The Market for Lemons and the Signaling Problem

Paul Sulzberger interviews Luigi Muzii.

This interview was originally published on Paul Sulzberger’s blog.

Used car dealerWhat comes to mind when you think about used car salesmen? The words “sleazy”, “disreputable”, “untrustworthy” or “con man” perhaps?

Of course, this assessment may be totally unfair to used car salesmen…

But the market for translation has many similarities with the second-hand car market. Luigi Muzii[1] has frequently made this comparison[2]. Luigi is someone who is never too shy to elaborate on a controversial viewpoint, so I asked him what he had in mind:

Paul: Luigi, do you really think that translators are like second-hand car dealers?

Luigi: No, not at all, Paul! Well… there might be the occasional one! But people who buy and sell translations find themselves in the same sort of economic predicament as people who buy and sell second-hand cars.

Paul: So what sort of predicament is that, Luigi?

Luigi: The sellers know a lot more about the quality of the cars they are selling than the buyers do. In economics this is called “information asymmetry” – and it has a really significant effect on price.

Paul: I can see the parallel with the translation industry. Buyers have huge difficulty in determining the quality of the services they buy – but the sellers, the translators themselves, have a very profound understanding of what determines a good versus a poor translation.

Luigi: Yes, just like the second-hand car market, the translation market is characterised by uncertain product quality – at least from the buyer’s perspective. This problem was studied by the American economist George Akerlof, currently Professor of Economics at the University of California, Berkeley. He won the 2001 Nobel Prize in Economics for his ideas on this problem – the very problem that bedevils the translation industry[3].

George Akerlof
George Akerlof

Paul: So how does the theory work?

Luigi: To make it simple, let’s imagine that there are only two kinds of used cars on the market – good quality ones (let’s call them “peaches”) and “lemons” (cars which frequently break down and require a lot of maintenance)[4].

Now suppose that the owners of lemons are willing to sell them for $1,000. There is a good market for cheap cars, and so let’s suppose that potential buyers are willing to pay up to $1,500 for a lower quality car. Let’s further suppose that the owners of peaches are willing to accept $3,000 for their better quality cars and potential buyers are willing to pay up to $4,000 for one.

Paul: If the quality of different cars on the market was clear and obvious to everyone, then the market would work well, wouldn’t it? Everyone would get the level of quality they expected and were prepared to pay for.

Luigi: Yes. Lemons would sell for a price between $1,000 and $1,500 and peaches would be traded between $3,000 and $4,000.

Paul: But the quality of second-hand cars just isn’t obvious is it? While the sellers know whether their cars are lemons or peaches, the buyers don’t.

Luigi: Correct! All the buyers know is that half the cars are lemons and half the cars are peaches. So there’s a risk of getting a lemon when you really want to buy a peach. Knowing this, buyers would only be willing to pay around the mid-point between the most they would pay for a lemon and the most they would pay for a peach:

1/2 ($1,500 + $4,000) = $2,750

Paul: The sellers of lemons would be delighted to get this price for their defective cars! But this certainly wouldn’t be acceptable to sellers who know that their cars are of good quality and are actually worth a lot more would it?

Luigi: Exactly. Can you imagine what a huge effect this has on the market?

Paul: I guess the sellers of poor quality cars would be encouraged and the sellers of higher quality ones would be discouraged. Soon enough the number of lemon sellers in the market would grow and the number of peach sellers would decrease.

Luigi: Correct. Once buyers understand that the chance of getting a lemon is now more than 50/50 the price begins to fall again. Let’s imagine that after a while, two-thirds of the cars on the market are now lemons and only one-third are peaches. The equation becomes

1/3 ($1,500 + $4,000) = $1,833

Paul: Yes, the average price has dropped! So this is what is called the “race to the bottom”.

Luigi: Yes! Naturally this example is a bit oversimplified, but we can see this mechanism at play in the translation market. The core of the problem is the inherent difficulty buyers have in distinguishing between the value different translation providers offer. When every translation provider in the market attempts to stand out from the crowd by claiming that they deliver “quality”, it’s no wonder that buyers find it hard to tell the difference between them. Promising to do “a good quality job” is hardly a unique selling proposition is it?

Where both peaches and lemons are all labelled as “top quality”, buyers are confronted with a market of uncertain product quality. This is how the translation market looks to buyers. The consequence of this sort of market is that bad-quality tends to drive out better quality as we discussed previously.

Paul: What happens to the sellers of peaches in these circumstances?

Luigi: This sort of market provides an incentive for many translation companies to pass off a low quality product as a higher quality one; decreasing profit margins encourage sellers to look for cheaper, lower quality resources. This just puts more “lemons” on to the market, and so prices continue to spiral downwards making the problem even worse.

Paul: If “information asymmetry” is the problem, would educating the customer help fix the problem, do you think?

Luigi: Just think how many years it has taken you and me to develop an understanding of what translation quality means. Even then, you and I probably have very different opinions on what translation quality is or how we should assess it. What chance has the average translation buyer got to understand the question in any meaningful way?

Educating the customer may sound like a nice idea – but it’s not even a very practical approach. Customers are rarely willing to be instructed by those who are not considered to be their peers – their translators, after all, are only service providers!

