Say you and I own hot dog stands on the same block. We both sell a moderate amount of hot dogs at a fair price, and each make $20 an hour.
You, tired of competing with such a formidable foe, come up with a scheme to rip off our customers. I reluctantly agree.
The agreement is to stop competing. We both will double the price of our hot dogs, and because customers won’t be able to find hot dogs anywhere else, they will agree to pay that price. In this outcome, we each make $30 an hour.
But here’s the catch: if I cheat on the agreement and you don’t, I’ll be undercutting your price and taking all your customers. I’ll make $40 an hour and you’ll make $0. If you cheat and I don’t, you’ll get the same fortune.
There is no penalty for cheating. What will the outcome be?
The outcome that’s collectively best for both of us would be if neither of us cheated. We both would make $30 an hour, neither of us would be undercut and we both would be better off than we would have been without the agreement.
However, I have an incentive to cheat. If you don’t cheat, I’ll make $40 instead of $30 if I choose to cheat. If you cheat, I’ll make $20 instead of $0 if I choose to cheat. Therefore, regardless of the choice you make, the rational choice I will make is to cheat.
Whether you cheat or not, I should cheat. Whether I cheat or not, you should cheat. The equilibrium ends with the both of us cheating.
Collusion is when firms set up agreements to gouge prices. As the hot dog example shows, collusion does not work when there are no penalties for cheating.
When firms collude, they are called cartels. A famous example of a cartel is OPEC, an international organization of oil-producing countries who together gouge world oil prices to bring in steady incomes for their member nations. The countries all agree to produce a limited quota of oil, so that prices will remain high. When cheating on quotas becomes flagrant, Saudia Arabia, the biggest producer, tries to punish cheaters by over-producing itself, lowering prices and profits for everybody. Even with this threat, member nations continue to cheat and over-produce, because the economic incentive in the short-run is high, and punishment for cheating is rare and unlikely.
Then there are illegal drug cartels, where if gangs cheat, people die. If you and I agreed to be murdered if either of us cheated, maybe our hot dog cartel would’ve been more successful.
I am focusing on cartels here because this is a microeconomics course. But game theory, the study of strategic decision making among self-interested agents, can be applied in a wide range of scenarios. Why do countries tax each other’s imports when it would be beneficial for both countries if neither country taxed? Why do gangs go to war when it would be beneficial for neither gang to fight? Why do candidates spend millions of dollars on attack ads when it would be beneficial for neither party to pay for attack ads? Why do romantically interested partners avoid texting each other a lot when it would be beneficial for both people to text a lot? It’s because parties in all these scenarios are more concerned about their own local interest than they are with the global interest of the group, and there is little mechanism to punish them for this selfishness.
Of course, there are still countries who don’t tax imports too much, gangs who don’t fight each other too much, candidates who don’t attack each other too much (lol) and romantic partners who don’t avoid each other too much. For these scenarios, there may be punishing mechanisms that diminish each party’s selfishness.
If I’m nice to you, and you smack me in the face, will I be nice to you the second time around? Probably not.
That’s what was missing from the first hot dog game we constructed: there was no opportunity for revenge, when there would be in real life. When a game is iterated, it happens multiple times, and firms change their behavior according to information from past rounds.
Say we have a new hot dog collusion agreement. I stick to the agreement, but you cheat. On the first round, yes, you’ll make $40 and I’ll make $0. But on the second round, I will be smarter, and I will cheat too, meaning we will both make $20. On the third, fourth and fifth round, we both would continue cheating, making $20 each round. Through the five iterations, you make $120 (40+20+20+20+20) and I make $80 (0+20+20+20+20).
Now imagine the same game, except this time, we both stick to the agreement through all five rounds. That would mean we would each make $30 per round. That’s $150 total for each of us!
Clearly, in a long term agreement, we could recognize how we would benefit more from cooperating if the other person cooperates too. So if we trust the person to not cheat, we would have an incentive to not cheat ourselves.
If you think about it, the lesson of iterations acting as punishment is applicable to more personal aspects of life. People will generally wrong you for their own benefit if they expect you to be a small part of the rest of their life (since they can avoid punishment then). But they will generally cooperate if they expect you to be a large part of the rest of their life (because they will dread retaliation).
Or they’ll cooperate because they actually value your friendship and respect you as a human being; whichever pretext helps you sleep at night.
…
Okay, sorry for the cynicism. Economic theories tend to assume that people have zero altruistic motivations, that everything they do is self-serving. Obviously, that assumption isn’t always true. Some people genuinely help others without expecting anything in return. I guess it’s harder for economists to make predictions based on those behaviors, so they focus on the alternative. Self-interested people are more common in the business world anyway.
Oh, what was my main point here? Right, cartels are willing to cooperate if cooperation will pay off more than cheating would in the long-run. The only condition is that they trust the other firm to cooperate too.
Imagine a third person had to enter our hot dog cartel. When this happens, cheating becomes more likely. Before, the cartel could collapse if any of two people cheated. Here, the cartel could collapse if any of three people cheat.
