payoff functions:
seller payoff: P p p Ϫ c(q),s
buyer payoff: P p (v Ϫ p)q, v p $80, p ෈ [$5, $80], q ෈ [0.1, 1]. (1)b
8 journal of political economy
All payoff information was common information, and before the experiment
began, several hypothetical exercises were completed to ensure
that everyone understood the instructions and payoff functions.
Subjects were also aware that one of the five periods would be selected
randomly and that that particular period would determine payoffs. After
the fifth period, subjects were paid in private after they completed a
survey (see List 2005, app. B).
These parameter values yield a standard prediction under the assumption
of common knowledge, self-interest theory, and appropriate
backward induction. Since product quality is costly, sellers will choose
the minimum level ( ). A buyer’s best response is to chooseq p 0.1min
pmin, which is . Thus the subgame perfect equilibrium outcomep p $5
is and , with associated profits of andq* p 0.1 p* p $5 P p $5 P ps b
, much less than more efficient profit levels (i.e., and$7.50 p p $30
yields and ).q p 0.5 P p $24 P p $25s b
Previous experimental efforts have found that typically andq 1 q*
and that in a reduced-form regression model, leadingp 1 p* Ѩq/Ѩp 1 0
authors to conclude that reciprocity is important in economic interactions.
The reciprocity inference is generally traced to Rabin’s (1993)
model of reciprocity (Fehr et al. 1997, 839), which describes a person
with positive reciprocal motives as someone who responds to acts that
are perceived as kind in a kind manner, even though there is no future
pecuniary gain tied to this action. For the purposes herein, the literature
has taken the qualitative implications of Rabin’s model as meaning that
the probability of nonshirking is increasing in the level of the perceived
generosity of the offer. How generous an action is perceived is a difficult
question to answer, however, and is surely quite heterogeneous across
agents, inducing the literature to operationalize reciprocity as meaning
that in a reduced-form regression model (see, e.g., Fehr et al.Ѩq/Ѩp 1 0
1993; Ga¨chter and Falk 2002).
In column 2 of table 1, Treatment Lab-RF (RF denotes replication
with field values) simply manipulates the environment in Treatment
Lab-R by setting7
seller payoff: P p p Ϫ c(q),s
buyer payoff: P p v(q) Ϫ p, p ෈ [$5, $80], q ෈ [1, 5]. (2)b
For values, I use , $5, $8, $15, and $50 for , 2, 3, 4,c(q) c(q) p $4 q p 1
and 5; for values, I use $6, $8, $15, $30, and $80 for , 2, 3,v(q) q p 1
7
The payoff function for the buyer is now similar to the S13–S16 treatment in Fehr et
al. (1997). In this case, now the price represents a pure lump-sum transfer, which differs
from the earlier joint profit equation, which was characterized by price increases leading
to an increase in the sum of payoffs when .q ! 1
social preferences 9
4, and 5 (PSA 6, 7, 8, 9, and 10).8
While these chosen values are admittedly
only a rough estimate of the gains to trade available in this
market, use of these parameters provides the necessary tension between
the dominant strategy and the joint-profit maximization actions. Under
this design, the Nash purely selfish prediction is ; and for sellersp* p $5
to send minimal card quality, . These actions result inq* p 1 P p $1s
and . The efficient quality level is , which ensures a jointP p $1 q p 5b
surplus of $30. Note that there could be losses of up to $74 (the buyer
sends $80 and receives the lowest-quality Frank Thomas card); as in the
other induced value laboratory treatments herein—Treatments Lab and
Lab-Context—after these treatments were carried out, I had subjects
participate in other unrelated experiments that did not involve interaction
to ensure that they would leave with positive cash balances.
Treatment Lab-RF1 (RF1 denotes replication with field values in a
purely one-shot setting) is identical to Treatment Lab-RF in every manner
except that it is not executed over five periods with five different
partners; rather it is a one-shot game. Since, in the above treatments,
by design subjects should have construed the setting as one-shot, Treatment
Lab-RF and Treatment Lab-RF1 should yield similar data patterns
if (i) subjects interpret Treatment Lab-RF as several one-shot games and
(ii) experience does not unduly influence play. In total, Treatment Lab
yields 77 data points for buyers and 77 data points for sellers in the gift
exchange game.
In row 2 of table 1, Treatment Lab-Context adds context to Treatment
Lab-RF1. In this case, rather than buyers and sellers transacting with
abstract commodities, Treatment Lab-Context adds context that resembles
the subjects’ naturally occurring environment. For example, buyers
make an offer to a seller to buy one 1990 Leaf Frank Thomas card, and
the buyer requests a certain PSA grade. As in Treatment Lab-RF1, sellers
have five PSA grades available (PSA 6, 7, 8, 9, and 10) and subsequently
choose the quality of the Thomas baseball card to give the buyer if they
accept the buyer’s offer.9
Treatment Lab-Context includes 32 buyers and
32 sellers.
Completing the laboratory treatments is Treatment Lab-Market, more
specifically, Treatment Lab-Market($20) and Treatment Lab-Market
($65). Treatment Lab-Market is the laboratory market parallel to Treatment
Lab-Context: buying agents approach dealers in the experimental
market to purchase 1990 Leaf Thomas baseball cards in face-to-face
transactions. Each participant’s experience in Treatment Lab-Market
8
Please see app. C in List (2005) for a discussion of how these values were obtained.
9
PSA grades 6–10 were chosen because little trading of Thomas cards below PSA 6 is
carried out in the actual market. Note that in this treatment I am not actually having
agents transact with real commodities; rather subjects are told to act as though they are
using graded Thomas cards.
10 journal of political economy
followed four steps: (1) consideration of the invitation to participate in
an experiment, (2) a session to learn the market rules, (3) actual market
participation, and (4) conclusion of the experiment and exit interview.
In step 1, potential subjects approached the monitor’s dealer table
on the floor of the sports card show and inquired about purchasing
late 1980s/early 1990s baseball cards displayed on the table. If the subject
was a white male roughly 25 years of age, the monitor asked if he
was interested in participating in an experiment that would last about
30 minutes.10
If the agent agreed to participate, the administrator explained
that at a prespecified time the subject should enter an adjacent
room to take part in the experiment. Directions to the room were provided,
and the subject was informed that he would receive $20 to participate
in the experiment. To gather the dealer subject pool, I visited
numerous dealers’ tables and examined whether the dealer had a fair
number (more than five) of Thomas ungraded 1990 Leaf cards for sale
that were of sufficiently heterogeneous quality. If the dealer had a sufficient
number, he was asked if he would like to participate in a market
experiment in which he could potentially sell some of the Thomas cards.
Directions to the room and the appropriate times to enter the room
were then provided to those dealers who agreed to participate. No showup
fee was given to dealers.
Upon subjects’ arrival to the experimental market, in step 2 a monitor
thoroughly explained the market rules to them privately (buyers in one
room and sellers in another). Consumers were informed that they would
be “buyers” of 1990 Leaf Thomas baseball cards in the experiment. The
agents were told that they (typically in groups of five) would enter the
market and approach a prespecified dealer, who had his Thomas cards
displayed on his table in the experimental market. Importantly, in the
spirit of the literature that suggests that contracted negotiations can
crowd out reciprocity (see, e.g., Fehr and List 2004), I was careful to
instruct buying agents to avoid haggling, while keeping the transaction
as natural as possible.11
In practice, negotiations are typically quite short
or do not occur at all in this market (see List 2004a, table 2); thus,
besides realism, this approach gives social preferences their best shot,
10
Given the results in List (2004a), I wished to avoid any confounds associated with
statistical discrimination in this marketplace; hence I opted to use “majority” subjects as
my buying agents in all treatments. This design choice may well give social preferences
their best chance since the data in my earlier paper suggest that these types of buying
agents receive the best offers from dealers. Note, however, that any agent who desired to
participate in an experiment was able to do so since the minority agents were asked to
participate in an unrelated pilot experiment.
11
Macaulay (1963, 56) reports that “detailed negotiated contracts can get in the way of
creating good exchange relationships between business units.” Sitkin and Roth (1993,
376) assert that “legalistic remedies can erode the interpersonal foundations of a relationship
they are intended to bolster because they replace reliance on an individual’s
good will with objective, formal requirements.”
social preferences 11
since buying agents are signaling a fair amount of trust in the dealer
when purchasing nongraded sports cards without much detailed negotiations.
To ensure that buying agents did not aggressively bargain,
their payoffs were not tied to quality or price; rather, they were paid
$20 for approaching two dealers. And, to maintain consistency with
Treatment Lab-Context and afford the dealers reasonable price offers,
the buying agent offered $20 (or $65) and requested a 1990 Leaf Thomas
card that would merit a PSA 9 (10) if graded.
