We made a case for market decentralization in Chapter 3, arguing that it was informationally cheap (in fact, costless) in terms of information needed by the planner and that under some circumstances it could be used to generate desired outcomes. However, we began to see some limitations to its use in Chapter 4, where we developed a whole class of goods that could not be allocated through standard private markets. So now we ask whether there are more general allocation procedures that will work for more general economic environments.

Of course, planning problems would be relatively simple if the planner could costlessly obtain all the relevant information; he would then be in a position to compute and administer directly the optimal allocation. However, such informational requirements are far too stringent. The planner cannot possibly obtain and verify the accuracy of all necessary information. Consequently, any reasonable planning mechanism is going to involve some decentralized elements; the planner is going to have to rely on economic agents to provide some information (and perhaps make some decisions) that cannot be monitored (or verified). Clearly, once planning procedures have this feature, incentive issues are sure to arise as well.

Toward the end of this chapter, we will discuss planning mechanisms in the abstract, making precise the nature of decentralization requirements and associated incentive problems. However, we lead into the subject first with a discussion of some examples that are close in spirit to the market mechanism.