|
...prices but do not receive better quality and, in fact, are defrauded more often. Online seller reputation is effective for identifying good-faith sellers. But conditional on completed auctions, reputable sellers do not provide better quality. Evidence also suggests that high-claim sellers target less-experienced buyers. We attribute these patterns to two loopholes in the eBay rating system.
1. Introduction
* Theorists argue that if consumers do not observe product quality before purchase, they may use price, advertising, or seller reputation to infer quality. The inference is validated by monotone relationships between these "signals" and true quality. For example, in Nelson (1970), only high-quality sellers can afford costly advertising because high-quality products are more likely to attract repeat purchases. (1) In a slightly different setting, Klein and Leffler (1981) argued that high price in the company of high quality generates a continual stream of rental income, which motivates sellers to provide high quality in the long run rather than cheat and disappear in the short run. Similarly, Shapiro (1983) suggested that sellers who sold high quality-products in the past have an incentive to maintain their reputation, because good reputations are rewarded with high prices and high profits in the future.
An ideal test of the theory is to link true quality with price, advertising, and reputation. However, since quality is rarely observable, researchers often document positive correlations between price, advertising, and reputation and conclude that the positive correlation(s) support the theory. (2) Such exercises are particularly popular in the context of Internet auctions, because most auction sites (such as eBay) offer the complete record of transactions plus feedbacks from previous transactions. (3)
Unlike previous studies, we design a field experiment in Internet auctions, obtain a key variable--true quality (4)--and conduct a direct comparison between quality, price, seller claims, and seller reputation. We focus on baseball cards whose physical qualities have not been professionally certified at the time of sale. As detailed in Section 2, quality is one of the most important determinants of card value, but the quality of an ungraded baseball card is very difficult to detect online. In fact, the only information available to buyers in these auctions is quality claims and descriptions made by sellers in the auction listings.
The evidence was gathered in two steps. First, a seven-month eBay watch revealed a stunning fact: 33%-51% higher prices were paid to sellers claiming to be selling high-quality cards (as opposed to modest or no claims). This would be reasonable if sellers making grand claims actually delivered better cards or provided more reliable service. We checked this in the second step. Immediately following the market watch, we purchased ungraded cards with the same types of claims and had them professionally evaluated. These cards were purchased systematically so that half of the sample came from high-claim sellers and the other half from modest- or no-claim sellers. In a reverse of the theory, sellers making the best claims were more likely to commit fraud by defaulting (no delivery) or sending counterfeits. Even conditional on actual deliveries of authentic cards, the average quality from high-claim sellers was indistinguishable from the average quality from sellers with more modest claims.
Given the freedom to make nonverifiable claims (aside from the reputation concerns that we discuss below), the incredibility of seller claims should not be a surprise to theorists. What is surprising is the amount of price premium that buyers are willing to pay for empty claims. After examining a number of alternative explanations, we conclude that the most likely answer is some buyers being uninformed. These buyers, unlike the uninformed buyers assumed in most of the literature, (5) are not only uninformed of the quality of a specific seller (or the quality of goods he or she sells), but also hold incorrect beliefs on the distribution of quality in the market. More important, the prices we observed would not correct such misbeliefs. On the contrary, if an uninformed buyer watches the eBay market as we did and is naive enough to believe the application of the classical theory to the eBay setting, he is likely to fall for the empty claims.
Does the eBay rating system help lower the risks of cheating and signal better product quality? Our data suggest that reputable sellers are less likely to make bold claims and less likely to default or send counterfeits. However, conditional on authentic delivery, the actual card quality is uncorrelated with seller ratings.