These parameter values were guided by the empirical results in Treatment
Lab-Context (discussed below), current market values of sports
cards, and what would be naturally demanded in this environment. First,
since the average buying agent sent $20 to dealers in Treatment LabContext
and requested a PSA 9 Thomas card, Treatment Lab-Market
($20) is the naturally occurring analogue. Treatment Lab-Market($65)
used the same dealers who were visited in Treatment Lab-Market($20)
and was identical in every sense except that in this case buying agents
offered $65 for the Thomas card and requested a PSA 10. I chose $65
because it is roughly 33 percent greater than , matchingc(10) p $50
the relationship of and the $20 value chosen in Treatmentc(9) p $15
Lab-Market($20). Second, use of lower-quality card levels would have
been unnatural since the bulk of demanded volume in the market is
for higher-end card types, such as PSA 9 and 10.
In step 3, the buying subjects each approached one dealer in round
1. Each interaction lasted less than three minutes and resulted in the
purchase of a Thomas Leaf sports card. Upon completing the transaction,
the buyer departed the experimental market and physically gave
the monitor the Thomas card in an adjacent room. After all transactions
in round 1 were completed, buyers received instruction on which dealer
to approach in round 2. Dealers were not allowed to communicate
during this time period. The buying agents then reentered the experimental
market and approached a different dealer for the final buying
period. Every dealer was approached twice: once with an offer of $20
and a request for a PSA 9 card, and once with an offer of $65 and a
request for a PSA 10 card. The ordering of the offers was random.
Step 4 concluded the experiment: after subjects completed a confidential
survey (see List 2005, app. B), they departed. In total, I observed
the behavior of 30 dealers who were each approached by two different
buying agents offering either $20 or $65; thus I have a sample size of
60 in Treatment Lab-Market.
Following the received gift exchange literature, if social preferences
play a role in this case, then the card’s grade and the offer price should
be positively correlated: in a reduced-form regression model.Ѩq/Ѩp 1 0
Once the buying agents had purchased each of the cards in these treat-
12 journal of political economy
ments, the last step was to have the cards professionally graded. This
was completed by having every card graded by a PSA representative.
Treatment Floor moves the exploration out of the laboratory and into
the marketplace in which these agents actually consummate business:
the floor of the sports card show.12
Treatments Floor-$20 and Floor-$65
represent the naturally occurring analogues to Treatment Lab-Market
and are identical whenever possible. Again, the buying agent’s experience
typically followed four steps. In step 1, white males roughly 25
years old who were interested in late 1980s/early 1990s baseball cards
were asked to participate in an experiment. If the agent agreed to participate,
in step 2 a monitor thoroughly explained the experimental
rules. The agent was informed that he would be a “buyer” of 1990 Leaf
Thomas baseball cards in the experiment.
The agent was told that he would approach 10 different dealers on
the floor of a sports card show to purchase the Thomas card. I was able
to preselect the dealers to be approached before the show by visiting
their dealer tables and examining whether they had more than five
Thomas ungraded 1990 Leaf cards for sale that were of sufficiently
heterogeneous quality. It is common practice for dealers to mill around
the show looking at others’ goods, and I was merely behaving in accordance
with this norm when visiting dealer tables. As in Treatment
Lab-Market, I was careful to instruct buying agents to avoid haggling,
while keeping the transactions as natural as possible. And the buying
agent offered $20 (or $65) and requested a 1990 Leaf Thomas card
that would merit a PSA 9 (10) if graded. In this spirit, much as in
Treatment Lab-Market, buying agents represented sophisticated buyers
in that they expressed interest in a PSA-graded card.
In step 3, the subject approached dealers one at a time. As in Treatment
Lab-Market, each interaction lasted less than three minutes and
resulted in the purchase of a Leaf Thomas sports card. It should be
noted that throughout the experiment the sports card dealers were not
aware that an experiment was occurring. This ensured that the process
was as natural as possible for the dealers, whose behavior was of primary
interest in this field experiment. Step 4 concluded the experiment: after
subjects completed a confidential survey, they were paid $20 in private.
A few noteworthy design issues should be mentioned before I proceed.
12
As I have noted elsewhere (e.g., List 2004b, 2004c), with the rise in popularity of
collector sports cards and memorabilia over the past two decades, markets that organize
buyers and sellers have naturally arisen. Temporal assignment of the physical marketplace
is typically done by a professional association or local sports card dealer, who rents a large
space, such as a gymnasium or hotel conference center, and allocates 6-foot tables to
dealers for a nominal fee. When the market opens, consumers mill around the marketplace,
haggling and bargaining with dealers, who have their merchandise prominently
displayed on their tables. The duration of a typical sports card show is a weekend, and
subjects enter the market ready to buy, sell, and trade.
social preferences 13
First, each dealer was approached twice: once in Treatment Floor-$20
and once in Treatment Floor-$65. Visits were spaced so as to attenuate
any suspicion: one example is that dealer i was approached by agent n
on Friday night and by agent m on Sunday morning. And, the ordering
of the visits was random: some dealers were approached in the $20
treatment first, others were approached in the $20 treatment second.
I observed no ordering effect, so I suppress further discussion of this
issue.
Second, in contrast to audit studies that test for market discrimination,
I am directing the agent to buy the good. In this sense, these are not
transactors who obliquely discontinue bargaining if the dealer accepts
an offer; these are actual transactions. And, since transactions are typically
in cash at sports card shows, I provided the necessary funds to
purchase the cards. Third, note that great care was taken to ensure that
the data were gathered from interactions that would naturally occur in
the marketplace. Subjects were entering the market to buy goods that
were very similar to the good that I had them buying. Fourth, Treatments
Lab-Market and Floor were carried out at several different sports card
shows in the same region in the United States, from October 2002 to
July 2004.
In total, I observed the behavior of 50 dealers who were each visited
by two different agents (one in Treatment Floor-$20 and one in Treatment
Floor-$65); thus I have a sample size of 100 in Treatment Floor.
As in Treatment Lab-Market, the last step of the experiment was to have
the cards professionally graded. In addition, I should note that in every
case I was able to obtain important subject-specific information from
the dealers, either via a survey they completed during an experiment
in which they later participated or through a survey (see List 2005, app.
B) they filled out in exchange for a payment of $1.
To explore a level deeper into the underlying structure that organizes
behavior in this market and control for potential selection effects, I
completed three final treatments making use of natural exogeneity that
the market offered during the sample period: while a third party (PSA)
has graded sports cards since 1987, no service existed prior to June 2003
to grade sporting event tickets and ticket stubs. PSA announced its
grading intentions in April 2003, but it did not provide grading criteria
until June 2003. As noted earlier, Brown et al. (2004) highlight the
attractiveness of such natural variation by arguing that such exogeneity
is impossible to find in field data. I believe that these three field experimental
treatments offer this useful characteristic.
Treatment Floor-NoGrading (denotes no grading available) is identical
to Treatment Floor in that buyers approached dealers on the floor
of a sports card show (from October 2002 to March 2003) with a low
and a high price (either $10 or $30) to purchase an unused ticket or
14 journal of political economy
ticket stub that would receive a PSA grade of 9 or 10 if tickets were
graded like sports cards. Given the thinness of the ticket market, it was
necessary to use five different ticket types in the purchasing tasks (Cal
Ripken’s last game at Camden Yards, his final game of “the Streak,” his
“consecutive world-record-breaking” game, and two World Series
games). I was careful to choose tickets that were in the same price range
to increase the likelihood of having the luxury of pooling the data. In
total, I observed the behavior of 30 dealers in this treatment and therefore
gathered 60 data points since each dealer was approached twice.
Treatment Floor-AnnounceGrading (denotes after announcement of
grading) was completed at sports card shows after PSA announced that
it would begin grading ticket stubs (April 2003) but before it released
its grading scheme (June 2003). In this treatment, I purchased the same
tickets and used the same protocol as in Treatment Floor-NoGrading.
As outlined in row 4 of column 2 of table 1, I observed 54 dealer
decisions in this treatment.
Completing the experimental design is Treatment Floor-Grading (denotes
grading available), which is identical to Treatments Floor-NoGrading
and Floor-AnnounceGrading but was completed after June
2003, a time period in which grading services of tickets and ticket stubs
existed. I observed 36 total dealer decisions in this final treatment.
Accordingly, I purchased 150 tickets in these three treatments; and as
in Treatment Floor, I subsequently had every ticket graded by a PSA
representative. In the Appendix, I provide the necessary institutional
details about the sports card and sports ticket grading industry to motivate
the experimental design.
