Why is it extremely difficult to prove high quality?
This article examines the difficulty in convincing a buyer about the quality of a product even when the product is of genuine high quality. The explanation is underpinned on an extremely important theory in economics, namely ‘The Market for Lemons’ introduced by George Ackerlof in 1970, which sheds light on buyer — seller interactions when the seller knows more about the product than the buyer does (which is the most common case anyway). While the first part of the article will be devoted to explaining this theory, the section that follows will provide some intuition on ‘real world imperfections’ that leave a bit of hope for the genuine seller.
Market for Lemons
To explain his landmark theory which applies to any kind of product, Ackerlof chose the market for used cars. The question at focus here is “Why are the cars that have just left the showroom (only run for few hundred kilometers by the first owner) significantly cheaper in the market than brand new ones?”. If you only count for the ‘wear and tear’ factor and the psychological factor of the car not being ‘perfectly new’, it must be sellable for a price that is close to the price of a brand new one. However, this is far from reality and the above two factors do not satisfactorily explain the huge price differences that can be observed often.
Ackerlof recognized the information asymmetry between the buyer and seller to be the key reason for this price difference. The explanation goes like this. When the car company sells a brand-new car, they know from statistics (data from their previous sales) the probability p of the car being a good one (peach). Accordingly, the car has a chance of 1 — p to be a bad one (lemon). However, it is important to note that the company is unable to tell which cars are the lemons out of the ones they have ready to be sold. All of these cars have gotten the green flag from their quality assurance process. How close p is to 1 depends on how good the quality assurance process is. It is fair to say that p is public knowledge and it is one of the factors that determines the price of a brand-new car.
Say, Bob buys one of these brand-new cars from the company at the price K and rides it for few hundred kilometers. During these rides he’d learn things about the car’s actual quality. Therefore, after riding the car Bob gets a better understanding about the probability of his car being a peach (or a lemon). Let’s name this new probability q. Since q is determined after acquiring more knowledge on the car’s quality (a knowledge which was unavailable to the company’s quality assurance staff) it is a better estimator than p. Therefore, if q > p, then Bob has got a good deal because his car is better than the average car of that type. Accordingly, the car was actually worth more than K when he purchased it. On the other hand, if q < p, then Bob is going to be an unhappy person as he had paid more than the car’s actual value to the company.
Suppose Bob wants to sell his car after running just few hundred kilometers and he has a prospect buyer Alice. What could be the price that Bob is able to convince Alice to pay? We know that Bob has some extra knowledge about his car’s actual quality than the knowledge publicly available about that type of cars. Bob has obtained this extra knowledge by riding the car for few hundred kilometers. Let’s take the case where Bob has figured that q > p. In English, this means that he knows for a fact that his car is of better quality than an average car of that type. However, this is private knowledge of Bob. Due to the way he acquired this knowledge it is impossible for him to convince the truthfulness of that knowledge to anyone else (Bob may be lying to sell the car for a higher price). If we assume that Alice is rational, she knows that the two events (q > p) and (q < p) are equi-probable, while she also understands that Bob has an incentive to convince her that q is greater than p because then he can sell his car to her for a higher price. Therefore, even if it is the case that q is actually greater than p (which is only known to Bob), Alice will never believe that. In fact, there’s no reason for her to believe that. It’s worthwhile to note that Alice is not able to testify Bob’s claims by examining the car because if that is possible, the car company itself would have figured it in first place. Therefore, the best response to be expected from Bob is to agree to a price between K1 and K2 where K1 is the price for a peach (q > p) and K2 is the price for a lemon (q < p).
Let’s put Bob again under the limelight of reasoning. If he knows that his car is a peach, he is losing the advantage he got from the original transaction (buying the car from the car company at a lower price than its actual value) entirely if he agrees to the mediocre price Alice is willing to pay. Therefore, if Bob is rational, he’ll not agree to the transaction with Alice. On the other hand, if Bob knew that his car is a lemon, he’d be really happy to go ahead with the transaction with Alice as it relinquishes him entirely from the disadvantage incurred from the original transaction with the car company.
