Monday, January 26, 2026

A Grand Unifying Theory: Groupthink / Power

This post is one of six posts in a series on this topic. The full list of posts are linked here for convenience.

It seems logical to assume the ultimate behavior of a collective of humans would be some sort of weighted average of the behaviors and goals of the individuals in that collective. If a group of one hundred people split 52/48 between tendency A and tendency B, it seems logical that the collective would "tend" towards A a majority of the time, maybe a super-majority of the time depending on decision making rules. If there is uncertainty about the exact split of those in that collective, it seems obvious that modeling the behavior of those individuals more accurately somehow would improve the accuracy of predictions about the behavior of the collective.

Reality seems to support vastly different conclusions. A collective with an arbitrary split of 52/48 between A and B might actually exhibit a significant skew in direction C, in a third dimension not predicted by merely analyzing A and B. This divergence has nothing to do with the stated purpose of the organization, be it business, political, social or legal. It stems from innate aspects of humans working within ANY sort of hierarchy, regardless of why that hierarchy has been adopted or imposed. Understanding this divergence is vital to understanding the nature of solutions that must be pursued when attempting to correct for bad behavior after calamity strikes or correcting behavior before calamities are created.


Groupthink: None of Us Is As Dumb As All Of Us

The author of this entire series hereby stipulates that this portion of the analysis has the least grounding in any concrete scientific or quantifiable measurement. It is thus the section most guilty of hand-waving and "pop-psychology" methodology. However, any time that a complex problem can be simplified to fit on a Despair poster and sell thousands of copies, there's a nugget of truth to be unearthed and polished up. In this case, these observations are based on over thirty years watching managers at all levels in their native habitat, the corporate organization chart. Those observations make it clear that the divergence between INDIVIDUAL tendencies and GROUP tendencies stems from these key factors:

  • differences in human affinity to hierarchy and power
  • the complexities made possible through specialization and its resulting hierarchy - complexities that grow exponentially with the size of the organization
  • differences in financial and social incentives offered to people based on position and status

In prior sections of this analysis, discussions about productivity and human creativity stated that human interaction with others accelerates learning by sharing experiences which can improve rote productivity or lead to improved processes that improve productivity. Here, an argument is being outlined that assumes working with others IMPAIRS productivity by triggering unexpected interactions that drain effort away from formal goals and expend it in directions none of the individuals supposedly want individually. Three factors make this possible.

Affinity to Hierarchy -- This parasitic drain stems from imposing hierarchy upon a collection of people who will have different individual tolerances for and interest in such hierarchy. Managing one's relationship with that hierarchy and attempting to influence one's role in that hierarchy becomes another task in the day of each member of the collective. Those who seek a sense of structure will devote time to syncing themselves to whatever direction can be gleaned from the hierarchy. Those that seek power and status will devote time attempting to advance their position. Those turned off by hierarchy will attempt to minimize their interaction with the collective and perform as much of their work independently as possible.

Hierarchy and Complexity -- Even for people with no particular aversion to hierarchy, larger organizations with more complex hierarchies make extreme complexity more manageable, but the complexity can never be completely eliminated. Large organizations creating complex goods and services require processes unique to the organization that have nothing to do with the service or product being delivered. This becomes its own area of specialization that can be VERY abstract, making information about such work prone to misinterpretation. Unfortunately, this complexity can also make it easier for bad actors to hide bad actions behind the complexity, delaying recognition of actions that require correction due to quality issues or outright malfeasance.

Incentives -- If the range of individual feelings toward hierarchy is wide, the range of incentives influencing the BEHAVIOR of people at different levels of a hierarchy is stupendous. Arguably, these widely varying incentives drive most of the unpredictable variance between goals of individuals and actual outcomes from an organization. Those familiar with the work environment in Corporate America are familiar with the trope about "executive hair" and how those in "Mahogany Row" tend to conform to certain stereotypes about appearance and dress. These sartorial similarities absolutely PALE in importance to the similarities in more critical areas of communication tendencies and risk-taking.

