Reprint: R1009B
Individual judgment and initiative are essential to the success of the modern capitalist economy. At the same time, rules and centralized systems are needed to bring order and prevent waste. Balancing decentralization and command-and-control modes of decision making has always been a struggle, and organizations have experience managing the tension.
In recent times, though, a new form of centralized control has taken root: mechanistic decision making based on top-down statistical models and algorithms. This has been especially true in finance, where risk models have replaced the judgments of thousands of individual bankers and investors, to disastrous effect.
The problem with the statistical approach is that it cannot adequately account for the uncertainty inherent in economic decisions or the idiosyncratic nature of human activity.
What finance in particular needs is a return to judgment. The author offers some broad guidelines for this: Computerized controls work best with inanimate products and processes that can be physically shielded so that the variations of conditions can be minimized. Computers also shine when, like the configuration of pieces on a chess set, the number of possible outcomes is vast—in fact, this vastness often gives the computer its edge—but they all conform to well-specified rules. Conversely, human judgment is favored when shielding is difficult, outcomes are ambiguous, and the possibilities are open-ended.
Ultimately, however, the “right-sizing” of judgment is itself a matter of judgment.