Remember, we said Alice is rational; so, there’s no reason for us to think that she’d not count on Bob’s choices here. Bob will sell only if he’s got a lemon. Given this knowledge, Alice will not agree to any price which is greater than K2. Now we are creeping into the exciting territory of Game Theory, which is a branch in mathematics that analyzes the rational choices for a player when interacting with other players with misaligned interests. Well, in this case, the only type of car that would really sell is the worst lemon.
According to the theory, no trade can happen! Although this sounds completely alien to the real world where trades do occur, Ackerlof’s theory is now regarded as one of the most important fundamentals of economics. In fact, the Wikipedia page for the theory says that the paper is the most downloaded economic journal paper of all time in RePEC. The theory is widely known as the theory of ‘adverse selection’.
A close look at price-driven markets in any domain would reveal this theory in action. Both the sellers and buyers implicitly agree on the impossibility of proving high quality and thereby expect the price itself to solve the equation for them. Due to this, lemons markets are predominant in almost all domains of trade. However, the theory of adverse selection does not perfectly govern trade as people have invented workarounds to deal with it with varying degrees of success. Two workarounds that have survived the test of time are:
1. Game theory based Signaling
2. Social Trust
Signaling
Given the game theoretic nature of the challenge posed by adverse selection, communities have historically come up with equally strong game theoretic techniques to deal with it in trade; namely signals. Two very prominent signals that have survived the test of time over decades (if not centuries) are brand building and advertising.
Brand building
A company that sells products decides to spend a lot of its resources on socially valuable causes such as charity and event sponsorships. Parallelly they build awareness in the society about them investing on such good deeds. Over the time this helps the company to accumulate enough social capital under its brand name to the level that their buyers consider (not necessarily consciously) that as a serious fact when buying the company’s products. The notion in play here is that the company has so much to lose so that they’ll not risk it by taking shortcuts with their product quality.
To explain this in an abstract form to the more interested reader, let’s suppose the expected profit for the company by cheating on their product quality is X and the damage to the brand name in case it gets exposed is Y. The higher the brand value the higher will be Y. If the probability of getting exposed is p, then a rational company will not attempt to cheat as long as X - pY < 0 (where the expected outcome from cheating is negative, in other words). The higher the brand value of the company, the stronger this condition will hold, giving confidence to the rational buyer to go ahead with the purchase even under the adverse effect of asymmetric information about the quality of the product.
Advertising
One fact needs to be cleared first; what can be recognized as informative advertising (occasional advertisement directed towards conveying information about a product) is excluded from here. Informative advertisements are not really signals — they are rather a mere form of information passing. The type of advertising in interest here are advertisements that are shown repeatedly in media often carrying no meaning. In fact, these are the type of adverts we see mostly today. Advertisements are nothing but attention-grabbing crap. The only message an effective advertisement needs to convey is that a particular company has spent a lot of money on advertising a particular product. How can this act as a strong signal, convincing the rational buyer to go ahead with the purchase?
The company burning a fortune on advertising a certain product signals that the marginal cost of the product (cost of producing an additional unit) is very high so that even an exorbitant expenditure of advertising is not going to make a significant impact on it (if the impact of advertising budget to marginal cost is significant, the company will lose in the long run to its competitor who will offer higher quality products at a lower cost with a smaller advertising budget). Higher marginal cost implies high product quality because high marginal cost with low product quality means the company is incompetent and will lose to the competition eventually. Therefore, a company spending a fortune on advertising a product while surviving the competition gives a rational buyer a signal that the product quality is high and it’s safe to purchase it!