Identification and Hypotheses
The sports card marketplace includes both local and nonlocal sellers
(dealers). Accordingly, by parsing dealers into types, my experimental
design permits two distinct identification strategies. First, with the maintained
assumption that nonlocal dealers have no reputational concerns,
any reciprocal behavior observed among the subjects in the floor treatments
can be attributable to social preferences. Alternatively, behavior
of local dealers might include reputational as well as social preference
effects. Second, even if I observe disparate data patterns across local
and nonlocal dealers, it might be the case that selection effects are at
work. Indeed, one of the key points of the theoretical work on social
preferences is that there is a distribution of types in the population—
some purely selfish types and others with social preferences. In this spirit,
it is important to consider a second means of identification: behavior
within dealer type—across the laboratory and field settings, as well as
within the field treatments.
social preferences 15
Under this design, a clean test of the predictions from a purely selfinterested
model and a model with social preferences is permitted. Table
2 provides a summary of what can be examined across the various experimental
treatments. In a general sense, table 2 highlights that laboratory
experiments that estimate propensities might not provide the
necessary environment to measure them accurately. Consider column
1 of table 2, which summarizes predictions in the lab treatments. In this
case, I delineate between two potential situations: one that has “experimenter”
effects and one that does not have experimenter effects. For
example, in the lab treatments, reputational concerns (i.e., sellers are
fully aware that the experimenter can document delivered quality), experimenter
demand effects, Hawthorne effects, or simply the fact that
the task is undertaken in an artificial setting can each potentially confound
any preferred interpretation.13
Though the definition is much
too narrow, for minimalism I denote such potential laboratory effects
simply as “reputational” concerns in table 2.
These effects are important to consider when delineating between
predictions of the self-interest and social preference models. If such
effects are present in the lab, then both models predict that there will
be gift exchange, or a positive relationship between price and quality.
Of course, in the field treatments (predictions contained in cols. 2 and
3 of table 2), by design, such effects are varied, permitting a clean test
of the two models. As I progress through the summary of the results
below, I shall highlight whether evidence is consistent with each of the
theories.
III. Experimental Results
Table 3 provides a summary of the raw data. The table can be read as
follows: Treatment Lab-R in row 1 denotes that the average price in this
treatment was $28.40, average quality was 3.5, and average requested
quality was 6.1. Note that in table 3, for comparability reasons, I have
scaled Treatment Lab-R data to range from 1 to 10, and PSA 6, 7, 8, 9,
and 10 are denoted as quality levels 1, 2, 3, 4, and 5.14
A first result
13
In social psychology, several studies due to Martin Orne, Robert Rosenthal, and others
discuss the important effects of the experimenter-subject relationship (see, e.g., Orne
[1962] and Rosenthal’s [2002] summary). While efforts to expunge such effects have been
explored in the experimental literature using double-blind (e.g., Hoffman, McCabe, and
Smith 1996), randomized response (e.g., List et al. 2004), and related techniques, such
approaches may attenuate a number of laboratory phenomena but seem incapable of
completely eliminating them, and might even introduce other biases.
14
Average individual payoffs (ranges) are as follows: Treatment Lab-R: buyers, $14.90
($6.50 to $24), and sellers, $18.60 ($5 to $34); Lab-RF: buyers, $2.40 (Ϫ$59 to $25), and
sellers, $8.00 ($1 to $61); Lab-RF1: buyers, $0.22 (Ϫ$25 to $25), and sellers, $9.81 ($1 to
$35); Lab-Context: buyers, Ϫ$0.09 (Ϫ$67 to $25), and sellers, $8.44 ($1 to $70).
16
TABLE 2
Predictions: Self-Interested Model vs. Social Preference Model
Floor Treatments
Tickets
Lab Treatments Sports Cards Floor-NoGrading
Floor-Announce
Grading and
Floor-Grading
Self-Interested Model
Without experimenter effects: no
relationship between price and
quality
Local dealers: positive relationship
between price and quality
due to, e.g., reputation effects
Local dealers: no relationship between
price and quality
Local dealers: positive relationship
between price and quality
due to, e.g., reputation effects
With experimenter effects: positive
relationship between price
and quality due to, e.g., reputation
effects
Nonlocal dealers: no relationship
between price and quality
Nonlocal dealers: no relationship
between price and quality
Nonlocal dealers: no relationship
between price and quality
Social Preference Model
Without experimenter effects:
positive relationship between
price and quality due to social
preferences
Local dealers: positive relationship
between price and quality
due to, e.g., reputation effects
and social preferences
Local dealers: positive relationship
between price and quality
due to social preferences
Local dealers: positive relationship
between price and quality
due to, e.g., reputation effects
and social preferences
With experimenter effects: positive
relationship between price
and quality due to, e.g., reputation
effects and social
preferences
Nonlocal dealers: positive relationship
between price and
quality due to social
preferences
Nonlocal dealers: positive relationship
between price and
quality due to social
preferences
Nonlocal dealers: positive relationship
between price and
quality due to social
preferences
Note.—Each column represents predictions of the self-interested model vs. the social preference model across the three major experimental types. In a split of the dealer types, a dealer is
labeled as a nonlocal if he or she is unlikely to be concerned with reputation effects; e.g., if he or she rarely attends sports card shows in the area (fewer than three times in a typical year), does
not plan to attend more frequently than this in the future, does not own a sports card shop, and does not have an Internet sports card business. All other dealers are labeled as locals.
17
TABLE 3
Results Summary
p
(1)
q
(2)
qr
(3)
Treatment Lab
Treatment Lab-R 28.4
(16.1)
3.5
(2.0)
6.1
(2.1)
Treatment Lab-RF 22.6
(20.7)
2.3
(1.4)
4.1
(.9)
Treatment Lab-RF1 24.8
(22.1)
2.5
(1.7)
4.0
(1.3)
Treatment Lab-Context
Treatment Lab-Context 19.5
(19.6)
2.3
(1.5)
4.2
(1.1)
Treatment Lab-Market($20) $20 3.1
(.9)
4
Treatment Lab-Market($65) $65 4.1
(.6)
5
Treatment Floor (Cards)
Treatment Floor-$20 $20 2.1
(.9)
4
Treatment Floor-$65 $65 3.2
(1.0)
5
Treatment Floor (Tickets):
Treatment I
Treatment Floor-NoGrading $10 2.7
(.6)
4
Treatment Floor-AnnounceGrading $10 2.9
(.6)
4
Treatment Floor-Grading $10 3.1
(.8)
4
Treatment Floor (Tickets):
Treatment II
Treatment Floor-NoGrading $30 2.7
(.7)
5
Treatment Floor-AnnounceGrading $30 3.4
(.8)
5
Treatment Floor-Grading $30 3.6
(1.1)
5
Note.—Summary statistics from one (or two in the case of Treatment Floor [Tickets]) unique treatment. p is
the average price, q is the average quality, and qr is the average requested quality. Treatment I-R data are scaled
to range from 1 to 10, and PSA 6, 7, 8, 9, and 10 are denoted as quality levels 1, 2, 3, 4, and 5 in the table.
Standard deviations are in parentheses.
18 journal of political economy
relates to the comparison between the behavior of this subject pool and
that of students. As Fehr and List (2004) note, a typical criticism levied
against experimental results concerns the fact that most economics experiments
are conducted with students. This may be problematic for
several reasons. For example, owing to selection effects, those who do
not behave like students may have selected into roles and be overrepresented
in certain parts of the economy (e.g., sellers in the marketplace).
The first result addresses this issue.
Result 1. Behavior of sports card enthusiasts in laboratory games
is in line with the gift exchange literature using student subjects, and
the results extend well to one-shot environments.
Evidence for result 1 is contained in the raw statistics in row 1 of table
3, which are consistent with the raw data gathered in laboratory experiments
with student subjects (see, e.g., Fehr et al. 1993; Charness 1996).
Overall, a graphical depiction of the trajectory of the data clearly shows
that product quality and prices are positively related (raw data figures
that complement table 3 appear in the working paper of this study [List
2005]). In addition, when I examine the temporal aspect of the data,
there is little variation over time, consistent with previous studies on
gift exchange (for an exception, see Charness, Frechette, and Kagel
[2004]).
To provide the necessary statistical link to the literature, I follow
previous work and estimate Tobit random-effects regression models. The
dependent variable in the regressions is the quality of the good, which
is regressed on the price transfer and controls for dealer-specific effects:
q p bp ϩ q . (3)it it it
In equation (3), represents the product quality that dealer i sent toqit
the buyer in period t, denotes the buying agent’s offer price to dealerpit
i in period t, and qit includes a white-noise error term with mean zero
and a constant in the Tobit model. This specification is augmented by
inclusion of dealer random effects in the Tobit random-effects regression
model.15
Regression results presented in columns 1–3 of table 4 provide evidence
that dealers reward buyers for paying higher prices. In each of
the three treatments the marginal price effect is positive and statistically
significant at the level with a two-sided alternative. This result isp ! .10
consistent with the received gift exchange literature. When applicable,
I also present an estimate of v in table 4, where v is equal to Ѩv(q)/ѨP
15
Including a time trend does not change the qualitative results. And when the models
converged, controlling for buyer-specific effects does not change the qualitative empirical
results presented below. In addition, I have explored empirically modeling the relationship
between buyer and seller rents: in a regression framework. Insights similarv Ϫ p p f(p Ϫ c)
to those presented below are obtained, so I suppress further discussion.
