What is a "Market for Lemons"?

I have had a number of eBay merchants write me about a post that I made earlier this week suggesting that eBay seems to be displaying the early signs of a Market for Lemons problem (Some would argue it is far beyond the beginning stages, but I will let you debate that with your comments below).

Without focusing on eBay or eCommerce in general, it has been requested that I illustrate a simple example of how a "Market for Lemons" develops.  Obviously, there is a lot of academic research on this subject (I posted a Wharton case study by Dr. Eric Clemons earlier this week, and you won't want to miss it since it discusses eBay's challenges in greater depth.), and almost everybody has heard the "Market for Lemons" term utilzed at some point.  Wikipedia provides a basic overview of the Market for Lemons economic concept, so I won't regurgitate what you can read elsewhere.

In short, the "The Market for Lemons: Quality Uncertainty and the Market Mechanism" is a paper written by George Akerlof in 1970 that describes what happens to markets that suffer from information asymmetry problems.  Ultimately, Dr. Akerlof won the Nobel Prize for Economics in 2001 (along with Michael Spence and Joseph Stiglitz) for his analysis of markets with information asymmetry.  While I was at Wharton, I was fascinated by the implications of this research on eCommerce marketplaces.

The basic premise of the Market for Lemons concept is that in markets where information asymmetry exists, a "Market for Lemons" can ultimately develop in the absence of an effective signaling mechanism.  It is important to note that in eCommerce, this is completely natural and to be expected given the following facts.   Buyers and sellers are typically thousands of miles apart.  Often and especially with small and medium-sized retailers, the buyers and sellers don't know each other (SMBs don't have recognizeable brands) and they haven't previously transacted together.  Although it is easy to review feedback and merchant ratings, it is almost impossible (certainly it is extremely difficult and expensive at the least) for buyers to perform adequate due diligence before the transaction (Is the merchant legitimate, professional, trustworthy and reliable?  Is the merchant financially stable and going to stay in business?  Is the merchant licensed?  Is the merchant selling counterfeit products?  etc...).  The buyer is also at a structural disadvantage in eCommerce transactions because the buyers pays for the item before ever seeing or receiving the real product.  I could go on and on, but you get the point.  The buyer is at an information disadvantage to the merchant when transacting online.  The ecommerce buyer suffers from information asymmetry relative to the seller.

I know some eCommerce marketplaces/channels would like you to believe that their ratings systems solve the information asymmetry problem, but that is obviously not true given where we are today. Please see my post from yesterday for a substantial review of the REAL problem.  Ratings systems are merely a small part of the solution at best, and if requested, I can go into this issue in much greater detail as well in future posts.

I have been researching and analyzing the eBay's information asymmetry problems since my days at Wharton in 1999-2001, so I could probably bore you forever on this eCommerce issue.  Therefore, let's move on to a simple illustration.

Let's start with Figure 1 below and let's assume that we are talking about an electronic marketplace that hasn't yet enabled adequate signaling mechanisms to effectively mitigate the inevitable challenges associated with information asymmetry (What is an adequate signaling mechanism? I will define that for you in detail in a later post, but for now, you should understand that ratings are not adequate without meaningful certifications and strong guarantees... a.k.a. something like buySAFE and its bonding of sellers.). Let's also assume that we are simply looking at the "widget" market. Widgets sell at a low price of $20 and at a high price of $100.  For merchants, if they own widgets that are in perfect, mint condition...widgets of the best quality...their widgets will be worth $100 to a fully informed buyer.  If they owned "fake" widgets, produced by counterfeiters in China, their counterfeit widget will only fetch $20 from fully informed buyers.  In essence, there is a range of widget quality from counterfeit, fakes to perfect, mint-condition widgets, and the fair prices range from $20 to $100.  Obviously, in this example, an average widget would fetch an average price of $60.

As a buyer in this marketplace, you suffer from information asymmetry relative to the seller.  It is extremely difficult, if not impossible, to know whether the online listing you are looking at is for a "perfect" widget or a "counterfeit" widget.  You have no way to know for sure.  All you have is the merchant's description of the product and a few photos that may or may not be of the item you are in fact buying.  Also, when you look at the ratings of the sellers to try and discern what you might be able to expect, everyone looks the same. Hundreds of thousands of sellers have 1,000 ratings or more.  The ratings are all almost perfect.  It is hard to find any negative comments that could provide you context because you have to scroll through page after page of positive comments.  Then you realize that no rational buyer is going to leave a negative rating for the merchant if the merchant can retaliate and hit the buyer with a negative as payback.  You know that the marketplace is very competitive and you know from experience that small businesses go out of business quite often (60+% of small businesses will go out of business within a few short years).  Unfortunately, you have no way of knowing whether the merchants are financially stable or not.  You want to make sure that the merchant is who he says he is, but you can't check that either. Bottom-line, as a buyer, you are flying blind.  You can't tell which widgets/sellers are counterfeit/fraudsters and which are perfect/professional.  This is a major information asymmetry problem.