Paul: So what is the answer to this dilemma, Luigi?

Luigi: When translation providers in the market signal that they all deliver “top quality”, it’s no wonder that buyers have difficulty in telling them apart. It simply provides them with a strong incentive to choose lower-priced services. The translation industry has a “signalling problem”[5].

Paul: Signalling? You’re talking about the American economist, Michael Spence!

Luigi: Yes. He figured out how we are often able to make good decisions when we are not experts in the subject or have insufficient information. He got the 2001 Nobel Prize for Economics for his ideas.

Andrew Michael Spence
Andrew Michael Spence

The idea Spence developed is called market signalling[6]. He looked at what makes signals credible enough for us to choose between alternatives when we really don’t have enough knowledge or information to understand the difference.

Paul: That’s the situation translation buyers find themselves in!

Luigi: Precisely. If translator A can deliver a better service than a lower-priced competitor B, but a potential buyer is unable to independently check which provider would actually be more advantageous, Michael Spence’s ideas on market signalling provides a way to help us understand on what basis the buyer might choose the better quality service from translator A.

Paul: So how does Spence’s theory work?

Quality as a USPLuigi: Imagine that you want to buy a second-hand car and you find two cars you like equally well—but one has a warranty and the other doesn’t [7]. You’d probably prefer the one with the warranty (and if you have the budget for it, the chances are good that you’d also be willing to pay a bit more for it). If something goes wrong with the car, the seller will fix it free of charge.

Paul: Fixing up “bad” translations by the original translator at no cost is no news in the translation industry.

Luigi: Indeed. But getting translations “fixed” or car repaired can be a real hassle!

Paul: Yes. The very thought of a car breaking down is a nightmare. The inconvenience of not having it while repairs are being done would a big consideration for me when buying a car.

Car Break Down

Luigi: That’s right! You are probably more interested in getting a quality car—one that won’t break down—than you are in having the dealer pay for any repairs. In fact, you’re hoping that there won’t be any repairs!

The meaning behind the offer to fix the car at the seller’s expense goes way beyond just saving money for the buyer. Spence’s insight was that the potential cost of a warranty to the seller is interpreted by the buyer as a signal about the true quality of the car.

Paul: So what’s the reasoning behind the seller’s strategy in offering a warranty then?

Luigi: As most of us don’t have any special expertise as mechanics, it’s difficult for us to independently verify whether the car with a warranty is any better than the car without one.

Paul: Again, this is the same dilemma that most translation buyers face!

Luigi: Yes. If the seller wants to get a better price for a car which is unlikely to give the buyer any trouble, then he needs an effective method of communicating the truth about its actual condition. His message needs to be one that we are going to take seriously.

If the seller knows that the car has had a good track record (or perhaps he has had it checked out mechanically), then he can be confident about offering a warranty because he knows that the chances that we’ll come back to him demanding costly repairs is quite low. If the car is actually in good condition, then offering a warranty is unlikely to cost him anything.

Paul: I get it: the seller’s very offer to pay for any repairs is a signal that he knows that the chances of a breakdown are actually low. So that makes me more inclined to believe him because I know that he wants to increase his profit and he will not want to pay my repair bills if he can possibly avoid it!

Luigi: Exactly. It’s the potential loss to the seller that makes his message credible. He’s now in a position to demand a higher price and we are more likely to pay it because we have some confidence that he’s telling the truth. Our payoff is that we will escape the hassle of a breakdown. It’s a win-win situation.

Luigi: That’s right! You are probably more interested in getting a quality car—one that won’t break down—than you are in having the dealer pay for any repairs. In fact, you’re hoping that there won’t be any repairs!

The meaning behind the offer to fix the car at the seller’s expense goes way beyond just saving money for the buyer. Spence’s insight was that the potential cost of a warranty to the seller is interpreted by the buyer as a signal about the true quality of the car.

Paul: So what’s the reasoning behind the seller’s strategy in offering a warranty then?

Luigi: As most of us don’t have any special expertise as mechanics, it’s difficult for us to independently verify whether the car with a warranty is any better than the car without one.

Paul: Again, this is the same dilemma that most translation buyers face!

Luigi: Yes. If the seller wants to get a better price for a car which is unlikely to give the buyer any trouble, then he needs an effective method of communicating the truth about its actual condition. His message needs to be one that we are going to take seriously.

If the seller knows that the car has had a good track record (or perhaps he has had it checked out mechanically), then he can be confident about offering a warranty because he knows that the chances that we’ll come back to him demanding costly repairs is quite low. If the car is actually in good condition, then offering a warranty is unlikely to cost him anything.

Paul: I get it: the seller’s very offer to pay for any repairs is a signal that he knows that the chances of a breakdown are actually low. So that makes me more inclined to believe him because I know that he wants to increase his profit and he will not want to pay my repair bills if he can possibly avoid it!

Luigi: Exactly. It’s the potential loss to the seller that makes his message credible. He’s now in a position to demand a higher price and we are more likely to pay it because we have some confidence that he’s telling the truth. Our payoff is that we will escape the hassle of a breakdown. It’s a win-win situation.

