Review of Satisficing and Maximizing: Schmidtz
Of the many things argued in this essay, two are interesting. First, Schmidtz claims that incomplete information and non-zero transaction costs, such as costs of acquiring information entail that in many cases satisficing is the proper strategy. We need a stopping point “that limits how comprehensive a body of information we insist in gathering before stopping the search…” (Satisficing and Maximizing, 37) Alright, suppose that you indeed find something “good enough.” But if you could instead get something you perceived was better and felt contributed to your happiness more at zero cost, wouldn’t you do it? What could possibly stop you? It seems that the reason to limit your resources spent on information gathering is to avoid psychic loss which occurs if the costs of getting the information necessary to make a decision outweigh the benefits. Now Schmidtz might argue that without proper information you can’t optimize at all. Let’s grant him this point for the sake of argument. Then satisficing will be the optimal strategy. Or, rather, trying to make the best choice based on incomplete information will be fully equivalent to choosing something “good enough.” But satisficing can never produce results better than optimizing; if it does, then it itself becomes the optimal course of deliberation.
Suppose, for example, that your strategy for searching for a new house is to wander around the neighborhoods aimlessly hoping to find something suitable. You don’t know the state of the market, relative house prices, where to go, or when to stop. Assuming in our example that the information costs are prohibitive, delineating the criteria for an acceptable house and stopping after finding something that satisfies them, though it seems like satisficing, might well be optimal, as well.
Second, Schmidtz gives examples of situations in which there seems to be no optimum at all.
(1) “[S]uppose you are immortal and are also fortunate to have in your possession a bottle of EverBetter Wine. This wine improves with age. In fact, it improves so steadily and so rapidly that no matter how long you wait before drinking it, you would be better off, all things considered, waiting one more day.” (42) You have to drink the wine at some point, lest the bottle proves to be useless to you, yet no matter which day you pick, you don’t want to drink it then. Reply: Pleasure from the sense of taste cannot be infinite, so the example is contrived.
(2) Computer hardware has improved in accordance to an empirical trend called the Moore’s law. That means, quite fast. The moment you buy a computer, it is already obsolete. Must you therefore always wait for a better machine, never buying one? Reply: Waiting has disutility, so that has to be weighed against the utility of a computer that allows you to do everything you want for a period of time. A future computer may be technologically superior but it need not satisfy your desires any better than a less powerful device. In that case, it is rational to buy.
(3) Suppose the bank offers you 100% per year interest on your loan. Does it mean that you will never use the money, because it will seemingly always pay to keep the money in the bank, growing at this enormous rate? Reply: Once you have, let’s say, a few billion dollars, the marginal utility of money (the utility of an extra dollar) will become negligible; the disutility of waiting will make itself felt; and it will become reasonable to withdraw and spend the money.
(4) Buying a better house will cost you $1000; living in that new house yields $100 worth more happiness than living in your original house. Abstracting from discounting the future, you will recoup your investment in 10 months. Suppose now that after 4 months of living in the new house, an opportunity arises to move into a still better house which costs $2000, and living in this third house is $200/month better than living in the first house. And so on ad infinitum. But if you keep moving forever without stopping, you will never profit. Yet stopping at any point seems arbitrary. What’s going on? Reply: Again, there seems to be a natural limit of how pleasurable a new house is going to be compared with the old house. At some point the utility of a marginal physical improvement will be reduced to 0. Hence this example is also dubious and does not entail the reasonableness of satisficing.
(The first and fourth example are Schmidtz’s; the second and third are mine.)
To be continued…