7.3 Why should we follow the rules?

Social contract theory is not free from difficulties. Of the two major problems facing it, the first is a lack of clarity concerning the status of the agreement that creates moral rules. Is this supposed to have been a real agreement between real people at some point in the past, or a hypothetical agreement about what people would agree to if they were faced with the task of creating moral rules from scratch. The second problem with has to do with its adequacy as a justification for moral rules in general. Granted that it shows we have a need for such rules, it fails to provide an adequate account of why we should bother to follow the rules when we have constant incentives to cheat.

What agreement?

This first problem facing Social Contract Theory has to do with its claim that the authority of the rules of the social game can only be based on the free and willing consent of those under the authority of these rules. If social rules are conventional rules, they can only be binding to those who have freely accepted the restrictions they impose on us. But most of us, who never lived in any situation remotely resembling Hobbes’ state of nature, probably cannot remember ever being asked to endorse a social contract. Instead, we are born into a society with a set of rules already established and taken for granted as legitimate rules and we are asked simply to accept these rules. But then how could these rules really be binding on us if we were never asked whether or not they fit in with our ideas of what would serve our interests best? There are two standard ways of responding to this question.

First we might say that although the social contract was in fact a real historical event establishing a set of conventional moral rules, once these rules have been set in place, they are no longer subject to debate or rejection. Our ancestors who lived in a state of nature were in a unique position to found a social order and, short of the kind of complete social breakdown that occurs during a civil war or political revolution, subsequent generations have no choice but to accept the rules authorized by the original parties to the social contract. Perhaps subsequent generations can tweak particular details of the rules of the social game, as has definitely happened, for example, concerning relations between men and women in the western world. But the basic rules that result in the creation of the social world in the first place must henceforth be taken as given.

This response is, however, unsatisfactory and even threatens to undermine the very thing that seems unique about Social Contract Theory, its claim that social rules are inherently conventional and based on the free choices of those to whom they apply. That is, if only the original parties to the social contract are in a position to accept or reject some set of social rules, then these rules can count as conventional only for them and not for anyone born into a later generation. But this means that the social rules are just as arbitrary for later generations as they would be if given by a particular culture, by God, or by some conception of human nature.

The second way to answer the question of why moral rules might be binding on us even if we have never explicitly approved of them starts by rejecting the idea that the social contract should be understood as a real historical event. Instead, the contract should be understood as a hypothetical contract reflecting what rules would be acceptable to free and rational agents if they were living in a pre-social state of nature. That is, the Social Contract Theory is to be recast as an idealized set of rules that any free and rational individuals would be logically compelled to accept in order to avoid the general types of problems that would plague their lives in the absence of such rules. Just like physicists idealize in their accounts of the laws of motion by talking about things like frictionless planes and perfectly elastic collisions, us philosophers are licensed to talk about the requirements of free and rational agents in general, regardless of the historical details of their lives since what we are after is an account of the sorts of rules we should accept. This approach is by and large the approach taken by the majority of contemporary philosophers who take Social Contract Theory seriously as a way of justifying social rules philosophically, and it does seem like a reasonable way to proceed.

Thus let us accept that Social Contract Theory is not dependent on the claim that a real agreement is at the basis of currently existing social rules. Instead let us accept for the sake of argument that the social contract is a hypothetical device that enables us to talk about what rules would be acceptable to free and rational agents whoever they happen to be. We can now ask a more difficult question of this theory – why would free and rational agents accept any rules at all that limit their options in the way moral rules do? Moral rules, we recall, should be understood as rules that might get in the way of self-interest. For example, if there is a moral rule against lying to others, this means that whether it suits us at the moment to lie is not important because the rule against lying should overrule our immediate interests. We have seen that Social Contract Theory is based on the idea that we can all agree that certain sorts of behavior should be restricted since they tend to lead to chaos if enough people engage in them. Thus it certainly seems like moral rules have an important role to play in our social lives.

The prisoners dilemma

We may wonder, however, whether Social Contract Theory goes far enough by pointing out how moral rules can serve our collective interests. What is to prevent us from accepting a set of rules as long as it does not get in the way of our individual interests but then ignoring the rules whenever it seems to us that it pays to do so? If moral rules are necessitated by the fact that we are essentially self-interested individuals but also need not appeal to anything besides self-interest, doesn’t this make moral rules highly unstable?

To see why this is the case, let us consider a famous puzzle often called called “The Prisoner’s Dilemma.” This puzzle condenses into a very clear picture the fundamental problem of morality – why should we trust each other when we are all constantly facing temptations to violate trust and why should we stick to the agreements we make when a situation presents itself where it is in our interest to violate them?

