Source
Automatically imported from: http://commons.somewhere.com:80/rre/2000/RRE.The.market.in.market.html
Content
| | | | --- | --- | | Red Rock Eater Digest | Most Recent Article: Wed, 30 Aug 2000 |
``` ---
This message was forwarded through the Red Rock Eater News Service (RRE). You are welcome to send the message along to others but please do not use the "redirect" option. For information about RRE, including instructions for (un)subscribing, see http://dlis.gseis.ucla.edu/people/pagre/rre.html
---
The market in marketplaces: Some notes on the dubious case of eBay
Phil Agre http://dlis.gseis.ucla.edu/pagre/ May 2000
2400 words.
This is the draft of 7 May 2000. You are welcome to forward to it anyone who might be interested, but please don't quote from it.
The wired world reconfigures intermediaries, but how? One case study is provided by the online auction service, eBay. EBay is exhibit A for those who see the Internet bringing about a frictionless world of efficient markets, and so it makes an especially important study in that mysteriously underinvestigated topic, the market in marketplaces. Some positive results are available. Peter Kollock, for example, investigated the extent to which eBay sellers with high "feedback" ratings were able to command higher auction prices for their goods. Clearly in this sense the eBay market is working correctly: higher feedback ratings did in fact predict higher prices. I don't quarrel with this. But I want to suggest that eBay's success derives not entirely from its efficiency as a marketplace, but also from the inefficiency of the market in marketplaces.
A note on method. Although my arguments are motivated by personal experience with eBay trading, I do not claim that my experiences prove anything. Rather, I will offer a series of six in-principle economic arguments that make the same kind of internal sense as most economic arguments, and I invite others to test or refute them.
First argument: Incentive to leave overly positive feedback. Auctions among geographically distant strangers who may lack any other social bonds pose obvious problems of trust. The advertised goods may not exist, or may have been described misleadingly, and even when fraud is absent transactions can go wrong in many ways both large and small. EBay claims to overcome this problem in part through its "feedback forum", in which buyers and sellers can leave comments on one another. Newcomers are routinely struck by the ecstatic tone of most of these comments:
Fast payment, friendly emails. Great ebayer! Highly Recommended!!!
Prompt Email; Prompt Payment; Perfect Purchaser. A++++
GOOD COMMUNICATION, F AST PAYMENT +++++
Quick payment, excellent transaction. Recommended !! A+++++++++++++++
Amazingly efficient. Great Buyer. Would recommend to any seller. Thanks, Phil.
These are all part of my own feedback, and except maybe in one case (where I ended up giving the guy some advice about sending his son to college) I can testify that they all greatly exaggerate interactions that, while efficient enough, were not overwhelmingly fast or pleasant or otherwise wonderful.
What explains this kind of exaggeration? Observe that, from an economic standpoint, the feedback mechanism exhibits externalities. The feedback that I leave on another eBay community member is useful primarily to strangers -- other members of the eBay community with whom I will probably never have any dealings. Those people, moreover, usually have no way to determine whether my feedback is accurate. I am providing those people with free information. So why even do it? Simple altruism is part of it, and curiosity about the workings of the new medium. But even slight experience with eBay suggests another answer: the buyer has an interest in providing positive feedback to the seller so that the seller will provide positive feedback to the buyer, and vice versa. The two parties have a pecuniary interest in collaborating to provide a misleadingly positive view of one another. After all, as Kollock points out, positive feedback can increase the prices one gets for one's goods. And my experience is certainly that this kind of collaboration is routine, and even overt.
Second argument: Incentive against leaving negative feedback. If a buyer feels abused by a seller, negative feedback provides a mechanism for warning other potential buyers of a potential problem. But the buyer receives no benefit for issuing this helpful warning, and at least two kinds of incentive not to. To start with, the eBay feedback mechanism provides community members with the opportunity to respond to any negative feedback, and the person who provided the negative feedback cannot surreply. As a result, someone who receives negative feedback can simply lie, and the person who left the negative feedback will have no further recourse -- except by issuing a separate, second negative feedback, which may not even be possible if the time allowed for feedback has elapsed, and which in any case invites a nuclear war of mutual recriminations. Another example from my experience:
Complaint: goods not as described, responded to very mild complaint w/ belligerent nonsense
Response: SENT ITEM LISTED, RECEIVED NASTY INCOHERRENT EMAILS! OFFERED $JUST WANTED GRIPE
My complaint is self-explanatory. What you wouldn't know is that the seller's reply is just false. For example, far from "offering $", the seller had falsely told eBay that I had not paid for the goods, which caused the eBay server to send me a message (presumably automated) demanding that I send $ to the seller! I replied with bank details of the payment that had already cleared and that was that, but these lies about me are now posted in a global public forum.