On the surface, senior leaders usually APPEAR to be very calm, deliberative and conservative in their interactions with nearly ANYONE in the organization. This veneer of inscrutability is thought to reflect confidence, leadership and decisiveness (if not infallibility). In reality, in many corporate settings, pay incentives are so extreme that leaders exist in a vastly different bubble than average people in the same organization. The stakes for the next "win" are so high that leaders certainly have an incentive to swing for the fences to achieve that goal but their pay is already so high at its minimum that their worst case scenario (failing, getting fired, getting a golden parachute) is still one hundred times better than the life outcome of nearly anyone else. Leaders in this situation do NOT have the same incentives to avoid DOWNSIDES as everyone else and consequently, such leaders are predisposed to accepting higher risks with extreme "personal NPV" for them, even if the "collective NPV" for the business is much worse.

The real problem with this incentive structure is it affects the promotion process because risk takers don't want a layer of management beneath them constantly reminding them of potential downsides. They want people to take orders, rally the troops and take the hill. This creates a self-selecting feedback loop in the organization that filters downward, layer after layer, hire after hire and begins altering the culture of the entire organization.


An Anecdote

The prior analysis made a case that the goals and outcomes from an organization of individuals cannot be accurately predicted by knowing or controlling the goals and motivations of those individuals. There is an even more important corollary to this initial claim. The corollary is that if the behavior of an entire organization cannot be PREDICTED by knowing and controlling the goals of its individuals, it can also be assumed that the behavior of the organization cannot be CHANGED by even conscious attempts to "change the culture" by hiring new members, even at senior levels. The organization itself isn't human and doesn't care WHO is in charge. The organization itself has its own "inertia" beyond the control of the humans within it.

Here's a possibly long but hopefully enlightening anecdote to illustrate this concept.

A third of the way through my career, at the end of the 2000-era Internet bubble, my employer imploded and I landed a new role at another firm in town in roughly the same industry as before. The new job started in October of 2001 and I had two weeks time off to decompress from the prior firm that ended with a whimper and get dialed in for the new role. I had worked previously for a large telecom firm then the small Internet firm that tried to operate as a creative, entrepreneurial startup but nothing prepared me for the culture shock I encountered from the first day at employer #3.

I arrived on my start date at the HR office at 8:00am, assuming the new boss would be there to escort me to my office and begin doing on-boarding paperwork, getting a laptop, etc. Nope. The HR person didn't even show up until 8:40. The new VP boss didn't show up until 8:54am. He escorted me to his section of the building but conducted three other conversations with people he encountered in the halls and elevator along the way, treating me like an annoying runaway dog being dragged back to the family yard. No sense of professionalism towards a new hire and rude to not only me but everyone he encountered in that six minutes.

Within an hour on that first day, I was pulled into a CRUCIAL meeting. It seems the company had partnered with another company to actually operate the gear delivering services to roughly forty percent of the entire company but that company was declaring bankruptcy and had posed an ultimatum. Give us some bridge money to keep your stuff alive while you try to migrate it to your own stuff or don't pay us anything, "go dark" on December 1 and lose all of your customer revenue until you can turn up your own network with whatever customers are still willing to stay with you after dropping their service.

The "leaders" in this meeting including my boss were collectively flummoxed as to which alternative to take. They were completely at a loss as to how to even model the problem from a financial and calendar perspective. Being unsure of how input from a first-day employee would be received among ten corporate VPs, I said nothing in the meeting. As I walked out with the new boss, I outlined how the problem could be modeled to clarify the decision for execs:

  • identify all key locations
  • identify cost of new OC3 or larger circuits
  • identify equipment intervals for new gear
  • identify time required to pre-build new network configurations for new year
  • identify revenue/sub and subscriber counts at each locations
  • calculate revenue lost between go-dark and our expected turn-up date
  • identify keep-alive bridge costs between go-dark and our expected turn-up date
  • calculate the difference: if revenueloss > bridgecost, pay bridge cost

He and a peer VP working the problem with him heard this and said, great, mock it up and we'll review that in the next meeting. It took me about 20 minutes to model the structure of the model and sanity check it with mock figures and another day to find people with access to appropriate figures and get those numbers plugged in (remember, I'm a Day One employee who knows no one).