TABLE 4
Marginal Effects Estimates for the Sellers’ Provided Quality
Variable
Treatment Type
Lab-R
(1)
Lab-RF
(2)
Lab-RF1
(3)
Lab-
Context
(4)
Lab-
Market
(5)
Floor
(Cards)
(6)
Floor-
NoGrading
(7)
Floor-
Announce
Grading
(8)
Floor-
Grading
(9)
Floor-
Pooled
(10)
Price .05
(1.8)
.05
(3.3)
.07
(4.3)
.05
(4.3)
.02
(4.4)
.02
(6.6)
Ϫ.001
(.01)
.02
(2.1)
.02
(1.1)
.02
(2.6)
Constant .6
(.7)
Ϫ.4
(.7)
Ϫ.9
(3.3)
Ϫ.8
(2.9)
1.6
(6.2)
.6
(3.1)
1.7
(8.0)
1.7
(5.8)
1.8
(3.3)
1.7
(7.3)
v . . . $.72
(3.6)
$1.1
(6.9)
$.65
(4.7)
$.45
(2.1)
$.21
(5.0)
$.01
(.3)
$.17
(1.1)
$.23
(1.1)
$.19
(2.3)
Dealer random effects yes yes no no yes yes yes yes yes yes
Observations 25 25 27 32 60 100 60 54 36 90
Note.—The dependent variable is the sellers’ product quality given to the buyer. Floor-Pooled pools Floor-AnnounceGrading and Floor-Grading data. v is the estimate of the monetary gift
exchange, computed as . t-ratios (in absolute value) are beneath marginal effect estimates.Ѩv(q)/ѨP
20 journal of political economy
and provides a natural benchmark of gift exchange expressed in monetary
units. In the case of Treatments Lab-RF and Lab-RF1, both estimates
of v are significantly different from zero, suggesting that gift
exchange occurs at the margin. In terms of economic significance, a v
estimate of 1.1 in Treatment Lab-RF1 suggests that a $1 increase in p
leads to a $1.10 increase in the reciprocated gift, .v(q)
While these results provide a robustness check of the data gathered
in the laboratory with student subjects and represent good news in that
the major laboratory results seem to spill over to different pools of
subjects who are commonly engaged in similar exercises in their everyday
lives, one can push the comparability notion a bit harder by adding
field context to the laboratory environment. This approach is inherent
in Treatment Lab-Context, which yields the following result.
Result 2. Adding natural context influences behavior, but gift exchange
remains alive and well.
Evidence for this result can be found in the summary of the Treatment
Lab-Context data contained in table 3. Treatment Lab-Context data
reveal that average prices and quality levels are only slightly lower than
what was observed in Treatment Lab-RF1 (the comparable context-free
treatment). Slight behavioral differences are also revealed in comparisons
of scatter plots of these data (List 2005), which show (i) that the
positive relationship remains in the contextual data but that there is a
slightly greater mass at the subgame perfect equilibrium prediction: 13
of 32 (41 percent) observations in Treatment Lab-Context versus nine
of 27 (33 percent) observations in Treatment Lab-RF1; and (ii) that
there is a greater number of price (quality) realizations at $25 (three)
and below in Treatment Lab-Context.
For the data from Treatment Lab-Market, table 3 shows that the positive
relationship between price and product quality is evident in the
aggregate data: whereas the average quality was 3.1 (PSA 8.1) in Treatment
Lab-Market($20), it was 4.1 (PSA 9.1) in Treatment Lab-Market
($65). Figure 1 provides a visual view of these quality differences. Comparing
the proportion of sellers who provided various quality levels
across the $20 and $65 treatment yields a discernible rightward shift in
the distribution of Treatment Lab-Market($65) data.16
In terms of the
average monetary value of the return gift ( ), sellers provided $19.73v(q)
in Treatment Lab-Market($20) and $41.33 in Treatment Lab-Market
($65).
To compare gift exchange on the margin across these two treatments,
I return to equation (3) and estimate a Tobit model. For Treatment
16
In some instances dealers made quality claims, and these included statements that
they could not provide the requested quality. As in Treatments Floor (Cards) and Floor
(Tickets), I still had my buyers purchase the good in Treatment Lab-Market and provide
this information to me in the survey. I consider mendacious claims below.
social preferences 21
Fig. 1.—Treatment Lab-Market
Lab-Context data, the marginal price effect is positive and statistically
significant at conventional levels (see col. 4 of table 4). It is interesting
to note that the marginal effect estimate (0.05) is slightly lower than
the marginal effect estimate in Treatment Lab-RF1 (0.07), and v is considerably
lower: $0.65 versus $1.10. When Treatment Lab-RF1 and Treatment
Lab-Context data are pooled and equation (3) is estimated, however,
a likelihood ratio test suggests that the homogeneity null should
not be rejected, suggesting that behavioral differences do not exist across
Treatments Lab-RF1 and Lab-Context.
Considering Treatment Lab-Market data, I provide marginal effects
estimates from a Tobit random effects model in column 5 of table 4.
The marginal effect estimate of 0.02 is positive and significant at conventional
levels; this estimate suggests that card quality increases by
roughly one grade when the buyer offers $65 rather than $20; in this
case, .17
Accordingly, the overall pattern of results suggests thatv p $0.45
gift exchange is alive and well, even when market context is utilized in
the experimental design.
Results 1 and 2 provide a nice validity check of the extant gift exchange
literature. Yet, as table 2 highlights, this evidence does not unequivocally
show that these subjects exhibit social preferences. It might
be the case that these dealers are purely selfish and, owing to effects
associated with the lab environment, they behave in a manner consistent
with gift exchange. A necessary next step is to explore behavior in
17
In addition to the Tobit random-effects estimation strategy, which is heavily utilized
in the literature, since there is a natural ordering in the data and there are only five cells
(i.e., PSA 6–10), I supplement these results by using a panel data ordered probit model,
as described in app. D in List (2005). Empirical estimates from the panel data ordered
probit model are suppressed because they always coincide with insights gained from eq.
(3).
22 journal of political economy
naturally occurring environments in which the controls of the experiment
are relaxed appropriately. In such a setting, experimenter demand
effects, Hawthorne effects, and the like are absent since, unbeknownst
to them, experimental subjects (sellers) are randomly chosen from the
dealers who have certain goods. A first insight from the field treatments
is as follows.
Result 3. When third-party verification is available, behavior in
naturally occurring transactions is consonant with both gift exchange
and a concern for reputation.
Tables 3 and 4 as well as figure 2 provide evidence for result 3. Row
3 in table 3 shows that the positive relationship between price and
product quality is evident in the aggregate data: whereas the average
quality was 2.1 (PSA 7.1) in Treatment Floor-$20, it was 3.2 (PSA 8.2)
in Treatment Floor-$65. In terms of the average monetary value of the
return gift ( ), however, sellers provided much less than they providedv(q)
in Treatment Lab-Market: roughly $8 in Treatment Floor-$20 and $20
in Treatment Floor-$65 (vs. $19.73 and $41.33 in Treatment Lab-Market).
This difference is highlighted via a comparison of figures 1 and
2, which reveals the significant leftward shift in the quality distributions
across both the $20 and $65 price offers when one moves from the lab
to the field. Concerning the ticket stub data, I find that Treatments
Floor-AnnounceGrading and Floor-Grading displayed in row 4 of table
3 support the positive relationship found in the sports card data.
Regression results in table 4 yield similar insights: estimates in column
6 of table 4 provide evidence that product quality and price are positively
correlated in Treatment Floor (Cards), since the marginal effect estimate
of 0.02 is positive and significant at conventional levels. This estimate,
which is quite similar to the marginal effect in Treatment LabMarket,
suggests that card quality increases by roughly one grade when
the buyer offers $65 rather than $20. In this case, however, since the
quality changes from PSA 7 to PSA 8 (rather than PSA 8 to PSA 9 in
Treatment Lab-Market), , considerably lower than the v esti-v p $0.21
mate of $0.45 in Treatment Lab-Market. A similar result is found in the
Treatment Floor-AnnounceGrading and Treatment Floor-Grading data
presented in columns 8 and 9 of table 4, although the marginal price
effect is not statistically significant in the Treatment Floor-Grading data
at conventional levels. When the Treatment Floor-AnnounceGrading
and Treatment Floor-Grading data are pooled (a likelihood ratio test
indicates that pooling is appropriate: ), however, the marginal2
x p 5.8
price effect, contained in column 10 of table 4, is statistically significant.