Paul: Now, what about the lower quality cars on the salesman’s lot?

Luigi: If the seller knows that a car is actually in poor condition, he can be pretty certain that if he offers you a warranty it is going to cost him. In theory, he could push up the asking price by offering a warranty on a poor quality car. But that’s risky—he knows that the worse the condition of the car, the more likely he will make a loss from inevitable and expensive repairs later on. As buyers, we’re not so silly, and we know that too!

To make a profit, his best strategy may be to sell cars of lower quality at a lower price — and without a warranty. After all, there’s still a good market for cheap cars, just as there is a good market for lower value translations!

Paul: I can see that just promising quality is not enough. As buyers, we are unlikely to be persuaded by some sign over the car yard which promises that the seller only offers good quality cars. “Talk is cheap,” as they say. I guess a promise to deliver quality becomes credible only when the seller puts his money where his mouth is.

Luigi: Yes. This explains why the offer of a warranty is much more than a simple “added benefit” to induce the customer to buy. A good warranty is a credible signal of the actual quality of the car. Even though we may not be able to verify the quality of the car ourselves, we figure that a canny car salesman wouldn’t volunteer to pay for potentially expensive repairs unless there was a very low chance of things going wrong.

We can now see the important principle here—one that earned Michael Spence a Nobel prize. Here’s how some other economists have expressed it:

Actions speak louder than words. To be an effective signal, an action should be incapable of being mimicked by a rational liar: it must be unprofitable when truth differs from what you want to convey… [8]

Paul: I’ve got it. The salesman with low-quality cars on his lot simply can’t afford to offer the same sort of warranty as a competitor with good quality cars. It simply doesn’t make economic sense to offer a warranty and then to lie about the quality.

Luigi: Exactly. The seller with high-quality cars can afford to offer a really outstanding warranty because he knows he won’t have to honour it very often. In a situation where the sellers know the truth, but we don’t, we have to read the signals. If a salesman won’t match the warranty of his competitor you can be sure that he’s done his profit and loss calculations and has figured that it is not in his own financial interest to offer it to you. We may not be able to figure out the quality of a second-hand car by looking under the hood, but we can intuitively figure it out by reading the market signals.

When we can see that the seller faces a significant cost or penalty by not telling the truth, then we believe him. That’s the trick!

Paul: So how can we apply this principle to the translation industry? Could good quality translators earn more if they offered a money back guarantee?

Luigi: Ha, ha! Translations are not cars, and the analogy with the second-hand car market can only be taken so far. Offering a simple warranty is not enough. Buyers wanting a “quality” second-hand car might be willing to pay more for one with a warranty because their ultimate payoff is avoiding the inconvenience of a breakdown.

But does the average translation buyer think about “quality” in these terms when she goes out looking for a translator?

Paul: I suspect that most buyers are not really sure about what translation quality actually means—at least not in the same way that translators tend to think about it.

Luigi: Exactly. So why would they want to pay more for some vague notion of “quality”—even with a guarantee? To get a better price for better quality translation work, we not only have to ensure that our signalling is credible, but we also need to be clear about what customer’s payoff is—we have to be able to communicate quality in her terms, not ours…


Notes:

  1. AKA “il barbaro”, Luigi Muzii is a founding member and associate at sQuid. He has been working in the language industry since 1982 as a translator, localizer, technical writer and consultant. He spent 12 years in several departments of a major Italian telecommunications company, and two years in a broadcasting service company. In 2002 he started his own consulting firm to act as an information design and delivery consultant. He was visiting professor of terminology and localization at the LUSPIO University in Rome for almost ten years, and is the author of a book on technical writing, and of many papers and articles. He was one of the founders of the Italian association for terminology (ASSITERM) and of Gruppo L10N, a group of localization professionals volunteering in educational programs.
  2. Muzii L. (2012). A Contrarian’s View on Translation Standards, eBook downloadable from Lulu, page 6.
  3. Akerlof, G. (1970). The Market for “Lemons”: Quality uncertainty and the market mechanism. The Quarterly Journal of Economics. 84,3:3:488-500
  4. The example here is taken from Dixit A.K. & Nalebuff B.J. (2008). The Art of Strategy, W.W. Norton & Company, NY, USA, pp 243-244. See also Rosenthal E.C. (2011). The Complete Idiot’s Guide to Game Theory, Alpha Books, NY, USA, pp 102-105.
  5. See “Signaling Games and The Lemons Problem”, Chapter 24 in Dutta P.K. (1999). Strategies and Games, Theory and Practice, MIT Press, Cambridge, MA, USA
  6. Spence, A. M. (1974). Market Signaling: Informational Transfer in Hiring and Related Screening Processes. Cambridge: Harvard University Press.
  7. The argument here follows the one outlined by Dixit and Nalebuff in Chapter 8 (Interpreting and Manipulating Information) of their book, The Art of Strategy (2008), W.W. Norton & Company, NY.
  8. Dixit A.K. & Nalebuff B.J. (2008). The Art of Strategy, W.W. Norton & Company, NY, USA, p. 239.

Author: Luigi Muzii

Luigi Muzii