Local knowledge problem

The Local knowledge problem concerns how a group can benefit from what its members know in particular places and roles without losing that knowledge when decisions are made collectively. The book treats decentralization as both a source of valuable independence and specialization and a risk: information may remain trapped where it was gathered. Its central distinction is between centralizing authority and aggregating knowledge. A group can connect local insights into a collective judgment without requiring a single leader to direct every contribution.

What Local Knowledge Is

The concept is closely associated with tacit knowledge: knowledge that is difficult to summarize or transmit because it arises from a particular place, job, or experience, yet can be highly valuable. The closer a person is to a problem, the more likely that person may be to understand a workable response. In this view, local expertise is not a lesser substitute for managerial or central expertise; it is a distinct resource that a distant decision-maker may not possess.

That principle has a long history in the examples the book invokes. In ancient Athens, local festivals were left to the demes rather than the assembly, while regional magistrates handled most less serious crimes. In Exodus, Moses’ father-in-law advised him to judge only great matters and leave other decisions to local rulers. Both examples express the same practical premise: decisions can be made nearer to the circumstances and knowledge they concern.

The Decentralization Trade-off

Decentralization encourages independence and specialization, allowing people to acquire knowledge from their own situations. But it does not guarantee that useful discoveries will travel beyond the part of the system where they arise. Information can remain undistributed and therefore less useful than it could be. The challenge is to preserve the local conditions that produce knowledge while enabling the wider group to combine it.

The book frames this as a design problem for any crowd, whether a market, corporation, or intelligence agency. Individuals should be able to develop specific knowledge, while the organization also needs a means of making that knowledge collectively useful. Google’s use of local knowledge from millions of Web-page operators is offered as an example of aggregation that can make searches smarter and quicker. The broader point is not that every institution should be organized identically, but that specialization and collective use of information must be made compatible.

Aggregation Without a Single Center

The pre-September 11 U.S. intelligence community illustrates the difference between decentralization and effective organization. Division among agencies could support specialization and varied analysis; as the book notes, information one set of analysts considers irrelevant may reveal important relationships when another set examines it. The failure was that the system lacked a real means of aggregating judgments as well as information. There was decentralization, but no mechanism to turn the agencies’ separate perspectives into a collective assessment.

A proposal to create a truly central intelligence agency could make information sharing easier, but at a cost: it could sacrifice the diversity, local knowledge, and independence that decentralization provides. The book’s alternative is aggregation rather than centralization. Linking agency computer databases, for example, could improve information flow while allowing the agencies to retain autonomy. The book points out that even two years after September 11, the government still lacked a single unified watch list drawing on all parts of the intelligence community. The example makes the problem concrete: valuable information can exist within a system and still fail to become useful to the system as a whole.

Organizations and the Loss of Information

The same tension appears inside corporations. A rigid hierarchy can separate senior executives from workers who have direct knowledge of operations, while multiple layers make it harder for information to move upward. At one General Motors plant in Van Nuys, two assembly-line workers had been installing a bracket for two years without knowing that it held the car’s sunshade. When a supervisor showed them what happened if it was missed, they learned that the repair required removing the carpet and welding the bracket into place. Their lack of information was not a lack of experience at the task; it reflected a failure to connect that task with its purpose.

A chart from an unnamed company traced 223 different bureaucratic paths a new product idea might have to travel before acceptance. Such structures can obstruct the flow of ideas and leave executives unsure whether their picture of the company resembles conditions on the ground. The book also cites research linking anxiety about advancement with less accurate communication of problem-related information, and studies associating upward mobility among middle managers with withholding reports of failures. In these cases, local knowledge is not merely distant from decision-makers; organizational incentives can discourage its transmission.

The proposed remedy begins with assigning local problems, as much as possible, to people close to them. Companies should not assume that every problem must travel up a hierarchy and every solution back down; experience and local knowledge can outweigh managerial expertise in particular decisions. Yet real employee involvement remains unusual: one estimate cited in the book places the share of American companies making substantial use of high-performance work systems below 2 percent. The case for decentralization therefore concerns both access to relevant knowledge and the ability of people to act on it.

Hayek and the Limits of Aggregation

The book places this argument in relation to Friedrich Hayek’s account of local knowledge and markets. Hayek emphasized that knowledge dispersed among individuals could not be meaningfully condensed into a single central judgment. He viewed markets as decentralized systems in which people communicate and transmit information through prices, and was skeptical of treating the market as a giant calculating machine searching for a right answer.

The book accepts the importance of Hayek’s emphasis on decentralized coordination but argues that his opposition to centralized authority led him to underestimate the possible benefits of aggregation. A pari-mutuel betting market, for instance, has a central authority taking the other side of bets and produces one set of odds from bettors’ wagers. In the book’s account, such a mechanism can yield an uncannily accurate picture of the future. The contrast clarifies the distinction at the heart of the Local knowledge problem: aggregation need not mean replacing dispersed knowledge with the judgment of a central expert. It can instead provide a way for separate contributions to form a collective estimate while their sources remain distributed.