Interestingly, across all three specifications the marginal price effect
estimate is 0.02, and v is approximately $0.20.18
18
In computations of v in the ticket specifications, is equivalent to one-half thev(q)
value of in the sports card data.v(q)
Fig. 2.—Treatment Floor: a, sports cards; b, local dealers; c, nonlocal dealers
24 journal of political economy
As table 2 shows, since this data pattern is observationally equivalent
to predictions from a model based purely on reputational effects (e.g.,
Klein and Leffler 1981), again, these insights are not unequivocal evidence
in favor of gift exchange. One can explore a level deeper by
recognizing that some of the dealers in the sample may have had an
economic reason to uphold their reputations, whereas others may not
have had similar incentives. A next result follows.
Result 4. When third-party verification is possible, local dealer
behavior in naturally occurring transactions is consonant with both gift
exchange and a concern for reputation, whereas nonlocal dealers’ behavior
is in line with self-interest theory.
Table 5 and figures 2, 3, and 4 provide evidence for this result. In a
split of the dealer types, a dealer is labeled as a “nonlocal” if he or she
is unlikely to be concerned with reputation effects, for example, if he
or she rarely attends sports card shows in the area (fewer than three
times in a typical year), does not plan to attend more frequently than
this in the future, does not own a sports card shop, and does not have
an Internet sports card business. All other dealers are labeled as “locals”;
in practice, these are primarily dealers who frequent the area often.
This information was obtained from a survey (see List 2005, app. B).
Note that besides this difference, across all other observables, such as
years of experience and age, dealers are similar. I return to the issue
of selection effects below, however.19
The raw data displayed in figures 2, 3, and 4 provide initial support
for result 4. In transactions with local dealers, higher price offers yield
superior quality in Treatments Floor, Floor-AnnounceGrading, and
Floor-Grading, as illustrated in figures 2b and 3. Alternatively, while
delivered quality is positively related to price across these three treatments
among nonlocal dealers (see figs. 2c and 4), the differences are
minute.
Table 5 provides regression results to support result 4. Columns 1 and
2 split the Treatment Floor data into two subsamples: Floor(Cards)L
(local dealers over sports cards) and Floor(Cards)N (nonlocal dealers
over sports cards). In the former subsample, the marginal price effect
is positive and significant at conventional levels. In terms of economic
significance, the coefficient estimate in column 1 of 0.03 results in an
estimated marginal effect of roughly 1.5 grades; that is, in the $65 treatment,
local dealers provided a quality that was 1.5 grades above the
quality level they provided in the $20 treatment. Measured at the sample
means, this 1.5 quality increment yields the buyer a PSA-rated 8.6 card
19
When I categorize dealer data from Treatment Lab and Treatment Lab-Market in a
similar manner, I find that there is no difference in behavior across local and nonlocal
dealers.
TABLE 5
Marginal Effects Estimates for the Sellers’ Provided Quality Split by Dealer Type
Variable
Treatment Type
Floor
(Cards)L
(1)
Floor
(Cards)N
(2)
Floor-No
GradingL
(3)
Floor-No
GradingN
(4)
Floor-
Announce
GradingL
(5)
Floor-
Announce
GradingN
(6)
Floor-
GradeL
(7)
Floor-
PoolN
(8)
Floor-
PoolL
(9)
Price .03
(8.6)
.004
(.7)
.002
(.2)
Ϫ.005
(.5)
.04
(2.1)
.003
(.3)
.04
(2.7)
.003
(.1)
.04
(4.8)
Constant .6
(4.1)
.6
(4.6)
1.6
(5.0)
1.8
(5.2)
1.7
(5.2)
1.5
(4.6)
1.8
(5.0)
1.8
(1.7)
1.8
(10.0)
v $.31
(5.2)
$.01
(.5)
$.02
(.4)
Ϫ$.006
(.5)
$.32
(1.4)
$.02
(.6)
$.42
(1.5)
$.03
(.1)
$.35
(2.1)
Dealer random effects yes yes yes yes yes yes yes yes yes
Observations 70 30 36 24 30 24 20 16 50
Note.—The dependent variable is the sellers’ product quality given to the buyer. Floor-PoolL pools Floor-AnnounceGradingL and Floor-GradingL data. v is computed as . t-ratios (inѨv(q)/ѨP
absolute value) are beneath marginal effect estimates. Subscripts N and L after treatment type denote regressions with local and nonlocal dealer data only.
26 journal of political economy
Fig. 3.—Price/quality relationship for local dealers
Fig. 4.—Price/quality relationship for nonlocal dealers
rather than a PSA-rated 7.1 card. With the values discussed earlier,v(q)
this quality increase maps into an increase in market value of roughly
$20, much less than the extra $45 spent to obtain the card. A v estimate
of $0.31 complements this finding.
Alternatively, for nonlocal dealers, gift exchange is not evident in
Treatment Floor (see col. 2 of table 5), since the marginal price effect
is not statistically significant at conventional levels. Regression results
for Treatments Floor-AnnounceGrading and Floor-Grading provide fur-
social preferences 27
ther support for result 4: in both cases the marginal price effect in the
local dealer data is positive and significant at conventional levels (cols.
5 and 7 of table 5), whereas there is no such effect found in the nonlocal
dealer data (cols. 6 and 8 of table 5). For both the Treatment FloorAnnounceGrading
and Floor-Grading local dealer data, the marginal
effect estimate is 0.04, and v p $0.32 and $0.42, though neither v
estimate is statistically significant at conventional levels. When these data
are pooled (likelihood ratio test: ), v p $0.35 and is significant2
x p 1.4
at the level (col. 9 of table 5). Treating nonlocal dealer datap ! .05
similarly by pooling data from Treatments Floor-AnnounceGradingN
and Floor-GradingN provides little new information: gift exchange is not
evident among nonlocal dealers. Accordingly, as table 2 suggests, the
nonlocal dealer data are consonant with the self-interest model.
A natural question that arises concerns whether the local dealer behavior
is driven primarily by reputation effects or social preferences:
given the identification problem, from the above results alone one cannot
determine the extent to which reputation effects and social preferences
are influencing market outcomes. One nice characteristic of
the current experimental design is that I can examine behavior in markets
that are void of third-party verification to explore this issue. In such
cases, in economic terms the situation faced by the local and nonlocal
dealers is identical. Treatment Floor-NoGrading provides a first result.
Result 5. When third-party verification is not available, supply-side
behavior in naturally occurring transactions is consonant with purely
selfish money-maximizing theory, suggesting that reputational considerations,
rather than social preferences, are driving the earlier results.
Evidence for this result can be seen in tables 3–5 as well as figures 3
and 4. Table 3 shows that there is very little quality difference between
the $10 and $30 offers in Treatment Floor-NoGrading. This result is
highlighted in figures 3 and 4, where both local and nonlocal dealers
do not provide different quality levels across offers of $10 and $30 in
Treatment Floor-NoGrading. Empirical results displayed in tables 4 and
5 support the raw data patterns, since the marginal price effect is insignificant
in the aggregate data (col. 7 of table 4) and in both specifications
that split the data by dealer type (cols. 3 and 4 of table 5).
This finding, which according to the predictions outlined in table 2
is in line with the self-interest model, leads to the tentative conclusion
that reputation effects rather than social preferences are responsible
for driving a large part of the price/quality tendencies observed in the
naturally occurring data. While there is some evidence in favor of social
preferences in this market, as price and quality are directionally related
(positively) in various places in the nonlocal dealer data and in the local
dealer Treatment Floor-NoGrading data, it seems to be of second-order
28 journal of political economy
importance in real market transactions.20
This insight can be viewed in
List (2005, app. E), which provides several supplementary figures that
summarize the raw data across local and nonlocal dealers in the ticket
stub treatments.
Mendacious Claims
Empirical estimates presented above provide measures of gift exchange
in the spirit of the extant literature and highlight a framework that can
measure social preference effects and reputation effects. Yet it is important
to recognize and examine the degree of mendacious claims in
the marketplace. If dealers do not have the necessary inventory to fulfill
the quality request (e.g., as a result of my misjudgment of quality during
my perusal of sales during the show) but provide quality disclaimers,
then it is important to explore this aspect of behavior. In this spirit, an
important complement to the above results is a thorough analysis of
the statistical association between quality claimed and quality delivered.
A first result follows.
Result 6. When third-party verification is possible, local dealers
provide fewer claims of quality than nonlocal dealers in the field and,
conditional on claiming quality, shirk less frequently.