On the day of the next CRUCIAL meeting, I expected to email one of the VPs the spreadsheet and expected one of THEM would actually drive the projector in the boardroom and talk through the explanation. Nope. As I provided the quick summary to them on which variables to flip back and forth to better highlight the contrast between the alternatives and how they varied in different markets, they said we don't know how to do that, you bring your laptop to the meeting and drive the screen and talk through the analysis.

So on day three or four of my job, I sat in the boardroom with the CEO, a few SVPs of operations and finance and probably ten other VPs talking through the model and explaining to them that given uncertain delivery intervals of replacement gear, uncertain intervals for delivery of new circuits to the new gear and the likely churn of customers to competitors if a go-dark approach was chosen, it made sense in most markets to pay the bridge costs, keep the customers lit while expediting our internal network turn-up.

In reality, ANY of the executives in that room should have possessed the basic "feel" of that business to think through those options verbally and come to the best decision without a meeting. But even with a spreadsheet prepared to outline that train of thought, none of them had the basic Excel literacy to scroll up and down in the spreadsheet, much less the acumen to make a multi-million dollar decision based upon the content without being spoon-fed the analysis.

Only a few days after that, work began to plan the connection of our gear to new backbone circuits to other carriers. Engineers in one region were concerned about how to incorporate the new circuit while keeping the prior circuit connected for an overlap period. In network circles, this is a relatively straightforward process of configuring "peer" connections using Border Gateway Protocol (BGP) and configuring ranges of your IP space to "advertise" as available to that upstream provider while keeping other ranges internal to your network. They seemed stumped so I told them, "I know BGP pretty well, export your current configuration file, hide the current passwords and mail me the file and I'll take a look."

They did. I took a look. The name of the router was fw1 (it should have been something like frtwtxbb01). The current configuration had no BGP configuration whatsoever. It had a static "default route" entry in another routing protocol (OSPF) not normally used on peer connections between Internet carriers that pointed all traffic out one port on the router. This was a router serving (at that time) about 120,000 customers. With no dynamic rerouting of traffic. No filters preventing leaks of private IP ranges to other networks. No policies to reject accidental advertisements of private IP ranges from upstream providers. Astonishingly amateurish.

I was with that firm for twenty plus years. Over that time, the CEO office changed hands five times. The SVP positions probably churned out every six or seven years on average. VP slots often changed every three to four years. Yet despite that turnover at the senior leadership levels, there are elements of that original 2001 culture I encountered my first day -- hitting me like walking into an invisible wall -- that persisted to the day I left. And many others who hired on over this same period reported the EXACT same jarring Day One experience upon joining the firm. The rudeness of leaders. The ignorance of leaders. The arrogance. The surprising lack of proficiency among technical roles. Home-grown tools and spreadsheets for tracking projects, costs and future budgets involving BILLIONS of dollars that changed every year but never proved more suitable for their stated purpose for anyone who had to populate them or make decisions off them.

The company remains the corporate management equivalent of the Salvador Dali painting, The Persistence of Memory. Things look normal after a quick glance but then you notice the melting clocks dripping off the table and hanging in the trees.


Power and the Persistence of (Bad) Culture

The mix of high churn in leadership roles with stagnant (often toxic) culture described above seems completely incongruous. The odds would seem extremely low that a multi-billion dollar firm could encounter THAT much leadership churn without any of them making ANY dent in ANY of its systemic cultural problems. One might think that purely based on averages, half of those leaders would have brought in some "best practice" and maybe HALF of those would have taken root and become reflected in the culture twenty plus years later. In fact, those working in the company that entire time would come to the exact opposite conclusion. Long-timers would conclude the company as an organization essentially developed an immunity that killed off any new directive or process that conflicted with established patterns and the inertia of those toxic patterns exceeded the power and influence of even the CEO of the company. (And this is assuming each of those new leaders entered with the absolute best of intentions and competence. Reality suggests otherwise.)

How does this organizational immunity to improvement develop?

