John Deighton, Daniel Rosenblatt, and Patrick Warren all made great comments on yesterday’s post and ideas began to flourish like an Elizabethan garden. Could I ask the reader who has not read yesterday’s post to take a look at it? I know you are smart enough to “reverse engineer”” that post from these comments. (It is what I would try to do.) But these comments are so good, they move as swiftly as Smartie Jones away from the original. You might just as well start at the beginning. So, please, page down and come back. John Deighton: The posting got me thinking about a conversation I had on a bus a couple of weeks ago. My thought went, ‘This speaks exactly to the impossibility of trust, and the need for structures of incentive. What was it the man on the bus said – trust should be banned from discourse except when talking of relations not continuously in negotiation?’ There are some relations – say the relation between a professor and his/her university, or between a board member and a corporation, or between a son and father (your posting about Father’s Day is very apt here) that admit of no give and take regarding the obligations of the role. The relation is categorical. There is a hard, bright line between performance of the duties demanded of the role and failure to perform – no room for rewriting a contract as new circumstances arise. Fiduciary and sovereign/subject relations are of this kind. You can’t say I’ll be your auditor and if I catch you cheating I’ll turn you in, and then make an exception when the auditee offers a bribe or a hard luck story. Everyone has to be able to trust an auditor not to do the decent thing or the self-interested thing, but to do the thing the role specifies and nothing else. The auditor/auditee relation is categorical. It is said that professors get tenure to remove their fiduciary relation toward the university’s best interests from taint of self-interest. A professor without tenure cannot be trusted to hire someone smarter than him- or herself. It seems to me, though, that in a world that cannot rely on universal belief in an all-seeing God with hell up his sleeve, there are no fiduciary or sovereign/subject relations, and hence no use for the work of the word trust. Grant’s reply to John: Nice one. I think the, or at least one, reason that we can trust a person defined by role has nothing to do with us, not our trust, not their incentive. It is precisely that the person is defined by role, so defined in fact, and so keen to be so defined in fact, that they cannot misbehave without sacrificing their claim to the, variously, majesty, honor, sanctity, verisimilitude, veracity of their role. What keeps them honest is their existential need to remain who they are, to keep the meanings they have. Our job, if we wish to make sure they will not misbehave, is to give them so much majesty, honor, sanctity, and acknowledgment, that they have at the end of the day, more to lose than gain from misbehavior. Or to put this in Shakespearean terms, a monarch might chose to break the faith but he/she would pay dearly for the benefit. Or to put this in anthropological language, their claim to identity, to a desired cultural definition of the self, demands the performed fulfillment of their identity. There is a “violation” clause. You can’t be king if you don’t act like one. And this is not because we will remove your role definition, but because you will. (One thinks here of Nixon, Johnson, and Clinton leaving office in disgrace.) This is perhaps why Elizabethans worked so hard to honor their monarchs. It wasn’t slavish deference. It was the need to press home the definition that would summon the right behavior. In effect, they were saying, “This is who you are. You know what to do.” The cultural meanings that define who you are the cultural meanings that protect us…because frankly we don’t trust you, and we can’t incent you. This adds a third term to the economics dyad, I think. There’s trust, there’s incentive and then there’s definition (or meaning). And we can break these out perhaps like this: trust is sociological, incentive is economic, and definition is cultural (that is, anthropological). John’’s reply: Your point about definition is very good, particularly the purpose of flattery and honor. I would not raise it as a third term though unless you can convince me that it isn’t a form of incentive. If the rewards of adhering to the role exceed the rewards from cheating, I stay in the role. I think the thing about fiduciary and sovereign roles is that they allow malfeasance with little risk of being caught. Are you saying that honoring a monarch inculcates a self-concept in the monarch that is endogenously rewarding, a sense of who I am that is so pleasing to me that I owe it to myself (me, not the flatterers) to live up to it? One of those “I couldn’t live with myself if I…” inhibitors? A transaction within myself, not between me and the courtiers? Quite interesting. If that’s your story, though, it seems to require a divided self, one that is venal (enjoys flattery) and one that isn’t. I think you like divided selves, or perhaps it is situationally engendered selves, while I regard them as vulnerable to infinite regress. If there is a venal me kept in check by a moral me, why not a tempter me to egg on the venal me and a hangman me to regulate the tempter and an absent-minded me who forgets he’s being watched and so on? That’s the great wonder of economics – so parsimonious yet explains so much. That said it should be noted that the self-interested hermit-like economic actor is under attack in economics. Particularly in finance, it is being discovered that if the actor is endowed with cognitive limitations, like limited foresight, confirmatory bias and poor memory, more can be explained. But every refinement that is made to Adam Smith’s simple homunculus of self interest seems a step away from parsimony and another crack in the foundation of the imperial social science. Grant’s reply: Thanks. First, it is true that the person who is defined as a monarch (or president) and treasures this definition of himself does have an incentive to continue in this role. But this incentive comes from the symbolic economy of kingship (or the office of the president) not the exchange of the marketplace. We give him honor, he gives us good behavior. But neither party has a alternative supplier. The value of the value exchanged is not set by the wee contracts by created constantly in the market place. The relationship comes from cultural convention with very strict “felicity” rules, as Austin would say. Or to put this another way, we are incenting the monarch, but we cannot choose the value he gives us in return. In sum, there is something like incentive at work here, but it doesn’t look very much like the “incentive” we speak of in the marketplace. I think “definition” (or culture or meaning) might well survive as a distinct third term. Second, we are creating an “I couldn’t live with myself” condition. And this works because misbehavior actually means “I wouldn’t live with myself.” I would no longer be a monarch in my own eyes. Clearly, monarchs or politicians do misbehave. The definition supplied by kingship or the presidency is sometimes obscured by the temptations of a break-in or an intern. But I don’t believe this temptation presents itself as a calculation of relative advantage. It’s a temptation, a momentary failure of calculation that occurs as a result of the siren call of an intern (and the rash assumption that no one will ever find out). I have edged into your third point: the need to posit several selves. I don’t think I am, for once, obliged to do this. All I need to do is posit temptation. I realize I am resorting to the psychological model supplied by my upbringing in the Protestant church of Canada. But I think it’s apt. Clinton’s famous dalliance did not come from a calculation of benefit vs. risk. It didn’t come for a dialogue between “moral mes” and ‘tempter mes”. It came as a sudden “shorting out” of judgment. There is no second self here. There is just brute temptation. And this is, though I hate to say it, where the “civility” argument comes in. In a civil society, everyone is much more aware of the responsibilities attached to their role, and perhaps more prizing of the definitions that come to them as a result. They know the right thing to do and they believe more completely in the right thing to do. It helps define them. It is just possible that we created Clinton’s moment by so often scorning the office and diminishing the “definitional” benefit he got therefore. Why not act in a non presidential manner if the collectivity is inclined to doubt the role from which I draw my meanings? And before you say that I have opened up a can of worms and must now account for all the meanings supplied by civility and by culture, don’t forget this is what anthropologists do. Late note: I will leave till tomorrow responses to excellent comments by Daniel and Patrick.