Administrative failures are likely---our author did the math!

Juries, however disinterested or expert, are unlikely to be able to predict the course of future developments or to pick the winners who will prevail through shifting and dynamic conditions. They are typically charged to limit their consideration to the proposals before them rather than to incorporate comprehensive information about the next best alternative on a larger current or intertemporal scale. The standard analysis of decision- making under uncertainty indicates that an option to flexibly defer choices will, holding other things constant, have a greater expected value than a “now or never” decision. Administered systems are likely to exercise options too early, leading to irreversible choices and monolithic path dependence. Markets, by comparison, engender higher-valued options owing to the ability to defer decisions until more information becomes available. For instance, modern patent grants maintain an open option during the life of the patent with a maximum term of twenty years. Markets will thus tend to have more prolific branches on the binomial decision tree than an administered process such as a prize system, which perforce will prematurely lop off a greater number of branches at the date when the prize has to be awarded.

Consider a prize system where n ideas are submitted at time t, but only the top three can be selected at time t + 1. This implies that n – 3 are rejected, even if their value is positive, and those paths are foreclosed and no longer available to the administrator. If market conditions change sufficiently in the future, in such a way that one of the n – 3 options is now viable and possibly more valuable, it is unlikely that the administrator will be able to revisit the decision. Administrators further have control over their institutions, and researchers show that the option to defer choices is less likely to be chosen when conflict is low.42 Planners do not internalize all of the returns from risk-taking, and are thus ill-equipped to confront an innovation process that Bill Janeway wryly characterizes as trial and error and error and error.43 Administered systems tend to be more closed, with little outside knowledge of the factors that led to the decisions, and once a prize is awarded with great fanfare, it is difficult to trace the subsequent outcome because attention is diverted to other competitions. Planners have an incentive to proclaim their successes and downplay or deny their failures, leading to an asymmetry in the information that is made available. Markets are less likely to have this sort of lock-in and will tend to have access to a greater number of the initial options, as well as more useful information about failures that can benefit all participants.