In 1972, Irving Janis, a research psychologist at Yale University, documented a series of catastrophic foreign policy failures that shared a singular, chilling characteristic. From the lack of preparation for the attack on Pearl Harbor to the escalation of the Vietnam War and the disastrous Bay of Pigs invasion, Janis observed that the architects of these failures were not incompetent or ill-informed. They were, in fact, some of the most intellectually gifted individuals in the American government. Yet, when placed in a room together, their collective intelligence seemed to evaporate, replaced by a psychological drive for unanimity that Janis termed "groupthink." The Bay of Pigs remains the quintessential example: a plan so flawed that even a cursory independent review would have exposed its tactical impossibilities, yet it was approved by a committee of brilliant men who were more concerned with maintaining the social cohesion of the group than with challenging a failing strategy.

The cost of this dynamic is not confined to the history books or the halls of the State Department. In the corporate world, the committee has become the default setting for risk mitigation, yet it frequently achieves the opposite. When the Swiss national carrier Swissair collapsed in 2001—a company once so financially stable it was known as the "Flying Bank"—the subsequent post-mortem pointed directly at a board of directors that had become a closed loop of consensus. They had pursued a "Hunter Strategy" of aggressive acquisitions that drained their liquidity, yet the board, comprised of elite figures from Swiss business and politics, failed to produce a single dissenting voice until it was too late. The mechanism of failure was not a lack of data, but the social pressure to remain "collegial" in the face of impending disaster.

The failure of committees is often a matter of simple social arithmetic. As a group grows in size, the number of potential interpersonal relationships grows exponentially, and with it, the pressure to maintain harmony. Research conducted by Professor Jennifer Mueller at the University of Pennsylvania’s Wharton School suggests that as group size increases, individual performance actually diminishes—a phenomenon known as "relational loss." In a committee of twelve, the individual feels less responsible for the outcome and more concerned with their standing within the hierarchy. This shift in focus from the problem to the person is where the quality of decision-making begins to degrade.

In a typical board meeting or executive committee session, the "Information Signal" is often drowned out by the "Social Signal." The Information Signal is the raw data, the market trends, and the technical risks. The Social Signal is the desire to be seen as a team player, the fear of appearing obstructive, and the subtle cues from the person with the most power in the room. When these two signals conflict, the Social Signal almost always wins. This is why committees tend to gravitate toward the "least contentious" option rather than the "most effective" one. The result is a regression to the mean—a decision that is safe enough to be agreed upon by everyone, but too diluted to be successful in a competitive market.

Consider the development of the Iridium satellite phone system by Motorola in the 1990s. It was a $5 billion project overseen by a series of committees and boards. Despite the fact that the technology was rapidly being overtaken by terrestrial cellular networks and the handsets were the size of bricks, the project continued for over a decade. The committee structure ensured that no single individual felt empowered to pull the plug. Each member assumed that if there were a fatal flaw, someone else would have raised it. This diffusion of responsibility allowed a multi-billion dollar mistake to proceed with the unanimous approval of some of the most experienced engineers and executives in the world.

We are conditioned to believe that a "cohesive team" is the pinnacle of organizational health. However, in the context of high-stakes decision-making, cohesion is often a liability. When a group becomes too tightly knit, the cost of dissent rises. To disagree with the group is no longer seen as a professional contribution; it is perceived as a personal betrayal or a threat to the group’s identity. This is the "Consensus Trap." It creates a psychological environment where the most valuable information—the dissenting view—is the most likely to be suppressed.

The University of California, Berkeley, psychologist Charlan Nemeth has spent decades studying the impact of dissent on group performance. Her findings are counterintuitive: even when a dissenting opinion is wrong, it still improves the quality of the group’s decision. This is because the presence of dissent forces the majority to process information more deeply and consider a wider range of alternatives. Without that friction, the mind takes the path of least resistance. In a committee, that path leads to "confirmation bias," where the group only seeks out information that supports the emerging consensus and ignores everything else.

This tax on dissent is particularly visible in the venture capital industry. The most successful investments are often "non-consensus" bets—ideas that look like failures to the majority but are recognized as opportunities by a single, insightful partner. If a venture capital firm requires a unanimous committee vote to make an investment, it will almost certainly miss out on the "outlier" successes like Airbnb or Amazon, which were widely ridiculed in their early stages. By the time a committee can agree that an idea is good, the opportunity for outsized returns has usually passed. The committee, by its very nature, is a machine for eliminating the unconventional.