Table 6 summarizes dealer behavior across Treatments Floor (Cards)
and Floor (Ticket Stubs). Evidence for the first part of result 6 can be
obtained by computing the percentage of local and nonlocal dealers
who claim quality in Treatments Floor, Floor-AnnounceGrading, and
Floor-Grading. The second part of result 6 follows from a comparison
of the quality claimed and the quality actually delivered. Before I discuss
the evidence for result 6, it is important to point out that in some cases
dealers provide quality ranges; for example, “this card would grade at
PSA 8 or 9.” In these cases I use the midpoint of the range (e.g., 8.5).
A few other dealers were agnostic about the grading system. I label these
types as not claiming quality (similar results are obtained if I simply
delete these observations). And, in some instances the dealer stated
“this one is top quality” or “this is a gem” when describing the good. I
label these dealers as not claiming quality, but note that if I take the
literal word of the dealer and pair these statements with the appropriate
PSA grade, the fundamental results do not change.
Upon pooling data from the Treatments Floor, Floor-Announce
Grading, and Floor-Grading in table 6, I find that 94 of 190 (49 percent)
20
I also gathered information on length of buyer/seller relationships. While speculative,
an upper-bound estimate of social preferences within long-term relationships (where a
long-term relationship is defined as one wherein the buyer and dealer have had five or
more interactions in the previous 12 months or have had two or more interactions annually
over the past three plus years) suggests that they influence the price and qualitycorrelation.
social preferences 29
TABLE 6
Summary of Results: Product Quality Claims
Claims
(1)
Quality
Claim
(2)
Delivered
Quality
(3)
Delivered
Promised
Quality
or Above
(4)
Overall
Treatment Lab-Market 10/60 3.95
(.4)
4.3
(.5)
10/10
Treatment Floor (Cards) 53/100 3.9
(.7)
2.7
(1.1)
15/53
Treatment Floor-NoGrading 36/60 3.8
(.6)
2.8
(.6)
8/36
Treatment Floor-Announce
Grading
24/54 4.2
(.5)
2.9
(.9)
4/25
Treatment Floor-Grading 17/36 4.2
(.6)
3.1
(1.1)
4/17
Local Dealers
Treatment Lab-Market 6/42 4.0
(.5)
4.3
(.5)
6/6
Treatment Floor (Cards) 27/70 3.9
(.7)
3.4
(1.1)
12/27
Treatment Floor-NoGrading 22/36 3.9
(.5)
2.8
(.6)
4/22
Treatment Floor-Announce
Grading
7/30 4.1
(.3)
3.9
(.4)
4/7
Treatment Floor-Grading 4/20 4.3
(1.0)
3.8
(.5)
2/4
Nonlocal Dealers
Treatment Lab-Market 4/18 3.9
(.3)
4.3
(.5)
4/4
Treatment Floor (Cards) 26/30 4.0
(.7)
2.0
(.6)
3/26
Treatment Floor-NoGrading 14/24 3.7
(.6)
2.8
(.6)
4/14
Treatment Floor-Announce
Grading
17/24 4.3
(.6)
2.5
(.6)
0/18
Treatment Floor-Grading 13/16 4.2
(.4)
2.9
(1.2)
2/13
Note.—Standard deviations are in parentheses.
dealer observations involve product quality claims. In a split by dealer
type, 38 of 120 (32 percent) local dealer observations involve product
quality claims, whereas 56 of 70 (80 percent) nonlocal dealer observations
involve product quality claims. Of those dealers who make quality
claims, local dealers deliver the promised quality (or above) in 18 of
38 cases (47 percent), whereas nonlocals deliver the promised quality
(or above) in only five of 57 (9 percent) cases.
30 journal of political economy
To complement these insights, I estimate the bivariate probit model
with sample selection due to van de Ven and van Praag (1981):
1 if Y* 1 01
Y* p F(b V ) ϩ e ; Y p (4a)1 1 1 1 {0 otherwise
and
1 if Y* 1 02
Y* p Q(b Z) ϩ e ; Y p (4b)2 2 2 2 {0 otherwise;
e , e ∼ bivariate normal(0, 0, 1, 1, r).1 2
Equation (4a) is the quality claim equation. The variable is unob-Y*1
served, but I can observe its sign since if the dealer provided aY p 11
quality claim, and zero otherwise. Variables in V include a dichotomous
variable indicating whether the seller is a local dealer and a control for
the price offer/quality requested. Equation (4b) is the shirking equation
and is observed only when ; hence the selectivity model arisesY p 11
since a mendacious claim can occur only among dealers who provide
a quality promise. In Z, I include a dichotomous variable for whether
the seller is a local dealer. To account for data dependencies, I calculate
the standard errors assuming that the error terms are independent
across dealers but not within each dealer (i.e., clustered standard errors
for dealers).
The individual quality claim and shirking equations could be estimated
separately, but unless , such estimation results in sampler p 0
selection bias. In estimation of the system, I use full-information maximum
likelihood, where the log likelihood is given by
ln f (b V, b Z, r) ϩ ln f (b V, Ϫb Z, Ϫr)͸ ͸2 1 2 2 1 2
y ,y p1 y p1,y p01 2 1 2
ϩ ln f(Ϫb V ), (5)͸ 1
y p01
where f2 denotes the bivariate standard normal cumulative density function
and f denotes the univariate standard normal cumulative density
function.
Empirical results in support of result 6 are presented in columns 1,
2, and 5–8 of table 7. The first part of result 6—when third-party verification
is possible, in the field local dealers provide fewer quality
claims—can be found in all three quality claim equations.21
For instance,
empirical estimates in column 1 suggest that in the sports card treat-
21
The Floor-Grading model failed to converge; thus I present estimates from equations
that are estimated separately.
social preferences 31
TABLE 7
Empirical Estimates for the Sellers’ Quality Claims and Shirking Rates
Treatment Type
Floor (Cards)
Floor
(Tickets)
NoGrading
Floor
(Tickets)
Announce
Grading
Floor
(Tickets)
Grading
Lab-
Market
Variable
Quality
Claim
(1)
Shirk
(2)
Quality
Claim
(3)
Shirk
(4)
Quality
Claim
(5)
Shirk
(6)
Quality
Claim
(7)
Shirk
(8)
Quality
Claim
(9)
Local dealer Ϫ1.4
(4.4)
Ϫ1.2
(1.8)
.06
(.2)
.24
(.6)
Ϫ1.3
(2.5)
Ϫ2.3
(4.8)
Ϫ2.2
(2.2)
Ϫ1.0
(1.4)
Ϫ.30
(.8)
Price .01
(1.9)
. . . Ϫ.03
(1.7)
. . . Ϫ.01
(.3)
. . . .04
(1.3)
. . . Ϫ.01
(.7)
Constant .6
(2.0)
1.1
(2.4)
.7
(1.7)
Ϫ.1
(.3)
.6
(1.5)
1.0
(2.8)
.3
(.5)
1.0
(2.4)
Ϫ.5
(.5)
Observations 100 53 60 36 54 24 36 17 60
Note.—The dependent variable in the quality claim specification equals one if the dealer claimed quality, zero
otherwise. The dependent variable in the shirking specification equals one if the dealer shirked on the quality claim,
zero otherwise. The Floor-Grading model did not converge; thus estimates are derived from equations that are estimated
separately. The second-stage shirking equation for Lab-Market cannot be estimated because no dealers shirked in the
Lab-Market treatment. t-ratios (in absolute value) are beneath coefficient estimates.
ment, local dealers provide fewer quality claims, and this estimate is
statistically significant at the level. As is clear from columns 5p ! .05
and 7 of table 7, similar insights are obtained in the ticket treatments.
Evidence in favor of the second part of result 6—conditional on claiming
quality, when third-party verification is possible, local dealers shirk less
often in the field—can also be found in all three shirking specifications.
Whereas the coefficient estimates in Treatments Floor (Cards) and
Floor-Grading specifications are significant only at the ( )p ! .07 p ! .17
level with a two-sided alternative (cols. 2 and 8), the coefficient estimate
in the Floor AnnounceGrading model is statistically significant at the
level (col. 6).p ! .01
As in the spirit of the inquiry into result 4, one can question whether
the increased quality promises and deliveries from local dealers are due
purely to reputational concerns or have an element of social preferences.
Examining data collected in Treatment Lab-Market and Treatment
Floor (Ticket Stubs) lends insights into this issue and leads to the
final result.
Result 7. In the laboratory, or when third-party verification is not
possible, local and nonlocal dealers make similar claims of quality, and
conditional on claiming quality, shirk to the same extent.
