Bad Repeated Dealings
The discipline of repeated dealings doesn't always make life better
You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. You can support our newsletter by signing up here:
It’s a sunny Saturday afternoon. You decide to go to the dealership in hopes of buying a used car. (Let’s assume it’s 2019 and you could actually find the car you wanted, not the mess we are in now.) You find the exact model you’re looking for at $2k less than you expected. Great! $2k of extra consumer surplus! That’s what I’d say at least. Normal people would say “I saved $2k!”
But wait. Why is it so cheap? What’s wrong with it? The dealer says everything is in great condition but you (like me) are not a car expert. You get cold feet and decide to drive home without the car, worried there is a problem you don’t see. To paraphrase Marx (Groucho, that is), “I don’t want to buy any car that they are willing to sell me.”
In economics, this is called the lemons problem and it is one of the problems we may worry about in markets. People may have an incentive to cheat (broadly defined here as not being truthful about all the problems with the car). That’s a bad thing. People often point to these sorts of “market failures” as requiring governments to step in and help out. For example, since supposedly only sick people will want to buy health insurance, health insurance won’t work as a business and the government should step in to subsidize or run the industry.
The common counterargument to the market for lemons story is that, yes, it’s good in theory but it fails in practice. People buy and sell used cars all the time. Any market failure is a profit opportunity for those who have a partial solution. Carfax comes along and gives people more data on the car.
One common solution is for dealerships to develop a brand or a reputation as honest dealers. This is one of the margins on which sellers compete.
The basic mechanism is what economists call the discipline of repeated dealings. It is true that for every buyer that walks into the dealership, the dealership has an incentive to lie about the quality of the car. But the dealership is trying to maximize the value of all future sales too. If they sell you a piece of crap, you post that on Google Reviews and that hurts their future sales since they will lose part of their reputation and therefore future profits. Brand name recognition is a type of repeated dealings between one seller and many buyers.
Th easiest way to see the power of repeated dealings is in the prisoner’s dilemma. The prisoner’s dilemma is an especially thorny problem because no matter what the other person does, each person wants to take the “bad” action. There’s no room for “you play nice and I’ll play nice” cooperation in the single interaction game.
For example, suppose you and I are fishing in a small lake. Any fish I catch lowers the number of fish available for you and vice versa. Even if we come together and say “let’s not fish too much,” each of us has an incentive to cheat and fish more than we say we will.
However, if you and I keep running into each other on the lake for many years, we can overcome this prisoner’s dilemma. For an extreme version, we both agree if we find the other one fishing too much, the other party will pollute the lake and kill all the fish. Knowing that if I cheat today and get caught, I will lose all future fish, I have an incentive to agree to the terms and not fish too much. (Let’s ignore the complication of whether you will go through with your promise to kill all the fish. Maybe the blockchain can make it happen automatically?)
People who are in favor of markets (myself included) often point to this mechanism as to why markets function well without regulation, despite the possibility of cheating or bad actors. People have an incentive to invest in a reputation that disciplines the market, even if imperfectly.
This story should be standard for people who have taken any game theory or even micro 101.
Repeated Dealings Can Make Things Worse
While the benevolent repeated dealings story above is helpful for understanding many interactions, like investments in brand recognition, it is not a Jedi mind-trick that shows that markets always work or that good things always happen.
To see this, consider the other time I teach the prisoner’s dilemma in 101: collusion. Now you and I are competing newsletter writers. Let’s imagine for a second that there was another equally great econ newsletter out there somewhere. Our newsletters are so close that most readers will just pick whichever is cheaper.
We would both like to raise prices to increase profits but worry about being undercut by the other party. Anticipating the other party is going to cheat our collusive agreement, each party cheats, and collusion breaks down. This is a standard argument for why collusion is hard since each party has an incentive to break the agreement and win all the market. Here the prisoner’s dilemma works against us as newsletter writers but increases competition and lowers prices for readers. This is a good (from a total welfare perspective) prisoner’s dilemma.
I hope we can see where this is going. Just as repeated dealings changed the fishing outcome from bad to good, it can change the collusion outcome from good to bad. Repeated dealings help sustain collusion! In both cases, repeated dealings open up the range of possibilities between the two repeated dealers. Instead of simply cheating, the repeated players can cooperate.
Take another example from outside of the market context. Suppose there are two political parties run by different subsets of elites. I am sympathetic to the idea that, if there are democratic institutions, the political parties will fight for political support from the non-elites and compete toward an efficient policy. This is a prominent idea within the Chicago School of Political Economy, such as Gary Becker, Sam Peltzmann, and George Stigler. Josh, Alex Salter, and I have a forthcoming paper in Public Choice on exactly this mechanism. (You can read the ungated version here.)
Unfortunately, this force towards efficiency can break down if there is a possibility for a sort of back and forth of power between the political parties. We can imagine that there is implicit collusion not to go too hard for the non-elites support. Never go full populist. By not “competing” too hard, they jointly protect the self-interest of both political parties over the longer term, even if in this particular election cycle one party is out of power. This could lead to an entrenched set of policies that favor elites.
The more we look for the downside of repeated dealings, the more we find it. Brands can be bad too. Yes, sometimes a firm has an incentive to establish a reputation as an honest seller. But people also have an incentive to establish a reputation as tough negotiators. They may bargain particularly hard in order to convince later buyers that the price will not be dropped at all. Gangs have an incentive to establish a reputation as violent, even if they would prefer not to be violent in a particular altercation.
The important takeaway is that reputations, brands, or repeated dealings are not a Jedi mind trick that markets always work. As always, the question is empirical and we need to look on a case-by-case basis. But economic theory can help us navigate what exactly is going on in each case.