Organizational hierarchy

Organizational hierarchy arranges authority and communication in levels, offering a way to coordinate work and process information but also creating bottlenecks between those who know a problem firsthand and those empowered to act. The book’s central distinction is between hierarchy as a useful organizing mechanism and hierarchy so rigid or deep that it suppresses independent judgment, conceals important facts, and makes a large organization less capable than the people within it.

Structure and the Case for Local Knowledge

Hierarchy is not the only way to organize decisions. Decentralization gives people close to a problem room to use knowledge acquired through a particular job, place, or experience—knowledge that cannot easily be summarized and passed upward. The governing assumption is that proximity can matter: a person who encounters a problem directly may be better positioned to find a workable solution than a distant manager. The book traces this principle to decisions in ancient Athens that were left to local demes and regional magistrates, and to the counsel in Exodus that Moses judge only great matters and leave other decisions to local rulers.

The organizational problem is to preserve local knowledge and specialization while making their findings useful to the wider group. Decentralized units can produce a richer store of information, but there is no guarantee that a discovery in one part of the system will reach another. Nor does centralization automatically solve that problem: putting information in one place may make it easier to oversee, while sacrificing the diversity, independence, and local understanding that produced it. The book’s alternative is aggregation without requiring all judgment to originate at the center.

The possibility that hierarchy can be useful is not dismissed. An analysis by Roy Radner emphasizes hierarchy as an efficient information-processing mechanism. Its value therefore depends on the work an organization needs to coordinate and on whether its channels can carry the necessary information. A structure may clarify responsibility and process decisions, yet still fail if the task depends on dispersed knowledge that cannot travel reliably through its levels.

When Levels Become Barriers

The contrast between Ford and Toyota illustrates how managerial depth can shape a company. In 1980, fifteen levels separated Ford’s chairman from a factory-floor supervisor; Toyota had five. More levels meant more distance between executives and the people doing the work. At a General Motors plant in Van Nuys, two assembly-line workers repeatedly failed to install a bracket that held a car’s sunshade. The omission forced a repair worker to tear out the carpet and weld the bracket in place at the end of the line. One worker had performed her task for two years without being told what part the bracket played. The incident makes the cost of disconnection concrete: workers could complete a narrow operation without understanding its consequence for the finished product.

The same distance obstructed the movement of ideas. A chart reproduced by Thomas J. Peters and Robert H. Waterman from an unnamed company mapped 223 different routes through its bureaucracy that a new product idea might have to traverse before acceptance. With so many layers between executives and employees in the field, senior leaders could not be sure that their picture of the company matched its actual conditions. A hierarchy might bring thousands of people into one organization, but it could fulfill that purpose only if they exchanged and acted on good information rather than allowing each level to become another obstacle.

The incentives within a hierarchy can make those obstacles self-reinforcing. A study of young executives found that those more anxious to advance communicated problem-related information less accurately. Research on middle managers likewise associated upward mobility with withholding news of trouble from a superior; successful executives were less likely to disclose fights, budget problems, and other difficulties. When advancement depends on a superior’s assessment, accurately reporting bad news can become personally costly, even when the organization needs it.

Silence, Targets, and Distorted Information

Resistance to dissent deepens the problem. Organizational theorist Chris Argyris describes inauthentic behavior as common within organizations and points to bosses’ hostility toward opposition from subordinates as an obstacle to exchanging real information. A top-down system can give decision-makers the illusion that their judgment is complete while encouraging everyone else to play along. Since people already tend to avoid conflict and trouble, an autocratic structure makes the emergence of unwelcome information less likely.

Performance targets can distort reporting even when they are meant to improve results. In the early 1950s, sociologist Donald Roy worked as a lathe operator under a piece-rate system: pay per piece rose at successive production thresholds and then was capped. Workers knew that producing too much too quickly could lead the company to raise the thresholds. They therefore limited output and spent effort managing the rate structure rather than maximizing production. Roy called his account “Goldbricking in a Machine Shop.” The example shows how a system that appears to reward productivity can encourage workers to hide their capacity when disclosure may simply bring a harder target.

A similar incentive operates in corporate budgeting. If a manager receives a bonus for meeting targets, the manager has reason to set an easily reachable target by understating expected results, then to protect the bonus once the target is fixed. Michael C. Jensen argues that managers may use accounting devices to improve current results at the expense of the future. The hierarchy’s information problem is thus not merely that facts get lost while traveling upward: people may strategically reshape the facts before they travel.

Coordination Without a Single Center

The book contrasts organizational hierarchy with arrangements that coordinate distributed information without concentrating all authority. In the pre-September 11 intelligence community, agencies such as the NSA, CIA, and FBI had specialized knowledge and distinct analytical perspectives, but lacked a real mechanism for aggregating judgments across them. Creating a central intelligence agency might have put one organization in charge of more information, yet would have risked losing the diversity, local knowledge, and independence that specialization supplied. The proposed remedy was not centralization alone, but aggregation: for example, linking agencies’ databases while letting the agencies retain their autonomy.

The January 2003 discussion within NASA’s Mission Management Team offers a contrasting warning about hierarchy at work in a specific decision. Engineers on the Debris Assessment Team had concerns about the foam strike and wanted images of the Orbiter’s wing before making an informed assessment, but their concerns never received a serious hearing from the Mission Management Team. Its discussion was shaped largely as questions from Linda Ham followed by answers, without a deliberate effort to invite each member’s analysis. When the team’s conclusion was already settled, new information was reinterpreted to fit it. Formal levels and meetings did not ensure that relevant knowledge reached the decision; the structure of discussion restricted what could be heard.

The alternative is not to abandon coordination, but to design channels through which distinct judgments can meet without making conformity a condition of participation. In classroom experiments by Angela Hung and Charles Plott, students who made sequential guesses about which of two urns had been selected often followed earlier guesses even when their own marble draw pointed the other way. When their pay depended instead on whether the group’s majority answer was correct, they paid more attention to private information; individual guesses could become less accurate while the group became more accurate. The experiment clarifies the organizational stakes: allowing people to make independent judgments, and then aggregating them, can outperform a chain in which earlier decisions steer those who follow.

Organizational hierarchy is therefore neither an automatic virtue nor a problem solved by replacing it with an unconnected collection of units. Its performance turns on whether authority and communication let specialized, local knowledge remain independent while making it collectively usable. Where layers discourage dissent or reward the concealment of difficulty, an organization loses access to what its members know. Where coordination joins their judgments without silencing them, hierarchy need not stand between information and action.