Few would expect a butchering contest at a county fair to change how we think about leadership. Yet in 1907, one in England showed something counterintuitive: decisions may be better made on the front lines than in the C-suite.
Fairs at this time were designed to give city folk a taste of rural life. Rows of stalls showed off the best produce and livestock that farmers could offer. Competitions in sheepshearing and field plowing took place amidst the hiss of meat over flame and the brays of cattle at auction.
In the midst of it all stood a butcher’s booth with an ox on display. Attendees were invited to guess the weight of meat the animal would yield once slaughtered and cleaned. They would write their estimates on paper slips and deposit them in a box, where other attendees could not see.
The polymath Francis Galton took an interest in the event. After the winner was declared, he convinced the organizers to let him bring home the ballots. Galton was interested in democracy, but unsure if it was a workable system. Perhaps he hoped to discover how wise the masses were. To his surprise, the median guess was within one percent of the real figure. In fact, it was closer to the truth than almost all individual ballots. Galton concluded that a reliable measurement was available in the aggregate, but not in any one person’s guess.1
It’s a finding that’s been repeated. Groups of motivated people are, as a whole, more likely to be correct than are their individual constituents, even when some of those constituents are well-experienced and informed. Nevertheless, this power can only be unlocked under a few conditions.
James Surowiecki describes these conditions in The Wisdom of Crowds. Groups are wiser than their members only if their members are diverse, independent, and decentralized (that is, able to draw on knowledge which is not available to their peers).2 Averaging their responses combines this disparate information into a coherent whole. Groups that don’t meet Surowiecki’s conditions tend instead to confirm their existing biases.
When the Many Outthink the Few
Today, organizations making high-stakes decisions use decentralization to gain a competitive edge. The video game company Valve has a “flat” structure, where employees are empowered to act without the approval of a supervisor.3 Google, meanwhile, uses online prediction markets to ascertain the likelihood of events it cannot otherwise foresee.4 By taking decisions out of the hands of individuals, these organizations improve their adaptability and tap into a wider diversity of perspectives. It raises an uncomfortable question for modern leaders: can decisions made from the corner office ever match the insight of those made on the ground?
Of course, many leaders may be wary of remaking their organizations according to trendy management principles. These too often lose their appeal and leave their adopters seeming out of touch, which is why it’s important to remember that decentralization is more of a lost art than a passing fad.
Organizational structures have been decentralized for much longer than they’ve been centralized. Before telecommunications technology, orders from management could be disseminated only as fast as their messenger could travel. This meant that, for governments and businesses, decisions needed to be made before anyone at a central office knew they were happening.
The Original Decentralizers
Adam Smith, the economist often credited with founding the field in its modern form, provided a theory of decentralization in his 1776 magnum opus, The Wealth of Nations. As he observed, workers who carry out a task daily are often more able to improve its efficiency than are their bosses, who must divide their attention among the tasks of all their subordinates.5 Smith gives the example of a boy who was employed on an early steam engine to open and shut a valve so that the piston would rise and fall. One day, he noticed that he could tie the valve to another part of the machine, which opened and closed at the same time. That other part would then do his work for him, leaving him free to go. This innovation spread to other boat crews and eventually became a standard, labor-saving automation. It was made possible not by a top-down directive, but by a worker acting on his own initiative. It was his narrow focus on one specific task that allowed him to notice and correct an inefficiency.
When Hierarchies Fail
When organizations centralize authority too much, they also risk blinding themselves to risks noticed by their less empowered members. In 2023, Titan, a submersible operating tourist visits to the wreck of the Titanic, imploded while on a voyage. All five of its passengers were killed instantly. OceanGate, Titan’s parent company, had been warned of flaws in the vessel’s fault detection system by its director of marine operations, David Lochridge. Despite his position, Lochridge was not given authority to suspend operations and was eventually fired for refusing to sign off on the voyages.6
The problem at OceanGate was, in part, due to a lack of decision-making ability afforded to employees. Lochridge had knowledge that was unavailable to company CEO Stockton Rush, but was barred from putting it into practice. It’s the same dynamic Galton uncovered at that country fair: a set of independent evaluators, taken as a whole, is more likely to approach complete information than that same group working under central direction.
behavior change 101
Start your behavior change journey at the right place
The Perils of Too Much Freedom
Decentralized decision-making does have some drawbacks. The employees and organizational units that benefit from this model often have conflicting interests and undermine each other’s efforts rather than enhancing them.
The department store chain Sears, for instance, collapsed after CEO Eddie Lampert restructured the company as a set of separate, competing business units, which were each expected to turn a profit on their own, and were ranked against one another on that basis. Decision-makers at the firm were, therefore, equally motivated to increase the value they brought to the company and to damage the standing of their peers. As a result, strategies in which one unit sacrificed for the good of the whole (such as offering discounts to lure customers) were impossible to realize. After six years of Lampert’s leadership, Sears was liquidated in bankruptcy.7
Nevertheless, research suggests that delegating power outside of the head office has a generally beneficial result. A study by the Center for Information Systems Research at the Massachusetts Institute of Technology found that businesses that decentralize grow their revenues 12.9% faster than those that do not.8 And in certain crucial fields, such as supply chain management, true centralization may not even be possible according to some analysts.9
What cases like Sears show is that, in a decentralized system, leadership still has a role — though not the traditional one. When front-line actors are empowered to take charge of their own processes, they need the support of a collaborative, resilient culture and an incentive structure aligned with the organization as a whole. Adam Smith reports that the boy who worked on the fire engine only developed his labor-saving automation because he was free to play with his friends afterward. Those who lead organizations should think of themselves as guides and mentors, rather than autocrats.
In its success and in its failure, decentralization has revealed something important about how humans behave. We actively seek ways to better ourselves and take charge of our situations. When leaders promote structures that reward the cooperative aspects of this (at Valve, for example), they create opportunities for individual ambition to align with collective success. When they pit their subordinates against one another (as at Sears), they turn self-improvement into a destructive force. The better path requires us to place trust in those around us. If we are too afraid to surrender control, too uncomfortable with uncertainty, we only allow ourselves to fail.
