A fundamental difference between knowing how best to obtain or produce a predetermined output and ensuring that the output produced represents the welfare-maximizing use of scarce resources.

From Doing Bad by Doing Good: Why Humanitarian Action Fails by Christopher J. Coyne. Page 83.

Trial and error is not a substitute for economic calculation because outside of the market, planners lack the crucial knowledge and feedback that the process of economic calculation facilitates. It is not simply a matter of speed, of markets adapting faster than a trial-and-error process under planning outside of markets. It is a fundamental issue of the planner’s problem whereby decision makers have no means of discovering a solution to the economic problem in order to achieve economic progress. Of course, adjustments can be made in future periods if output targets either are exceeded or fail to be met, but this does not mean that those determining the output targets have transcended the planner’s problem.

This is an important consideration in the context of proposed innovations in the delivery of humanitarian assistance intended to foster adaptability through trial and error (for example, RCTs). These strategies are fundamentally different from solving the planner’s problem through economic calculation. These methods potentially can contribute to humanitarian assistance better meeting predetermined outputs, but they cannot provide a solution to the economic problem because the planner must still predetermine what ultimate outputs should be supplied. There is a fundamental difference between knowing how best to obtain or produce a predetermined output and ensuring that the output produced represents the welfare-maximizing use of scarce resources.