Our eBay watch data also confirm a puzzle documented in other studies of online reputation: buyers are more willing to place bids on items offered by reputable sellers, but conditional on completed auctions, seller ratings have no significant impact on the winning price. Furthermore, we find that the boldest claims not only yield price premiums out also increase the probability of sale. This suggests that if a seller is concerned only about the current sale, the lower likelihood of sale associated with lower seller reputation can easily be overcome by making an incredible claim of superior quality. This explains why bold claims are often made by sellers with relatively low ratings. If naive buyers understood the negative correlation between seller claims and seller reputation, even seller reputation alone could be very effective in helping them avoid fraud. This suggests that the key problem is not the lack of signals; rather, it is the confusion that results from mixing good signals like the eBay seller rating with misinterpretations of bad signals like seller claims.
The confusion may be attributable to two reputation loopholes. First, eBay ratings are universal across categories and roles. It is easy to build up an eBay rating by buying cheap items and then use those reputation points to sell expensive items. (6) A random check of feedback profiles also suggests that most feedbacks focus on the common conduct behind various transactions: the ability to complete the transaction in a timely manner. Feedback on the actual quality of the delivered item is rare, as it is hard to compare quality across categories and the absolute level of quality is subject to interpretation. This explains why the information that seller ratings successfully convey in our dataset is the seller's inclination to cheat (by making bold claims, defaulting, or sending counterfeits) rather than the actual physical quality conditional on authentic delivery.
To make matters worse, one may switch between anonymous identities with little cost in eBay. In the experiment we encountered two fraudulent sellers who intentionally built up positive ratings, committed a series of defaults, received over 20 complaints, and abandoned the accounts soon afterward. Because the cheater can switch to a clean slate, warnings from today's buyers cannot reach new buyers tomorrow. Like universal ratings, this loophole reduces the precision and accessibility of information on past seller conduct, casting doubt on how applicable the classical theory is to eBay.
The rest of the article is organized as follows. In Section 2 we explain why baseball cards are especially suitable for this study. Section 3 presents empirical findings from our market watch, followed by experimental design in Section 4 and experimental results in Section 5. We discuss the implications from our empirical findings Section 6, and Section 7 concludes.
2. Baseball cards and eBay
* Three features of baseball cards make them an excellent example. First, each type of baseball card is a homogeneous good and card quality is the most important determinant of card value. Second, industry-accepted professional grading services exist to verify card quality, identity, and authenticity. This allowed us to directly observe quality and seller performance. Finally, online auction is one of the most complained-about markets for consumer frauds, and quality concerns pertaining to online ungraded baseball cards are likely to be present for a multitude of goods. We now present institutions in support of these arguments.
Each year, card companies design and print sets of cards depicting players and events from the previous season. Once the print run of a particular set ends, the supply of each distinct card in the set is fixed. (7) The value of a particular card depends on its scarcity, the player depicted, and the condition of the card. By card condition, we mean the physical condition of the edges, corners, and surface, and the centering of the printing. People often use a 10-point scale to track card condition. For example, flawless characteristics (even under microscopic inspection) will rate a perfect 10, while obvious defects to the naked eye like minor wear on the card's corners might result in a corners grade of 7. The card's overall grade is computed off all four characteristics.
Card flaws are often hard to detect with the naked eye. In retail markets, buyers often use a magnifying glass to carefully examine card condition before purchase. One may also refer to independent, professional grading for a more accurate measure of quality. Because professional grading is voluntary and costs $6 to $20 per card, (8) this practice divides the market into two groups: graded cards and ungraded cards. 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. The case indicates the grading service, grade received, and a bar code with serial number that identifies the particular copy of the card. Anyone with Internet access can go to the grader's web site and verify the card's grade by looking up the serial number. As of 2005, Professional Sports Authenticator (PSA) and Beckett Grading Service (BGS) remain the largest and most respected of the existing 10-15 grading services.