The more firms there are, the more likely cheating will occur. This is what makes it so pervasive in OPEC. OPEC started with 5 countries, but now has 12. If any of those members renege on the cartel, how would the other countries retaliate? If someone increases production to punish one rogue member, it would actually be hurting all the other members as well. It naturally ends with every member cheating to maximize their own interest, over the interest of the group.
A reason we value free markets is that it often (not always) allows a large number of firms to compete. This means less collusion and lower prices for consumers. Collusion is mainly an issue when there is an oligopoly, i.e. a small number of firms.
When cartels gouge prices, they are stealing consumer surplus. For this reason, cartels are illegal, and there are several laws in tact that prevent them from forming. For example, if you and I agree to form a hot dog cartel in the United States, and you confess first, then you’ll have immunity from the law, and I’ll get the punishment. I’ll go to jail, you’ll take a chunk of my market share and all you’ll have to give up is my trust and respect. It’s totally worth it, business wise, and it’ll effectively prevent the formation of a cartel in society. I’ll hate you for it, but it would be a fantastic move for your business.
I’ve said enough about games with cartels. Let’s move on to new games.
Say I tell you and 8 other vendors about an idea: let us plant a row of flowers on our block. The flowers will attract people to our particular block, which will lead to them buying our products. Getting the permit to plant those flowers is expensive, and not worth the benefit for any one individual business. But if we spread the costs among ourselves, and share the benefit, then it is worth every penny.
I set up a donation system online, leaving me legally liable for how all your money is used. Whatever money you guys donate will 100% go toward planting the flowers we all know we would benefit from, or it will get refunded.
By the end of the week however, I discover that even though all your businesses would benefit, none of you donated. What happened?
I’ll tell you what happened you selfish pricks: you figured that someone else will pay for it, and that you would enjoy the benefits. Because rational businesses want to get the most out of people, for the least amount of costs, you all gave as little as you could and hoped someone else would pay for the rest. There were no penalties for avoiding payment, so you all didn’t donate, and now we have no flowers.
This public goods game shows that large amounts of people can’t coordinate well in delivering public goods. It’s why we need government to force money out of people in order to build roads, bridges and schools. These are infrastructure that most of us benefit from, but that most of us wouldn’t be willing to pay for if we weren’t forced to do so.
Say I want to hire someone for a super-easy, high-paying job, and the only qualification I seek is a person who is on time, all the time. If I ask a guy on a job interview whether he’s ever tardy, and he says no, should I hire him on the spot just based on that answer?
I’d be tempted to, but of course not. If workers know that I want to hire somebody who is never late, they’re gonna say they’re never late, even if they always are. If I want to do business with this person, he will need to signal his ability of coming on time.
A good way for him to signal this would be to give me his high school transcript, with a cumulative record of his tardiness and absences. If he is unwilling to give it, I will assume that he has something to hide, and he will not get the job.
Of course, this is only a signal. Someone who has zero absences in high school could have 50 absences with me, and someone who has 50 absences in high school could have no absences with me. The transcript does not provide an absolute guarantee. But I rely on it, because even though there is no guarantee, I feel that people who were never late before are less likely to be late in the future. Probabilistic inductive reasoning makes me more trusting of them.
The need to signal to show your abilities makes participating in economic transactions more difficult. Loan applicants need to signal their sense of responsibility by building their credit score. Graduate school applicants need to signal their intellectual curiosity by acing their classes. Web developers need to signal their programming skills by explaining projects they’ve worked on. Vendors need to signal the value of their products by touting the reputation of their brand. If they can’t signal the desired attributes to their recipient agents, their economic opportunities will diminish.
Investing in signaling can be expensive and the need for it can exclude people from the economy. A smart, hardworking kid who can do well in college but can’t afford the tuition probably will not have an opportunity to signal to his employers his ability to work and learn skills required in a job, especially compared to job candidates who did go to college. If there was no information asymmetry, if people could read each other’s minds and assess each other’s habits and intentions, quickly and accurately, transactions would be smoother and we wouldn’t have to invest in so much signaling. But that of course is a pipe dream. The education system, whose economic purpose is to provide signals about employees to employers, is here to stay. Do well in school kids; earn that scholarship!
Game theory at its core is about coming up with thought experiments to determine a general trend. It gives us an idea about what decisions rational, self-interested economic actors will tend toward in reality, given the constraints of information asymmetry and other people’s encroaching interests. It can often help people understand economic phenomena that can’t be described using simple supply and demand.
Remember that this site’s main purpose is to give you an intuitive feel for subjects. If you took an actual economics course before, you would know that the explanations there are a bit more mathematically thorough and intensive (especially for game theory), and while that might be appealing to some nerds, I feel it deprives math-phobic students from the beauty of Economics. If you’ve read through the three economic courses so far, you probably have a good idea about how important the subject is to everyday life. It shouldn’t just be for geeks like me.