Here are likely contributors to this phenomena:

  • Flavor of the Month -- Churn at senior levels is nearly as toxic as stagnation with weak talent. Frequent role changes convince lower tiers to ignore the next new thing and wait for the new guy to churn out rather than adopting requested changes.
  • C-Grade Players Don't Attract A-Grade Players -- "Leaders" who swap senior positions every three years might be doing so to avoid looming accountability for major bad decisions. They're not "A" players, they're "C" players. When these mercenaries quit one firm and join another, they often hire their C-tier buddies and bring them in as well, further lowering the quality curve within the new company.
  • Slouching Towards the Mean -- The sheer size of large organizations guarantees their cross-section of employees / members will tend towards societal averages in every measure. A firm employing ten people specializing in a unique technology can hire above the norm and keep certain character traits out of the firm. A firm operating 200 retail locations open 16 hours a day across forty states is going to have an employee base that much more closely mirrors all of the pathologies found in the overall population. These pathologies WILL find occasional expression within the company and often trigger many of the odd outcomes that weren't formally stated by anyone as a goal.
  • Hidden Truces Thwarting Accountability -- Senior leaders often declare informal truces with competing leaders whose organizations SHOULD act as checks on each other to preserve autonomy. Let me do what I want with my financial systems and I won't second guess your IT spending plan in front of the CEO.

That is how a COLLECTION of individuals operating as a single entity can exhibit a specific behavior that no INDIVIDUAL member of the collective claims to support, even if some members of the collective hold positions explicitly designed to THWART that behavior. The other forces at work within the collective jointly exert more inertia allowing the behavior through indirect influence than any individual can identify and counteract.

This clash between these collective secondary forces versus individual actors holds true beyond corporate environments. It holds true in any sufficiently large organization -- charities, educational institutions. governmental bodies, religions, unions...


Organization Power and Society

If the strange immunity of organizations to the exercise of power within them by their leaders makes sense, it has profound impacts for society. First, once an organization begins exhibiting patterns of abusive behavior (monopolistic practices for businesses, wasteful extravagance for charities and universities, captured gerrymandered-to-death political bodies, abusive treatment of believers in the case of religions, etc.), the INSTITUTION will take actions to defend and perpetuate ITSELF despite what any member or leader claims to do to correct the issue.

Because of that first point, organizations that begin fighting off external efforts to correct perceived internal problems will devote increasing shares of internal resources to the defensive battle rather than whatever might have been the original "mission" of the organization. This dilution of focus can materially alter the glide path of the entity, with dire consequences in the longer term.

Even if punitive financial or criminal penalties are assigned to the entity or specific members, the culture that allowed those actions to take place will likely persist, The explicit form of offense targeted by the correction might never happen again, but those lower level innate characteristics that allowed it to occur are still in existence and may lead to different actions with similar negative consequences to still occur.

This pattern also means that no outside entities attempting to coerce an offending organization into altering its behavior can take solace in having personal relationships with that organization's leaders as a lever for forcing change. The leaders you know at that organization may be as upstanding as you could ever wish for and as well-intentioned as you could hope for but if their organization has systematically engaged in bad behavior, you cannot confuse the HUMANS with the ENTITY. The ENTITY is likely the root problem and deserves no mulligan or do-over.

It's also important to point out that this pattern is in no way based on the original intended purpose or original actual behavior of an organization. The pattern defined here is driven solely by the size of an organization, the resources it grows to control and these "ricochet" effects of human reactions to operating within hierarchies.

The real conclusion from these observations is that for some organization offenses, NO amount of penalty can eliminate the "organizational DNA" that permitted a fault to occur and might allow similar faults in the future to occur. For some organization problems, no single change in leadership will be capable of changing the culture. The inter-related incentives and secondary impacts are too complex to understand and unwind and the "muscle memory" of the organization itself is far longer than the muscle memory of any subset of employees.

That means that for some problems faced by some organizations, those problems are unsolvable as long as the organization continues to exist in its current form. In the case of a corporation that gambles heavily and triggers a market collapse, that means that firm should be literally eliminated from existence. For a monopoly abusing market share to cheat customers or stifle competition, that means breaking the company up, not simply hitting it with a billion dollar fine. For a government whose "control levers" have become completely controlled by uber-wealthy private interests or by the parties themselves rather than the public...?

That's a topic for the concluding entry in this series.


WTH
Posted at