Single baseball cards can be traded online or offline. Offline outlets include retail stores and local card shows, both involving substantial transaction costs in finding trading partners. These transaction costs are significantly reduced in online auctions. (9) eBay offered the first standardized, user-friendly auction site where individual sellers and buyers could meet and transact conveniently. Since its public stock offering in 1998, eBay has always been the most popular online place to trade single baseball cards. (10)
Despite transaction cost savings, online trading intensifies the information asymmetry between buyers and sellers, especially for ungraded cards. An online seller can describe the card condition in a brief title and web page. He can also supplement the text description by one or multiple scanned pictures. If it is a graded card, the picture will reveal the name of the grader and the serial number that identifies the exact card copy. If it is an ungraded card, however, it is impossible to prove that the card depicted is the one for sale. Even if it were the card for sale, poor scan quality reduces the informational value of the picture. While centering on the front may still be observable, a blurry scan makes it impossible to judge the condition of the card corners, edges, and surfaces. Also, sellers sometimes do not include a picture of the back of the card, which could be in a different condition from the front. The other auction settings are standard, as described in Lucking-Reiley (2000).
Whether buying or selling, all trading parties refer to a standard price guide: 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 Near Mint-Mint (8) copies. It is widely agreed that the retail transaction prices are closer to the low end than to the high end. For graded cards, Beckett lists price ranges by grading company and grade. Card value is convex in grade: for example, the July 2001 low price for a Griffey Jr. rookie card is $60 for PSA8, $150 for PSA9, and $1,200 for PSA10. The book price allows us to control for market differences across cards and across time.
Online auction is one of the most complained-about markets for consumer frauds. According to the Federal Trade Commission (FFC), Internet auction fraud was ranked as the second-most common category of consumer complaint, accounting for 10% of the 204334 Internet-related complaints in 2001. For three years in a row, auction fraud was the most frequent offense reported to the FBI's Internet fraud unit. More alarmingly, total reported losses involved in Internet auction frauds tripled from $17 million in 2002 to $54 million in 2003.
Quality concerns stated above for baseball cards apply to a multitude of goods in online auction. The most direct analogy is to other collectibles such as stamps, coins, and antiques, all of which have similar industry features and professional grading services. It is also common to observe seller claims of quality in other top-ten trading categories, such as used cars, consumer electronics, books, clothing, accessories, toys, and jewelry. In fact, the most common complaints about Internet auctions are that "after sending the payment, buyers may receive an item that is less valuable than promised, or worse yet, nothing at all." (11)
eBay has adopted several measures to combat fraud. First of all, eBay requires either a verifiable email address or a real credit card number for anyone who wishes to buy or sell in eBay. However, because identity theft is the most common form of fraud, it is obvious that a fraudulent seller could open multiple dummy accounts using stolen credit card numbers. Second, eBay has an entire division devoted to investigating fraud complaints. At the time of our experiment, the Fraud Protection Program in eBay reimburses up to $200 per item (minus a $25 processing fee) for buyers who did not receive any delivery after 30 days of payment. However, for a specific buyer, the reimbursement is limited to three claims per six months. Also, unlike a physical auction house, eBay positions itself as a marketplace organizer and therefore does not guarantee the quality of goods actually delivered. (12) Because it is difficult to identify a unique copy of an ungraded baseball card, even retail stores do not allow any return or exchange after sale. Due to the same reason, an online ungraded card that turns out to be counterfeit or below the claimed quality is usually difficult to be returned, exchanged, or reimbursed.
The third and probably most widely used device to combat fraud is the eBay feedback forum. Every eBay user has a "feedback profile" made up of comments from other eBay users--an official "reputation." This profile consists of a numerical score next to the user ID in the...
NOTE: All illustrations and photos
have been removed from this article.
More articles from RAND Journal of Economics
An empirical study of auction revenue rankings: the case of municipal ..., December 22, 2006
Looking for additional articles?
Search our database of over 3 million articles.
Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication
name or publication date.
About Goliath
Whether you're looking for sales prospects, competitive information, company
analysis or best practices in managing your organization,
Goliath can help you meet your business needs.
Our extensive business information databases empower business
professionals with both the breadth and depth of credible,
authoritative information they need to support their business
goals. Whether it be strategic planning, sales prospecting,
company research or defining management best practices -
Goliath is your leading source for accurate information.
|