Strange enough — No buyer (even the most rational ones) writes down above-mentioned mathematical formulas when making the purchasing decision based on signals such as brands and adverts, yet these signals serve their purpose. This is in fact the beauty of signals in contrast to information. The later needs explanation while the former doesn’t. Signals are unknown knowns (things we don’t know that we know) that are built into our DNA by the process of millions of years of evolution. We don’t need anyone to teach us them.
Social Trust
Communities from very early stages of civilization recognized that it’s extremely difficult to lead a practical life by limiting interactions between individuals and groups to the ones that always depend on guarantees. Practical social life is full of uncertainties; hence one would be stuck with indefinite indecision if he / she waits for rational justification before every action. Trust is one of the strongest mechanisms (if not the strongest one) communities have developed to bring in some degree of certainty into the cases where the level of uncertainty is either too expensive or impossible to deal with.
Take the previous example of selling a used car. Let’s say Bob knows after running the car that it is a peach. While he is nearly helpless in convincing it to a prospect buyer in a totally trustless setup, his chances can be improved a lot if he already has a friend network that trusts him. One of the friends in that network may believe Bob’s claim and buy the car for a higher price based on the trust he has on Bob. If not a friend in the network can recommend Bob as a trustworthy person when it comes to these types of deals to a friend in his network (who is not a direct friend of Bob), resulting in that third person buying the car for a good price. This is how people with strong networks are usually more competent in trade.
This scenario can extend to a case where the middle person (who recommends the item) is not in the buyer’s network, but an expert or a public figure trusted by the buyer. This is the way product endorsements work. An expert can do a thorough assessment of the product before endorsing it. This is not something the buyer himself / herself is able to do due to the lack of competency. The buyer sorts out the indecision (leading to inaction) that would otherwise result from this incompetency by trusting the expert to do a good, honest job in assessing the product. However, the buyer may not trust the same expert for a recommendation of a different type of a product or the buyer may lose trust on the expert in future if his recommendations seem to be wrong. Therefore, we can contextualize trust as: A trusts B for a task M at time T.
Among other numerous benefits, building a trust-driven culture in an eco-system can significantly reduce the cost of enabling high-quality trade as it reduces the need for more expensive risk management mechanisms such as insurance and hedging. This is one reason, for instance, why Japanese supply chains perform better than those from elsewhere in the world.
Traceability as a disruptor
Traceability (the practice of tracking data in various stages of a supply chain and making that information selectively available for different stakeholders involved) has been traditionally practiced for achieving regulatory compliance and internal tracking. Under the backdrop of information revolution with artificial intelligence, blockchain and IoT set to change the world in unprecedented ways, there can be no better time to think of elevating traceability into the next level as a modern weapon for driving value communication and nurturing trust in supply chains. As a result, it makes a lot of sense to seriously consider traceability as a progressive tool to enable trading high quality products while successfully confronting adverse selection. Above section briefing signals and trust as time-tested mechanisms to deal with adverse selection hints on how traceability will have to be reshaped to achieve this.
1. If traceability can be morphed into a signal from its current form of mere information, it stands up with the potential of becoming a much less expensive signal compared to brand building and advertising. What we are talking here is the possibility of converting traceability into a technological assurance in contrast to the economic assurance given by the other two signals. If done in the right way, this will open up a myriad of new opportunities for small businesses that produce high quality goods but lacks funds for brand building and high advertising to reach their target customer. The challenge for traceability in achieving this would be ensuring that it is nearly impossible for a rogue business to (ab)use it to make its inferior products look like top notch.
2. Moving traceability beyond organizational boundaries to connect all stakeholders in a supply chain with credible and meaningful information will open doors to inculcating a trust-driven culture where companies have better visibility into their supply and more empathy on customer needs and pain points. This will subsequently form strong bonds between supply chain participants, enabling a whole new level of possibilities for trading high-quality products.
This is what we at Tracified are aiming to achieve with our blockchain based traceability platform; to provide as many technological guarantees as possible on information exchanged while facilitating new types of trust relationships to deal with uncertainty. More on how we do this will be discussed in future articles.
