Hey all -
We’ve finally hit the escape velocity of Planet Complexity. Moving on…
* A market for lemons
Let’s say you’re buying a used car and you head to your local car dealership. You estimate that the cars in the lot are between $10,000 and $20,000 each, but you can’t be sure which is which.
Because you may get a “peach” ($20,000 car) - but also a “lemon” ($10,000) - you figure a safe offer for a car is the average: $15,000.
But, unlike you, the dealer knows what each car is worth - and the dealer would never sell you a $20,000 car for only $15,000. So you can be quite sure that if you offer $15,000, you will only get a car whose value is between $10,000 and $15,000.
Now would you still offer $15,000?
Nope - you’d revise your estimate downward. And in response, the dealer would propose a car of even lower value.
And so the cycle continues until the only cars which trade are lemons for exactly $10,000. Hence, “a market for lemons.”
What’s the root cause here? Information asymmetry.
When you cannot assess the quality of something - but “asymmetrically,” the other side can, they have an incentive to take advantage of your ignorance. And you - knowing that you are ignorant - decide that you cannot possibly get a fair deal. Information asymmetry distorts prices and discourages trading.
This is more pervasive than it seems.
Posting a job offer with a stated salary means the people who earn more than that won’t apply (“peaches”), and only the people who earn less than that will (“lemons”).
Going to meetups where you can’t assess the quality of attendees means that the “best” people - however you define it - cannot be sure there are other people of the same caliber, thereby driving out the “best” people from the pool of attendees.
Going on a date when you aren’t really sure what the other person is like, both in appearance and personality - say with Tinder - means you can’t tell if it’s worth your time, and ultimately renders you less likely to go on the date at all.
Buying clothes from a random shop on the street places the onus on you to assess the quality of clothing, while buying from established brands such as Nordstrom or Zara means you can largely assume their merchandise has passed the test of time and earned the approval of millions of customers.
So how do you resolve a market for lemons?
The fundamental problem in the market for lemons is that one party cannot evaluate the quality of something while the other party can. The solution - logically enough - is to give more information to the party with less information so that they can better assess the quality of that something.
Indeed, that transfer of information about something’s quality is called a signal. More specifically, signals are the observable pieces of information you get about something, which allow you to infer the quality of the unobservable parts.
For example, if you see a well-dressed salesperson offering a polished car from the lot of a prestigious dealership - all for a high price - you now have several observable signals that the car is high-quality. You hope that the unobservable parts - the quality of the engine, brakes, steering and so on - are as good as the observable signals.
In other words, we tend to think that the observable signals are indicative of the unobservable quality. If I show you a nice-looking car, you’ll probably think that it’s a nice-overall car.
If it is actually a good car - that is, the observable signals were honestly indicative of the unobservable traits - we say that the signal was “credible.” If the car is a lemon - despite looking like a good car - we say that the signals were “dishonest” or “non-credible.”
Returning to the market for lemons above: if I am selling you a used car, the least I can do is offer you some observable signals about its quality. Maybe I demonstrate the steering in front of you, or have its quality certified by an independent mechanic, or let you test-drive it for a day. By receiving more observable signals, you get a better sense of whether the car is actually good or not.
The most I can do is give you “credible signals.” If I offer you a guarantee or a warranty, then you are practically assured that the car will be good quality, otherwise you return it for the price you paid. Obviously this is costly, but credible signals are always costly.
Now we’ve come full-circle. Markets for lemons arise when we have too little information about something’s quality, and this discourages us from transacting. Since we need more information, we evaluate observable signals to infer unobservable quality. And the more costly the signal is, the more “credible” it becomes (because cheaters can’t cheaply copy it).
And so we see job offers without posted salaries, or panels of “experts” in meetups, or required photos on dating services, or the importance of brand. These are our observable signals.
Thanks for reading,