The most effective antidote to the failures of committee-based decision-making is the restoration of individual accountability. In the late 1990s, Amazon’s Jeff Bezos famously instituted the "Two-Pizza Rule"—no team should be so large that it cannot be fed by two pizzas. But the size of the group was only half the equation. The more critical component was the "Single-Threaded Leader" model. For every major project or decision, there is one person who is personally responsible for the outcome. They may consult a wide range of experts, but the decision is theirs alone, and they cannot hide behind a committee if it fails.

This model changes the psychology of the room. When one person is accountable, they no longer seek the "least contentious" path; they seek the "most correct" path. They are incentivized to seek out dissent because they know that an overlooked risk will be their personal responsibility. In this framework, the group serves as a resource for the decision-maker, rather than a collective shield that diffuses blame. The difference in rigor is measurable. When an individual knows their name will be uniquely attached to a $100 million investment, their level of analytical care is significantly higher than when they are one of twelve people signing off on a memo.

The British civil service, long criticized for its "management by committee" culture, saw a shift in this direction with the introduction of Senior Responsible Owners (SROs) for major infrastructure projects. Before this, projects like the early stages of the NHS National Programme for IT suffered from a lack of clear ownership, leading to billions of pounds in wasted expenditure. The SRO model was designed to ensure that for every project, there was a "single face" accountable to Parliament. While not a panacea, it introduced a level of personal consequence that forced a more rigorous examination of project viability than the old committee structures ever allowed.

If a committee must be used—as is often the case for legal or regulatory reasons—the social dynamics must be intentionally disrupted. Janis suggested several "antidotes" to groupthink, the most effective of which is the formal appointment of a Devil’s Advocate. This is not a casual suggestion to "think differently," but a specific role assigned to a member of the group whose job is to find every possible flaw in the prevailing argument. By making dissent a formal requirement, the social cost of disagreement is removed. The person is not being "difficult"; they are performing their assigned duty.

Another high-precision tool is the "Pre-Mortem," a technique developed by psychologist Gary Klein. Before a decision is finalized, the committee is asked to imagine a future where the project has failed spectacularly. They are then tasked with working backward to determine what caused that failure. This simple shift in perspective—from "why this will work" to "why this failed"—bypasses the social pressure to be optimistic. It gives members permission to voice their anxieties without appearing to be "naysayers." In a study of project teams, those using the pre-mortem technique identified 30% more potential points of failure than those using standard risk-assessment methods.

The most sophisticated organizations also employ "Red Teaming," a practice borrowed from the military and intelligence communities. A Red Team is an independent group tasked with challenging an organization's assumptions and strategies. During the lead-up to the 2011 raid on the bin Laden compound in Abbottabad, the CIA employed multiple Red Teams to challenge the evidence that bin Laden was actually there. They were specifically instructed to find alternative explanations for the data. This forced the decision-makers to confront the gaps in their knowledge and prevented the "certainty creep" that often infects high-stakes committees.

The ultimate failure of the committee is not a lack of intelligence, but a lack of humility. The group environment creates a false sense of security—a belief that "we can't all be wrong." This collective overconfidence is what leads to the most expensive errors in business and governance. The solution is not to abandon collaboration, but to recognize that the most valuable member of any group is the one who is willing to be uncomfortable.

The principle that emerges from fifty years of research into group dynamics is that the quality of a decision is inversely proportional to the comfort of the group making it. If a meeting ends with everyone nodding in agreement and feeling a sense of harmonious closure, the decision is likely flawed. True analytical rigor is messy, uncomfortable, and socially taxing. It requires a deliberate rejection of the easy consensus in favor of the difficult truth.

As we move into an era where algorithmic decision-making and artificial intelligence are increasingly integrated into the committee room, the human element becomes even more critical. The machine can provide the data, but it cannot provide the courage to stand against the tide of a group’s collective bias. The future of effective leadership lies in the ability to build structures that protect the individual’s voice from the gravity of the group. The most resilient organizations will be those that treat consensus not as a goal to be achieved, but as a signal to be questioned. When everyone in the room is thinking the same thing, it is a near-certainty that someone is not thinking at all.