So, what do you do?  Well, you have to assume that if you pay an average price, you have a 50/50 shot of getting either a counterfeit widget or a perfect widget.  At $60, you will get a great deal 50% of the time, and bad deal 50% of the time.  Those are odds you can live with, so you offer to pay $60.

The problem, as you can see from Figure 2, is that all of the best sellers with great customer service and the perfect widgets will loose money at a $60 sales price.  They can't afford to sell you their perfect widget at $60.  Therefore, they have no choice, but to leave the marketplace in search of a more profitable sales channel or drastically cut their costs by reducing customer service, hiring less qualified employees, doing less training, and buying inventory that is not so pristine.  In short, all the good sellers leave the market because they can't afford to stay.

Now the marketplace has changed quite a bit.  The good widgets and good sellers are gone.  Now, anyone whose widgets are worth $60 or less will still be on the marketplace.  The problem is that buyers still suffer from that nasty problem, information asymmetry.  Therefore, as you can see in Figure 3 below, the buyers have to assume that there is a 50% chance of getting a counterfeit widget and a 50% chance of getting an average widget.  That means that the new average price buyers are willing to pay is now only $40. 

Again, any seller with widgets worth more than $40 is now going to immediately decide to leave the marketplace or continue to drastically cut costs by reducing service and product quality.  See Figure 4 below.

Here we go again.  Buyers now have a 50% chance of getting a counterfeit widget and a 50% chance of getting a less than average widget.  Obviously, the marketplace is deteiorating quickly.  Buyers don't like the quality of the widgets or the customer service, and merchant economics are under huge downward pressure. 

The buyer still can't tell the difference between good products and bad products, and good sellers and bad sellers; and so again, the buyer has to again assume that, on average, the widget is going to be worth something in the middle.  Therefore, the new, new average price falls further to $30.  See Figure 5 below.

This is a vicious cycle.  Prices will go lower and lower, the widgets will get worse and worse, seller economics deteriorate more and more, and the remaining sellers will be less and less optimal.  Without an effective solution to information asymmetry, the marketplace will go into a Death Spiral (Figure 6) and ultimately implode. Obviously, if the price of the average widget falls low enough, no legitimate merchants can sell their authentic widgets and still break-even. They either have to leave the marketplace or go out of business like GlacierBayDVD did earlier this year.  If the price goes low enough, the only sellers that can continue to make a profit on this marketplace are the sellers selling stolen goods, counterfeit merchandise, or no product at all (the fraudster takes the buyer's money and never sends anything at all).

Doesn't this make perfect sense?  Doesn't this explain what you have been observing and experiencing lately?  To me, the situation and the solution are obvious.

Unfortunately, I don't believe the major ecommerce marketplaces/channels have figured this out yet.  Information asymmetry problems take a while to play themselves out.  Older eCommerce channels are clearly showing signs that, in my mind, indicate major information asymmetry problems that will inevitably lead to the Market for Lemons scenario, but they either don't have the will, the expertise, or the ability to adapt.  Others marketplaces/channels are simply too new and immature to see this problem occurring in a big way yet, but mark my words, this scenario will play out time and again with other eCommerce channels as well.  The dynamics are the same, and the more mature marketplaces should be a clear example for new marketplaces/channels to learn from.  Unfortunately, I believe that many eCommerce channels continue to believe that ratings are an effective signaling solution.  I think this is because they all buy into the new Web 2.0 theory of the Internet... the "wisdom of crowds", etc...  Much of this thesis appears to be correct, but certainly not all of it. 

eCommerce channels also continue to believe that back-end protection systems will enable them to avoid the Market for Lemons scenario.  They fail to understand that information asymmetry can only be solved by clear, observable, reliable signals that are hard and costly for bad guys to acquire.   If every seller has back-end protection on his auction listing, then both the good guys and the bad guys have the same back-end protections for buyers.  This is obviously not an effective signaling mechanism that differentiates good guys from bad guys.

Credible signals need to have teeth.  Credible signals need to be meaningful.  When you see a signal, you need to be able to trust it because you know that someone has something to loose if the signal is wrong. Very few eCommerce players have fully addressed this problem with stringent, meaningful seller certifications and strong, comprehensive performance guarantees.  Ratings simply don't cut it.  Bonding sellers certainly does.

The marketplaces also fail to understand that independence leads to credibility.  Moodys and S&P are independent organizations dedicated to providing credible signals about the economic health of their customers, and they are effective because they are independent and their reputations for independence are excellent.  It is going to be tough for marketplaces and payments providers to provide effective signaling mechanisms because they will always be viewed as biased towards one of the parties to the transaction.