Lies about me in a seller's feedback profile are unlikely to hurt me that much, since it would take some work for an individual to connect me to them. But this seller wasn't satisfied with a false reply to my negative feedback, and instead went on to retaliate:
Complaint: DO NOT RECOMMEND! AVOID! DID NOT READ AUCTION, COMPLAINTS MADE LITTLE SENSE!
Response: v. mild complaint brought belligerent nonsense, now retaliating for neg feedback
My words again speak for themselves, and obviously most reasonable person will understand that this person was a vengeful psycho, or else just having a bad month. Someone who seriously wanted to investigate the matter would also find that this particular seller has done this more than once. The point is, a rational eBay community member will think twice about providing free information to the commons in the form of negative feedback about a market wrongdoer.
Third argument: The surprising ease of accumulating feedback points. (I was going to say "deceptive", but that word has a legal meaning.) Between the incentives to leave exaggerated positive feedback and the incentives not to leave negative feedback, it is not surprising that the great majority of feedback on eBay is positive. Furthermore, it is my impression that a significant proportion of eBay activity is meant largely to accumulate feedback points, because those positive points are a kind of capital. Because most eBay auctions can be completed without human intervention by eBay personnel, fees are moderate, and so the cost of buying goods on eBay and then immediately selling them on eBay -- thus accumulating two points of feedback -- may be less than the value of the feedback. This is particularly true with categories of auctions that involve goods, such as random postage stamps, that are cheap both to buy and to ship.
Feedback points can be misleading in other ways. Entries in the automatically generated table of feedback for a user do not mention the dollar value of the transaction (although for a 30-day period the final price can be found by linking back to the auction page for the item), and most do not mention the nature of the goods either. As a result, an eBay community member cannot tell whether a seller with 100 points can be trusted to sell a $1000 computer. What is more, eBay does not make it easy to determine the total value of the auctions that a seller currently has under way (though one can compute the total by hand); yet (according to the newspapers) serious fraud has been committed by sellers who behave themselves long enough to build up an impressive point count, and who then commence a large number of fraudulent high-value auctions at one time.
Fourth argument: The inefficient consequences of auction addiction. From an emotional point of view, participating in an auction is very much like gambling. Bids are very much like bets, and the emotional charge of the bid dissipates as soon as somebody else outbids. EBay even makes a big point out of having "won" the betting, treating it as an occasion for congratulations. The eBay auction mechanism enables bidders to provide a secret maximum bid price, and so if all bidders were rational then nobody would have a rational reason to place a second bid (unless the product somehow became more valuable, perhaps because of new information one had discovered about it). But in reality, a significant number of eBay bidders relate to the auction in an addictive manner, meaning that their bidding strategies are designed in part to get an emotional kick that is not rationally related to the bidding process. Addictive bidding hurts the addicted individual's family members, it inflates prices in the marketplace, and it keeps goods out of the hands of the people who rationally value them the most.
Fifth argument: Limitations of the bidding mechanism. Addictive bidding has another consequence, much more subtle but also much more significant. Because of the secret maximum bid mechanism, it should not matter at what point in the auction one places a bid. If the real value that one places on the product is $100, the rational strategy would simply be to bid $100 and then forget about it until the auction is over. (Things are more complicated if one is participating in several auctions, however, or if one needs something in a hurry.) But in practice, the bidding mechanism does not work perfectly. Addictive bidders, or those who are simply misinformed about the logic of the bidding process, are more likely to wait until the last moment to place their bids. They determine their bids not relative to their own personal valuation of the product, but relative to the bids that others have placed. It follows that a rational bidders who value the product at $100 should also wait until the last minute to bid, because otherwise they will lose out to the irrational bidders. (This is not completely true, because bids for the same amount are resolved in favor of the earlier bidder at that amount, thus retaining a slight incentive for early bidding.) But now it is in everybody's interest to bid at the last minute. And the eBay server, inevitably, has a limited capacity to process large numbers of last-moment bids on the same product: each bid for a given product needs to lock the database entry for that product until the bidding transaction has completed. The auction thus turns into a lottery in which the victor is whoever is lucky enough to be the last bidder whose bid has cleared before the moment the auction ends.
