23. Morals by Agreement: Constrained Maximization

Note: This post is the twenty-third in a series. Click here for the full listing of the series.

This is the fifth of what should be a few posts on the book Morals by Agreement, by David Gauthier.

---



---

In the Seinfeld clip above, Jerry Seinfeld arrives at the car rental agency only to find that although they have his reservation they don't have a car for him. He then explains that they don't understand the concept of a reservation - anyone can take a reservation, the trick is holding the reservation.

Similarly, in David Gauthier's philosophy based on people who maximize their self-interest, it's easy enough to explain why people might make agreements to co-operate (as explained in chapter 5) but the hard part is to explain why people would hold to their agreements. This is the task of chapter 6 of Morals by Agreement, where Gauthier attempts to explain why people who are self-interest maximizers would keep to their agreements when breaking those agreements might be more beneficial to them than keeping them would be.

Gauthier notes that in Hobbes' book, Leviathan, Hobbes creates a character known as 'the Foole' who argues the case for breaking your agreements where this is beneficial to you. Hobbes' solution is the creation of the sovereign. A government that will ensure that it is in your interest to keep your agreements by punishing you harshly if you don't. He also points out that Hobbes' solution dispenses with the need for morality (understood as a constraint on pursuit of self-interest) by simply re-aligning self-interest with the good of society via the sovereign's presence.

Gauthiers introduces the terminology 'Straightforward Maximizer' to be someone who, like the Foole, does what is in their best interest even if they have previously agreed not to and 'Constrained Maximizer' to indicate someone who foregoes their self-interest in situations where a fair co-operative solution will yield them a benefit over what they would get if everyone pursued their self-interest.

The Constrained Maximizer will cooperate if he expects others to do so, but if he suspects that the people he is dealing with are straightforward maximizers, he will do the same to avoid being taken advantage of.

In the terms of the Prisoner's Dilemma, the Constrained Maximizer will not confess if he thinks that the other Prisoner is also a Constrained Maximizer, but if he thinks the other prisoner is a straightforward maximizer, he will confess to avoid being taken advantage of.

Gauthier explains that the Constrained Maximizers will achieve better outcomes than the Straightforward Maximizers in the case where everybody knows which type of person everyone else is (i.e. if Constrained Maximizers had to wear 'CM' sweatshirts and Straightforward Maximizers had to wear 'SM' sweatshirts.) This is only logical - the Constrained Maximizer changes his strategy based on who he is dealing with whereas the Straightforward Maximizer doesn't, so naturally it is in the interests of the Constrained Maximizer to want everybody to be easily identified in one role or another.

Sadly, of course, in real life it is not always obvious who you can trust.

This (I think) is why Gauthier later says that,

"Failing to exclude straightforward maximizers from the benefits realized by co-operative arrangements does not , and cannot, enable them to share in the long-run benefits to co-operation; instead it ensures that there are no benefits to share."

In other words unless you kick out the straightforward maximizers, co-operation will fail.

Gauthier argues that Constrained Maximization is not simply another way to pursue maximization of your self-interest given particular circumstances, but rather requires that the Constrained Maximizer choose a disposition towards not breaking their agreements. The reason that people must choose this disposition rather than deciding on a case by case basis is that people who decide on a case by case basis will be kicked out of society so they won't get the benefits that co-operation brings.

Finally, Gauthier points out that over time, a community where all are constrained maximizers will be expected to outperform a community where all our straightforward maximizers.

---

To be honest, I didn't find this chapter, which is the heart of Gauthier's theory, very convincing. A few comments:

1) Back in chapter 3, Gauthier argued that prudence - taking into account my future interests - was no part of rationality. But now it seems that choosing the disposition towards constrained maximization rests on the sort of prudential concern that was dismissed as irrational in chapter 3 (unless I am confused). This is OK by me, I never really agreed with the notion in chapter 3 that it was rational to be imprudent, but it doesn't seem consistent.

2) It's not clear how the Straightforward Maximizers are going to be excluded. Who will exclude them and what incentive do they have to undertake the substantial costs of doing so?

3) Although the Constrained Maximizer is disposed to switch to Straightforward Maximization when confronted with another Straightforward Maximizer, it seems likely that the person who Straightforwardedly Maximizes full-time is likely to be better at it. For example, I buy a piece of artwork from a man who, after taking my cash, pulls a gun and demands that I give back the art that I bought. I might choose to repay this act of Straightforward Maximization by responding with violence myself, but if the man is a professional crook, it seems plausible that he is better with a gun than I am. And if I myself develop a great skill at Straightforward Maximization in order to successfully battle these lifelong criminals, then it seems possible that I might reach a point where it is in my benefit to become a crook myself.

4) Most fundamentally, this notion of people 'choosing a disposition' doesn't square with my intuition of how these sorts of decisions are made. I'm reminded of the joke about, I believe, Abraham Lincoln who was known for his honesty and incorruptibility.

As the joke goes, a businessman walks into his office, offers Lincoln $100 to change a piece of legislation and Lincoln smiles and says no.

The next day, the man comes back and offers $1000. Again Lincoln smiles and declines the offer.

On the third day, the businessman comes back and offers $10,000. This time Lincoln gets angry and throws the man out of his office. The man asks, 'why was your reaction so different this time' and Lincoln replies, 'you were getting close to my price.'

In real life, people don't simply adopt a disposition to keep agreements and then dispense with any thoughts of breaking them. People are constantly weighing whether or not it makes sense to keep an agreement.

You can see this clearly in the different levels of security that attend situations with different rewards to agreement breaking. A dollar store has little in the way of security other than keeping more expensive items behind the counter. A jewelry store however, where the small size and high value and resaleability of the merchandise provides greater incentives for agreement breaking (theft), tends to have much higher security.

It seems more probable to me that people incorporate morals into their decisions, not by adopting a disposition based on long run maximization of their interests, but rather by applying utilities to the morals themselves, where actions that involve violating one of our morals attract a negative utility that offsets the potential gain from the action. This would explain why we are so concerned with the words used to characterize a situation: "it wasn't really a lie" we say or "technically it wasn't stealing", statements which don't make a lot of sense under Gauthier's framework. But I'm getting beyond the scope of the current discussion.

It seems likely that straightforward maximizers are not excluded from constrained maximization society based on society's perception of their disposition but rather in response to specific acts which demonstrate that disposition at a level beyond what society is willing to tolerate. Therefore, to the extent that straightforward maximizers could carefully choose their transgressions so as not to get caught or to the extent that they could acquire enough power that society was unable to exclude them (think organized crime) they might be able to avoid being excluded from society regardless of their disposition.