H. P. Grice, "Conversations with Vilfredo"

Pareto efficiency, also called Pareto optimality, a state of affairs in which no one can be made better off without making someone worse off. “If you are provided information, the one who gives you information loses.” “If you give information, you lose.” “If you influence, you win.” “If you get influenced,” you lose.” The  economist Vilfredo Pareto referred to ‘optimality,’ as used by Grice, rather than efficiency, but usage has drifted toward the less normative term, ‘efficiency.’ Pareto supposes that the utilitarian addition of welfare across conversationalist A and conversationalist B is meaningless. Pareto concludes that the only useful aggregate measures of welfare must be ordinal. One state of affairs is what Pareto calls “Pareto-superior” to another if conversationalist A cannot move to the second state without making his co-conversationalist B worse off. Although Pareto’s criterion is generally thought to be positive or descriptive (‘empiricist’) rather than normative (‘quasi-contractual, or rational’), it is often used as a normative principles for justifying particular changes or refusals to make changes. Some philosophers, such as Grice’s tutee Nozick, for example, take the Pareto criterion as a moral constraint and therefore oppose certain government policies. In the context of a voluntary exchange, it makes sense to suppose that every exchange is “Pareto-improving,” at least for the direct parties to the exchange, conversationalists A and B. If, however, we fail to account for any external effect of A’s and B’s conversational exchange on a third party, the conversational exchange may *not* be Pareto-improving (Grice’s example, “Mrs. Smith is a bag.”. Moreover, we may fail to provide collective, or intersubjective benefits that require the co-operation or co-ordination of A’s and B’s individual efforts (A may be more ready to volunteer than B, say). Hence, even in a conversational exchange, we cannot expect to achieve “Pareto efficiency,” but what Grice calls “Grice efficiency.” We might therefore suppose we should invite thet intervention of the voice of reason to help us helping each other. But in a typical conversational context, it is often hard to believe that a significant policy change can be Pareto-improving: there are sure to be losers from any change – “but the it’s gentlemanly to accept a lose.” – H. P. Grice. Refs.: “Conversational efficiency and conversational optimality: Pareto and I.”