I’m grateful to Stefano for his critique of my proposed social welfare theorem. I’ve replied to him several times via e-mails to clarify. One of those replies went to Walter also, but not the second one, I believe. This exchange went off in several directions. I’ll summarize where I think the matter stands, from my point of view while accommodating Stefano’s points. I am drawing from the e-mails I sent to Stefano yesterday before Walter’s latest publication of Stefano’s e-mail to the two of us.
Stefano argued that demonstrated preference for a group doesn’t exist. Walter disagrees with that.
I agree that the group has no material, molecular, atomic, sub-atomic, light, wavelike, or wave existence.
I agree that individuals within the group are the only ones that can act, have preferences, choose, etc.
Individuals create and make up groups, societies, organizations, orchestras, armies, bureaus, etc. It is common to use these and other such names as subjects. The band plays a song. The army attacks. The cavalry arrived. Congress passed a law. It’s understood that there is no material existence of the subject apart from that within the organization, people, equipment, instruments.
The group also has organized itself to accomplish certain things such as playing a C-major chord at a certain time and place.
But if the people in the group act in concert, I do not see what’s wrong with saying the group has a preference ordering, which means that the members have created a group preference ordering and have acted upon it together, albeit each one individually. If we say “the army attacked”, we know what we mean, that a body of soldiers each acted and together made a coordinated attack. This means that I agree with Walter that the group exhibits what may be called a “demonstrated preference” through the group action its members produce. That concept is not essential; it’s a generalization of terms used by Rothbard. We can exclude it and just say that the group has preference orderings and acts in coordinated ways to select them. The army attacks a given fort at a given time in a given way. The band plays a certain song in a certain way.
I agree that society does not act apart from those in it, and neither does an army or any other organized group that has taken pains to assure group behavior to achieve a purpose.
There is, in fact, contracting going on within the group, and there is a resulting agreement, shared purpose, and then actions carried out to achieve that purpose. The organizing of human actions can get very complex. Producing an aircraft carrier or a supersonic missile takes a lot of organizing and doing.
You play a C on your clarinet and I play an E on my trumpet, and we create a sound we want. What remains is to say how this relates to social utility and what happens when Walter plays an F when he agreed to play a G.
The shown preference (or exhibited or demonstrated preference) is the coordinated action, the attack, or perhaps the law that’s adopted and enforced, like a rule saying “Thou shalt not steal”, applicable to everyone in the group.
To say that action is always individual or particular is correct as far as it goes, but we need to allow for coordinated actions because they are pervasive. In doing so, we do not need to resort to special additional theories, like mob psychology, to grasp that a group may express a preference or construct a structure of obligations in which roles are defined.
My theorem does assume a shared rule and it does assume it’s constant or being held when the judge allows a thief to get away with his theft. Otherwise the theorem makes no sense. I of course assumed ceteris paribus.
Within the organization, there is not unanimity, even though it’s organized to achieve a common purpose. Some soldiers and officers may disagree with the attack, and some musicians may think the song should be played faster or slower. Nonetheless, the group goes ahead with its actions. The extent to which a particular goal is achieved is variable, because individuals deviate from their assigned actions. Unanimity is not essential to the theorem. I brought that up only to suggest that if there was a goal that the organization’s leaders had adopted and that guided its members in their actions, this was bypassing unanimity or proxying for unanimity. That is, the group was acting almost as if it were united in its choice. Workers had accepted pay to do their part. Soldiers had joined up expecting to be ordered around. The leaders had gotten their position expecting to be asked to formulate plans and motivate workers, distribute rewards, etc. We do not need to assume unanimity in the statement of the theorem.
At no point did I invoke a geist or spirit of the organization as an actor. There are only human beings doing the organizing and actions.
All of the preceding explains an organization, of which a society is one. The proposal I made is that we as observers need not devise measures of social welfare or the utility of a group to know in some cases whether utility has been lowered by some action in the group.
The matter to my mind is simple.
1. A society has a rule “thou shalt not steal”. It adopted it by assumption because it was a preferred rule.
2. It applies to everyone in the society and it’s meant to apply to everyone.
3. The rule expresses a preference of the group and is not in place or will not be ever adopted without raising social utility. This does not mean there is a being called a group. It means the group has created an actionable agenda applicable to all within the group, and that the members are committed to it (common purpose).
4. Someone breaks the rule and a judge lets the person get away with it.
5. This breaks the achievement of the aim, in all its plenitude or the full achievement. (It’s like one platoon fails to attack, or the violins fail to play the right notes.)
6. The outcome is a less desired state of affairs for the group, or in other words lowers the group’s utility.
The place where I think the problem arises is #6, in the words “less desired state for the group”.
To me, the group has considered a menu of options about theft and selected its most preferred one, which is no theft at all. When a judge fails to implement the rule, this lowers social welfare compared to the case where the judge would have abided by the rule and implemented it by enforcing it against the thief.
However, suppose that a soldier gained personally by shirking or a thief gained by stealing, is the result a less desired (or preferred) state for the group? I say it is less desired as compared with the initial goal. I am thinking in terms of a group goal and a group outcome relative to that goal, these being made known by the group. I am not thinking in terms of my amalgamating individual utilities within the group to form a social utility. And this difference in concept is perhaps the sticking point in this discussion. Am I forced into making a set of interpersonal utility comparisons? I don’t think I am.
In my previous blog on this process, I put it like this:
“Bringing everyone into the picture, we know that society has already decided that theft lowers social utility. It has prejudged the case. It did this when it made the law against theft. Therefore, society, because it has this law against theft, is telling us that this or any particular theft lowers its utility, even when an interpersonal transfer occurs within the society. By adopting a rule against theft, society has already compared the utility gain of the thief to the utility loss of the victim and decided that no matter what they may be, its own, society’s own, utility is lower by that judge not enforcing the prohibition.”
Perhaps what I’m saying is true but trivial. It seems that I’m simply saying that the group prefers to achieve its goal and when it doesn’t, it’s not as well off as if it had. However, in the original paper, the situation went further and was not simple. It involved a duel between two laws, one that forbid theft and one that allowed theft. The idea was to prove that the law forbidding theft had precedence.
Going back to Stefano’s first e-mail, it contained the idea that I was thinking in terms of material utility or confusing two different concepts of utility. Not at all, as the preceding exposition shows. I’ve always thought in terms of preference ordering. My answer to this suggestion of his was this:
“At no time did I have in mind that social utility is a material wealth concept. I’ve been thinking of it as a preference ordering. Of whom? Of what? Of a body of people. Does this mean that methodological individualism has been assumed away by me? I didn’t want it to be assumed away. What I’ve left unsaid is that if there is social utility, that is, if the concept is even viable, it is something derived from individual choices. I’ve always thought of it in this case as being a delegation by the individuals, who create the body and its rules. That is to say, there are created corporate bodies that individuals create and they delegate powers to it, as handled by their representatives. Stockholders elect a Board and the Board creates orderings of preferences and speaks for the individuals. Exactly how this is done in a social setting, a society, is unclear, or at least I never tried to make a theory of it. I clearly accept that process or something like it as being behind a social prohibition of theft and a social preference. And I do not regard social preference as a sum of individual utilities. That cannot be done. But the individuals can create groups and transfer decision rights and re-allocate these rights and create means of monitoring their use and altering structures and personnel, if need be. They can create rewards and sanctions for effecting their joint objectives. A corporation is not simply a sum of individuals, and neither is a society.”
7:59 am on July 23, 2020