Sixth argument: Pent-up demand and network effects as explanations for eBay's success. It can hardly be denied that eBay serves a real need. Information technologies allow intermediaries to expand their geographic scope, and to bring together numerous parties for complex transactions, even over low-valued items. But it does not follow from the flood of activity in this marketplace that the marketplace itself is efficient. In many cases, eBay is competing (very successfully) with market mechanisms (such as newspaper want ads) that are very inefficient: that is, very cumbersome and very limited in the number of buyers and sellers they bring together. As a result, the world is full of pent-up demand for marketplace services. Closets are filled with stamp collections, collectible toys, consumer electronics, and much else whose value in an inefficient market is not remotely worth the transaction costs that would be involved in selling them. With eBay, the contents of many such closets have came pouring onto the market.
From a traditional economic standpoint, this is a good thing. Wealth is being created as goods are moved into the hands of the individual who values them most. But the traditional economic standpoint leaves some things out. Consider the effects of this one-time flood on the market in marketplaces. EBay, as the first online auction house to achieve widespread brand recognition, enjoys considerable network effects: people buy on eBay because eBay has more sellers, and they sell on eBay because it has more buyers. A buyers can switch to a different auction site with relatively little trouble, but a seller's positive feedback points are not transferrable. The network effect, and particularly the first-mover advantage created by the backlog of pent-up demand, thus creates a lasting barrier to entry for would-be competitors, even ones as otherwise powerful as Amazon or Yahoo.
Having established what amounts nearly to a monopoly, eBay can now appropriate the rents that its barrier to entry creates, either by raising prices (or not lowering them as their economies of scale grow) or by imposing restrictive terms (for example by refusing to involve themselves in disputes). It would be hard, of course, to demonstrate in a courtroom that they had done any such thing, since one would have to determine the prices or terms that would exist in a competitive market in marketplaces that doesn't exist but probably couldn't exist. But economic theory points strongly in that direction, and that should at least be cause for concern.
What has been learned? Not that eBay is basically corrupt, or that its benefits are wholly illusory. The sorts of arguments that I have offered, abstract and qualitative as they are, cannot demonstrate anything about the magnitude of the pathologies relative to that of the indisputable benefits. Nonetheless, eBay is a portent of a world in which everyone is connected to everyone else. Such a world can seem like a utopia of decentralization, inasmuch as it appears to be ordered purely through the mutual agreements of individuals instead of the top-down dictates of the state. But appearances can be deceiving.
If everyone is to be connected to everyone else, then someone must provide the switchboard through which all of those connections happen. In this case that someone is eBay. If markets are to be the governing principle of our networked society, then markets will also determine the structure of markets. Markets only work if certain conditions obtain, and in the case of the market in marketplaces it is far from obvious that they do obtain. Indeed, as we have seen, principled arguments can be made to the contrary. Maybe I'm wrong: maybe these effects aren't significant in their magnitude, maybe other effects will arise to counteract them, or maybe the bad effects are simply the downside of the best possible means of realizing gains to trade. But let's not jump to conclusions. If my concerns are well-founded then the networked world is a world of monopolies that hold great power over the relationships between everyone and everyone else. Such a world may be better, relatively speaking, than the world we have now. But perhaps there is a better way still, and if so we should find it soon before it's too late.
end ```
| | | --- | | ProcessTree Network TM For-pay Internet distributed processing. | | Advertising helps support hosting Red Rock Eater Digest @ The Commons. Advertisers are not associated with the list owner. If you have any comments about the advertising, please direct them to the Webmaster @ The Commons. |