As table 6 reveals, in Treatments Lab-Market and Floor-NoGrading,
local and nonlocal dealers behave quite similarly. In Treatment FloorNoGrading,
local dealers make quality claims in 22 of 36 (61 percent)
cases, whereas nonlocal dealers make quality claims in 14 of 24 (58
32 journal of political economy
percent) cases. Likewise, conditional on claiming quality, local dealers
in Treatment Floor-NoGrading shirk in 18 of 22 cases—that is, in 82
percent of transactions, local dealers provide lower quality than promised—whereas
71 percent (10 of 14) of nonlocal dealer transactions
should be considered shirking. In Treatment Lab-Market, both dealer
types make considerably fewer quality claims and never shirk.
Estimating the bivariate probit model with sample selection in equations
(4a) and (4b), I find that the observed differences across local
and nonlocal dealers are statistically insignificant at conventional levels
for both Treatment Lab-Market and Treatment Floor-NoGrading data.
For Treatment Floor-NoGrading, columns 3 and 4 of table 7 show that
both local and nonlocal dealers make similar quality claims and, conditional
on providing a quality claim, shirk to the same extent. For
Treatment Lab-Market data, column 9 of table 7 provides similar evidence:
local and nonlocal dealers provide a similar number of quality
claims.22
IV. Concluding Remarks
This study provides a framework for measuring social preferences and
reputation effects using a series of laboratory and field experiments. In
doing so, it showcases the desirability of building a bridge between the
lab and the field. In a methodological sense, this bridge permits a test
of whether laboratory behavior is a good indicator of behavior in the
field. The finding that agents behave differently in tightly controlled
laboratory experiments than in their naturally occurring environment
poses an important challenge to laboratory studies that measure individual
propensities. More generally, these results underscore the role
that field experiments can play in empirical economics. For example,
experimentalists typically take stock in results from a series of laboratory
experiments. This study pushes this notion in a new direction by shedding
light on the importance of including results from field experiments
within this evidentiary system. In this light, the results show that field
experiments can help to uncover the causes and underlying conditions
necessary to produce data patterns observed in the lab.
While the data suggest that social preferences do not have a major
impact in these particular markets, such results, of course, do not necessarily
preclude social preferences from having import within other
economic domains. Some scholars have argued that such preferences
are evident in domains in which the pressures of the market are absent
(e.g., the charitable fund-raising work of Falk [2004]). I view this class
22
As table 6 reveals, the shirking equation cannot be estimated for the Treatment LabMarket
data because no dealers—local or nonlocal—shirked in the laboratory experiment.
social preferences 33
of studies as fruitful in that they represent good examples of domains
in which such preferences might be significant.
The data are also sufficiently rich to speak to how reputation effects
and professional certification influence market performance. For example,
the data support the view that reputation effects enhance the
quality of goods. This insight is consonant with results in the work of
Akerlof (1970), who provides evidence on the operation of markets in
developing countries that demonstrates a positive association between
reputation effects and the quality of goods. Furthermore, empirical results
suggest that third-party enforcement of contracts is important: the
addition of professional quality certifiers enhances market performance
and supports the conjecture that the private market can solve the lemons
problem through third-party verification. This result might be viewed
as a test of the Klein and Leffler (1981) model in that local sellers cheat
less than nonlocal sellers when quality is measurable, but this is not the
case when quality is not easily measurable. This finding indicates that
reputations cannot work without information, suggesting that reputation
and the monitoring of quality are complements.
Appendix
Further Experimental and Institutional Details
A. Determining and Valuesc(q) v(q)
The values of were chosen to represent the dealer cost to replace a 1990c(q)
Leaf Frank Thomas card of various quality levels.23
The values are taken from
the standard price guide for baseball cards: Beckett Baseball Cards Monthly. For
each single type of ungraded card, Beckett collects pricing information from about
110 card dealers throughout the country and publishes a “high” and “low” price
reflecting current selling ranges for several quality variants. The high price
represents the highest reported selling price and the low price represents the
lowest price one could expect to find with extensive shopping. Assuming that
dealers’ replacement costs are roughly equivalent to the reported “low” price,
I use the “low” prices from Beckett for 1990 Leaf Thomas cards that would grade
PSA 6, 7, 8, 9, and 10 to approximate values.c(q)
Determining values in equation (2) to approximate the gains from tradev(q)
is more difficult since consumer demand curves are not readily observable. In
this case I considered results from two approaches: (i) taking the “high” prices
from Beckett for 1990 Leaf Thomas cards that would grade PSA 6, 7, 8, 9, and
10 and (ii) gathering statements of value for 1990 Leaf Thomas cards that would
grade PSA 6, 7, 8, 9, and 10 via a contingent survey in the spirit of Cummings
23
I chose this particular card for all treatments because of my experience in evaluating
the attributes of the card over the past 15 years (as a dealer and a consumer), Thomas’s
popularity, and the fact that this variant represents his “rookie card”—typically a player’s
most sought-after card. These latter two factors help to explain the extensive interest in
the card among broad classes of collectors.
34 journal of political economy
and Taylor (1999).24
The contingent valuation experiment, which was run on
the floor of a sports card show, randomly allocated consumers into one of five
treatments (PSA 6, 7, 8, 9, or 10). Thirty subjects were placed into each treatment,
for a total of 150 subjects. Subjects were asked to state their true value for a
1990 Leaf Thomas card in a contingent valuation scenario. In addition, they
were warned about hypothetical bias, which oftentimes arises in such situations,
with a “cheap talk” script.25
Previous efforts have found that a contingent survey
that includes a cheap talk script has yielded consumer values that closely match
actual values (e.g., List 2001). Most important for our purposes, mean values
from the contingent survey are in the range of published Beckett “high” prices;
thus I use these values and make , $8, $15, $30, and $80 for , 2,v(q) p $6 q p 1
3, 4, and 5 (PSA 6, 7, 8, 9, and 10).
B. Sports Card and Sports Ticket Grading
Each year, sports card companies design and print sets of sports cards depicting
players and events from the previous season. Once the print run of a particular
set has been completed, the supply of each distinct card in the set is fixed. The
value of a particular card depends on its scarcity, the player depicted, and the
physical condition of the card, that is, the condition of its edges, corners, and
surface and centering of the printing. To track card condition, people often
use a 10-point scale. For example, a card with flawless characteristics under
microscopic inspection would rate a perfect 10, whereas defects, including minor
wear on the corners, would decrease the card’s grade to a 7. The card’s overall
grade is computed via the aggregation of the various characteristics.
Professional Sports Authenticators is the industry leader in grading services,
and its parent company became publicly traded in 1999 (Collectors Universe,
under NASDAQ ticker symbol CLCT). PSA has graded more than 7 million
sports cards since its inception in 1987. Professional grading is voluntary and
costs $6–$100 per card, depending on package size and requested turnaround
time. Importantly, the fee is independent of the actual grade received. Graded
cards are encased in plastic and sealed with a sonic procedure that makes it
virtually impossible to open and reseal the case without evidence of tampering.
PSA adopted integer grades from 1 to 10, where a 10 is considered gem mint
and commands a premium price. A PSA 9 card is considered mint and is the
next most valuable card type. As witnessed by the and vectors used inc(q) v(q)
treatments I and II, card values are convex in the grade received. Importantly,
Jin, Kato, and List (2004) provide evidence suggesting that even under PSA’s
coarse grading system, certification reveals important information to ordinary
24
I could have gathered willingness to pay (or willingness to accept) values by auctioning
off Thomas cards using an incentive-compatible auction institution (i.e., a Vickrey secondprice
auction), but market prices should influence bids, leaving me with a vector of bids
that roughly estimate the perceived market price adjusted for transactions costs.
25
The cheap talk script is similar to that in List (2001) and notes that “In most questions
of this kind, folks seem to have a hard time doing this. They act differently in a hypothetical
situation, where they don’t really have to pay money, than they do in a real situation,
where they really have to pay money. We call this ‘hypothetical bias’. ‘Hypothetical bias’
is the difference that we continually see in the way people respond to hypothetical situations
as compared to real situations. So, if I was in your shoes, and I was asked to make
a choice, I would think about how I feel about spending my money this way. When I got
ready to choose, I would ask myself: if this was a real situation, do I really want to spend
my money this way?”
social preferences 35
consumers. Yet they report that dealers gain no information from a card’s PSA
grade, suggesting that dealers are able to evaluate quality as well as PSA.
Sports tickets and ticket stubs have recently gained enough market acceptance
to merit professional grading. Ticket supply, of course, depends on the stadium
size of the event and the proportion of fans in attendance who preserved their
ticket stubs (or in the case of unused tickets, the number of fans who left their
tickets unused). Ticket grading is similar to sports card grading: an identical 10point
scale is used, and sharpness of corners, centering of printing, sharp focus,
and original gloss are very important. Furthermore, staining, printing imperfections,
and print quality of crucial game information are also important in
determining ticket quality.