Imagine that you and a partner in crime have just been arrested after a botched attempt to rob a convenience store. The police have reason to suspect that you two have also successfully robbed other stores in the area but lack sufficient evidence to convict you of these other robberies. Since the police chief has been taking a philosophy course he decides to present each of you the following offer in separate interrogation rooms:

We have reason to believe that you and your partner have been involved in a string of robberies around here and we would like to put at least one of you in jail for a long time. So if you cooperate with us by testifying against your partner, we are willing to make you a deal. If you testify against your partner and she stays quiet, you will go free and she will get a 10 year jail sentence. If both of you testify against each other then we will give you each 5 years in jail. On the other hand if both of you refuse to cooperate with us and keep your mouths shut, we can only hold you for 1 year for the attempted robbery yesterday. What is your decision?

Now suppose that you and your partner have already come to an agreement not to rat on each other in case the police try to convince you to do so with an offer like this. What should you do here? That is, what is the smartest thing to do in this case assuming that neither you nor your partner really wants to go to jail and will always opt for a shorter sentence? Is there even a rational solution to this puzzle? Is it better for you or your partner to stick to the agreement not to rat, or is it better to rat? (Note that there are no hidden negative payoffs here, such as, for example, the threat that your partner will take revenge on you later on if you rat since that would distort the puzzle the police have presented to you. Besides, we are assuming that if you testify against your partner, the police will reward you with a new identity through their witness protection program.)

The problem here, is that you cannot be sure about what your partner will do, since she is being held in a separate interrogation room. All you have is her word that she will not rat on you, and all she has is your word that you will not rat on her. The way to find a solution to this puzzle is to recognize that there are really two possibilities for each of you and thus four possible combinations. Either you partner will keep quiet or your partner will rat on you, and likewise for you. We can arrange these possibilities in a “payoff matrix” like so:

I keep quiet I rat
you keep quiet 1 year each I go free,
you get 5 years
you rat I get 5 years,
you go free
3 years each

So if your partner keeps quiet, is it better for you to rat or to keep quiet? Well since you would get less time in jail if you ratted (no jail time versus 1 year for sticking to the agreement not to rat) than if you kept quiet, if your partner keeps quiet then you should rat. But what if your partner rats – after all, you do not know what she will do – which is better, ratting or keeping quiet? Well here again ratting gets you less time in jail since if you keep quiet while she rats you end up with 10 years in jail, while if you both rat, you’ll only end up with 5 years behind bars. This is known in the jargon of “game theory” which studies strategic situations like this, as a dominant strategy, since no matter what your partner does it is always better for you to rat. Unfortunately, since your partner is thinking in exactly the same way as you are thinking, if both of you choose what clearly seems to be the best thing to do, you will both rat on each other and will end up with 5 years in jail each. Why couldn’t you have just kept your mouths shut and gotten only one year in jail each? It would be nice if you could just trust your partner not to violate the agreement you have made, and this seems like it gets both of you the better deal. It’s too bad that trust seems to be irrational here since it is always better to rat no matter what the other person does.

Much has been written about the prisoner’s dilemma as a model for certain kinds of basic social interactions. For our purposes here, I’d like to mention only two points. First, this kind of situation arises quite often in the real world, namely whenever we may make an agreement with each other to play by the rules but are constantly tempted to cheat for an immediate gain, in spite of the fact that cheating leads to an outcome that is worse if enough people do it.

Think, for example, about the problem of over-fishing. All fishermen recognize that over-fishing depletes fish stocks and thus threatens their long term livelihood. So suppose the fishermen all get together and agree that it is in their best long term collective interests to limit their own catch. The question for each fisherman, as he sets out his nets, “Should I stick to the agreement and not over-fish?” He reasons like so: Well, I have no idea if the other fishermen will stick to our agreement. Suppose they do – in that case my over-fishing won’t hurt anyone else’s long term livelihoods, but I will benefit by having more fish to sell. On the other hand, supposed they also over-fish – in that case I would be a fool not to also over-fish since my short term benefits would be better if I did and there won’t be any fishing for anyone after a few years of over-fishing, so my not over-fishing does nothing to protect my long term interests. So everyone, reasoning in a similar fashion over-fishes and depletes the resource that all need to survive. Why can’t they all just stick to their agreement?

This last example is an example of a “many person prisoner’s dilemma,” also known as a “free-rider problem” or the “tragedy of the commons.” With a little reflection you will probably be able to come up with numerous other situations that exhibit the same logic. The prisoner’s dilemma has thus earned the reputation of being a simple model illustrating a very real problem faced by social actors trying to coordinate their behavior in a way that secures their own long term interests, which is exactly what Social Contract Theory says moral rules are supposed to do.

This brings me to the second point I want to make about the prisoner’s dilemma, namely that it shows the ultimate weakness of Social Contract Theory as a justification for moral rules. Moral rules are supposed to be solutions to situations like prisoner’s dilemmas, since in the state of nature we all act for the sake of our immediate short term gain and thus cannot coordinate our behavior in a way that leads to peace and security. But moral rules alone, in the form of voluntary agreements not to break the rules that you yourself have agreed to, are not enough to get out of prisoner’s dilemmas. As we saw a moment ago, a promise to follow a rule that we ourselves can agree is a good idea is worthless given the ever present temptation to cheat in order to get a better outcome.