Obviously, I believe buySAFE solves this major eCommerce problem by providing the strongest certification of merchants on the Internet today, and then backing up the merchants' transactional performance with a bond guarantee of up to $25,000 per transaction.  We designed the buySAFE business around one objective, to solve the information asymmetry problem for merchants, buyers and eCommerce channels.  We have researched, studied, and analyzed the information economics dynamic on eCommerce marketplaces since the Summer of 2000 at Wharton, and I believe that bonding merchants is the key to avoiding the Market for Lemons problem in eCommerce.  Obviously, I hope so for the sake of all of us that make a living in this space.

Please provide me with your comments.  I am very interested in your thoughts.

****************************************** Please note for future reference that the URL for Steve Woda's Blog has moved to http://www.stevewoda.com

Comments

[Jeffrey Henderson]

This is the problem with unscrupulous people. But Americans are trained to consume at all costs, and that means that some people have to buy fakes just to keep up.

Consumerism is creating the problem, the unscrupuous people are only providing people with what they want, disposable consumerables.

[ed dickson]

Steve -

Extremely interesting and informative article that makes a lot of sense.

This is a great argument why sellers should be bonded and prudent consumers should only buy from trusted sources.

I have to say, I learned a great deal by reading it.

Regards,

Ed

[Manuela Valenti]

Thank you for this valuable information and definitely so true.

I'm seeing it everyday on eBay. I sell in the art category and is incredible the amount of fraudulent sellers that we have to compete with on a daily basis, from the guy in China to the one right around the corner that has a compedium of copies of many different artists.

Prices certainly have gone down and customers and collectors are more skeptical than ever because there is no way to know or prove who is who unless you really know how to spot a fraud or a fake or know who is copying who and selling massproductions.

I know of many sellers who are powersellers on eBay, with profits of $30,000.00 per month and yet they are not honest at all and have taken out of business quite a few reputable and honest sellers, does this service determines their standards based on monthly sales? becuase in my opinion that doesn't make any sense at all, not necessarily the one that sells the most per month is absolutely the honest one. The ones that sell volume in a dishonest manner could have send the honest ones down on their knees.

Thank you in advance.

MV./

[Mark]

Very interesting essay.

Can you post some links to additional published material on this subject. A variety of additional factors come to mind: cost of returning items, cost of investigating seller, cost of delayed purchase, etc.

I'm not convinced 'bonding' can successfully address the asymmetry at Ebay. Why should the buyer believe the 'bonding' agent provides a low cost method of relief from a bad seller and bid more on bonded-vendor auctions? To develop trust, the bonding agent needs to make enough payments that knowledge diffuses throughout the auction buying population. The costs strike me as far more than any potential profit. It only works when the difference between 'bad' and 'good' widgets times the potential volume through the improved channel is large enough to support the bonding agent costs. This situation may exist when the product is gold, fancy new automobiles, and new houses, but for the collectibles exchanged on Ebay, I remain unconvinced.

It seems the market owner, in this case Ebay, is best suited to manage market rules to maintain buyer-seller symmetry. Their costs are far lower than that of the 'bonding' start-up. Ebay should either innovate or copy the 'trust' measures employed in other markets with asymmetry issues (dating sites come to mind).

(btw, I'm reading here due to a negative reaction to my first encounter with BuySafe. When confronted with a BuySafe checkout, I concluded it was a spammer collecting personal information (the requirement that I provide my daytime phone number). This reaction was confirmed when BuySafe emailed me despite my best efforts to keep my email address private. Finding this site and the 'singing dean of Columbia Business School' has improved my opinion, though.)

[Rodney]

These facts are exactly what I am seeing and in growing numbers. If you look at fake items or fraudulent listings where there is a high demand for a product the Ebay site is flourishing with con artists. The feedback system is so seriously flawed and never did work as the originator intended. Not many people leave feedback for even the most serious offences and this is endorsed by Ebay. Listings are down 6%. Power sellers are down 25%. 1 in every 40000 sales results in a paid insurance claim. Ebay no longer publishes their fraud claim totals so it has likely gotten worse than the 40,000. If you add this to your model equation you will note that fees are up as well as profits due to the increase but the exodus as you so brilliantly state is the nail in the coffin eventually. It seems like Ebay will sustain a blind eye approach in the long term. I would personally pick the stock short for the reasons you state. Pick a popular PC, Car, and go to ebay with a search and it is unreal how many people are scamming. Even people that were selling fake software are still allowed to sell products on ebay. Soon all that will be left is Ebay trying to hold onto its worst offenders. As the quality of the sellers goes down so do the buyers.