References
Akerlof, George A. 1970. “The Market for ‘Lemons’: Quality Uncertainty and
the Market Mechanism.” Q.J.E. 84 (August): 488–500.
———. 1982. “Labor Contracts as Partial Gift Exchange.” Q.J.E. 97 (November):
543–69.
Akerlof, George A., and Janet L. Yellen. 1988. “Fairness and Unemployment.”
A.E.R. Papers and Proc. 83 (May): 44–49.
———. 1990. “The Fair Wage–Effort Hypothesis and Unemployment.” Q.J.E.
105 (May): 255–83.
Andreoni, James, and John Miller. 2002. “Giving According to GARP: An Experimental
Test of the Consistency of Preferences for Altruism.” Econometrica
70 (March): 737–53.
Berg, Joyce, John W. Dickhaut, and Kevin A. McCabe. 1995. “Trust, Reciprocity,
and Social History.” Games and Econ. Behavior 10 (July): 122–42.
Bewley, Truman F. 1995. “A Depressed Labor Market as Explained by Participants.”
A.E.R. Papers and Proc. 85 (May): 250–54.
Blinder, Alan S., and Don H. Choi. 1990. “A Shred of Evidence on Theories of
Wage Stickiness.” Q.J.E. 105 (November): 1003–15.
Bolton, Gary E., and Axel Ockenfels. 2000. “ERC: A Theory of Equity, Reciprocity,
and Competition.” A.E.R. 90 (March): 166–93.
Brown, Martin, Armin Falk, and Ernst Fehr. 2004. “Relational Contracts and the
Nature of Market Interactions.” Econometrica 73 (May): 747–80.
Camerer, Colin F. 2003. Behavioral Game Theory: Experiments in Strategic Interaction.
Princeton, NJ: Princeton Univ. Press.
Camerer, Colin F., and Keith Weigelt. 1988. “Experimental Tests of a Sequential
Equilibrium Reputation Model.” Econometrica 56 (January): 1–36.
Charness, Gary. 1996. “Attribution and Reciprocity in an Experimental Labor
Market.” Manuscript, Univ. California, Santa Barbara.
Charness, Gary, Guillaume R. Frechette, and John H. Kagel. 2004. “How Robust
Is Laboratory Gift Exchange?” Experimental Econ. 7 (June): 189–205.
Charness, Gary, and Matthew Rabin. 2002. “Understanding Social Preferences
with Simple Tests.” Q.J.E. 117 (August): 817–69.
Cox, James C. 2004. “How to Identify Trust and Reciprocity.” Games and Econ.
Behavior 46 (February): 260–81.
Cummings, Ronald G., and Laura O. Taylor. 1999. “Unbiased Value Estimates
for Environmental Goods: A Cheap Talk Design for the Contingent Valuation
Method.” A.E.R. 89 (June): 649–65.
Dufwenberg, Martin, and Georg Kirchsteiger. 2004. “A Theory of Sequential
Reciprocity.” Games and Econ. Behavior 47 (May): 269–98.
36 journal of political economy
Falk, Armin. 2004. “Charitable Giving as a Gift Exchange: Evidence from a Field
Experiment.” IZA Working Paper no. 1148, Inst. Study Labor, Bonn.
Falk, Armin, and Urs Fischbacher. Forthcoming. “A Theory of Reciprocity.”
Games and Econ. Behavior.
Fehr, Ernst, and Armin Falk. 1999. “Wage Rigidity in a Competitive Incomplete
Contract Market.” J.P.E. 107 (February): 106–34.
Fehr, Ernst, and Simon Ga¨chter. 2000. “Fairness and Retaliation: The Economics
of Reciprocity.” J. Econ. Perspectives 14 (Summer): 159–81.
Fehr, Ernst, Simon Ga¨chter, and Georg Kirchsteiger. 1997. “Reciprocity as a
Contract Enforcement Device: Experimental Evidence.” Econometrica 65 (July):
833–60.
Fehr, Ernst, Georg Kirchsteiger, and Arno Riedl. 1993. “Does Fairness Prevent
Market Clearing? An Experimental Investigation.” Q.J.E. 108 (May): 437–60.
Fehr, Ernst, and John A. List. 2004. “The Hidden Costs and Returns of Incentives—Trust
and Trustworthiness among CEOs.” J. European Econ. Assoc. 2 (September):
743–71.
Fehr, Ernst, and Klaus M. Schmidt. 1999. “A Theory of Fairness, Competition,
and Cooperation.” Q.J.E. 114 (August): 817–68.
Ga¨chter, Simon, and Armin Falk. 2002. “Reputation and Reciprocity: Consequences
for the Labour Relation.” Scandinavian J. Econ. 104 (1): 1–26.
Hannan, R. Lynn, John H. Kagel, and Donald V. Moser. 2002. “Partial Gift
Exchange in an Experimental Labor Market: Impact of Subject Population
Differences, Productivity Differences, and Effort Requests on Behavior.” J.
Labor Econ. 20 (October): 923–51.
Hoffman, Elizabeth, Kevin McCabe, and Vernon L. Smith. 1996. “Social Distance
and Other-Regarding Behavior in Dictator Games.” A.E.R. 86 (June): 653–60.
Holt, Charles A., and Roger Sherman. 1990. “Advertising and Product Quality
in Posted-Offer Experiments.” Econ. Inquiry 28 (January): 39–56.
Jin, Ginger, Andrew Kato, and John A. List. 2004. “That’s News to Me! Information
Revelation in Professional Certification Markets.” Working paper, Univ.
Maryland.
Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler. 1986. “Fairness as
a Constraint on Profit Seeking: Entitlements in the Market.” A.E.R. 76 (September):
728–41.
Katz, Lawrence F. 1986. “Efficiency Wage Theories: A Partial Evaluation.” In
NBER Macroeconomics Annual, edited by Stanley Fischer. Cambridge, MA: MIT
Press.
Klein, Benjamin, and Keith B. Leffler. 1981. “The Role of Market Forces in
Assuring Contractual Performance.” J.P.E. 89 (August): 615–41.
List, John A. 2001. “Do Explicit Warnings Eliminate the Hypothetical Bias in
Elicitation Procedures? Evidence from Field Auctions for Sportscards.” A.E.R.
91 (December): 1498–1507.
———. 2004a. “The Nature and Extent of Discrimination in the Marketplace:
Evidence from the Field.” Q.J.E. 119 (February): 49–89.
———. 2004b. “Neoclassical Theory versus Prospect Theory: Evidence from the
Marketplace.” Econometrica 72 (March): 615–25.
———. 2004c. “Testing Neoclassical Competitive Theory in Multilateral Decentralized
Markets.” J.P.E. 112 (October): 1131–56.
———. 2005. “The Behavioralist Meets the Market: Measuring Social Preferences
and Reputation Effects in Actual Transactions.” Working Paper no. 11616
(September), NBER, Cambridge, MA.
List, John A., Robert P. Berrens, Alok K. Bohara, and Joe Kerkvliet. 2004. “Ex-
social preferences 37
amining the Role of Social Isolation on Stated Preferences.” A.E.R. 94 (June):
741–52.
Macaulay, Stewart. 1963. “Non-contractual Relations in Business: A Preliminary
Study.” American Sociological Rev. 28 (February): 55–67.
Miller, Ross M., and Charles R. Plott. 1985. “Product Quality Signaling in Experimental
Markets.” Econometrica 53 (July): 837–72.
Orne, Martin T. 1962. “On the Social Psychology of the Psychological Experiment:
With Particular Reference to Demand Characteristics and Their Implications.”
American Psychologist 17 (11): 776–83.
Rabin, Matthew. 1993. “Incorporating Fairness into Game Theory and Economics.”
A.E.R. 83 (December): 1281–1302.
Rosenthal, Robert. 2002. “Experimenter and Clinician Effects in Scientific Inquiry
and Clinical Practice.” Prevention & Treatment 5 (October 18). http://
journals.apa.org/prevention/.
Sitkin, Sim B., and Nancy L. Roth. 1993. “Explaining the Limited Effectiveness
of Legalistic ‘Remedies’ for Trust/Distrust.” Organization Sci. 4 (August): 367–
94.
Sobel, Joel. 2002. “Social Preferences and Reciprocity.” Manuscript, Univ. California,
San Diego.
Stigler, George J. 1981. “Economics or Ethics?” In The Tanner Lectures on Human
Values, vol. 2, edited by Sterling M. McMurrin. Cambridge: Cambridge Univ.
Press.
van de Ven, Wynand P. M. M., and Bernard M. S. van Praag. 1981. “The Demand
for Deductibles in Private Health Insurance: A Probit Model with Sample
Selection.” J. Econometrics 17 (November): 229–52.