The Four Frequencies
The systems we depend on — power grids, supply chains, financial markets, the organizations that run them — share four patterns. These patterns operate at every scale: the infrastructure a society runs on, the company you work for, the friction you navigate in daily life but can’t quite name. The framework calls them frequencies because, like sound, they’re always present. You just have to know how to listen.
Every failure the framework has examined, across six sectors, two scales, and three distinct types of failure, maps to these four patterns. The observation is bounded by the evidence, not offered as universal coverage: six forensic case studies supported by verified citations from independent organizations.
The examples that follow pair infrastructure and organizational scale deliberately. The dynamics are the same. The vocabulary is the same. The scale is different.
Thinness
Where is there no buffer?
Thinness appears wherever reserves, redundancy, and slack have been optimized away, so that small disruptions produce consequences far out of proportion to their apparent scale.
Most organizations have pursued efficiency relentlessly. Inventory reduced. Headcount tightened. Backup systems decommissioned because the primary was “reliable enough.” That efficiency comes at a cost: when everything runs lean, there is no margin to absorb the unexpected. One disruption doesn’t just cause damage. It cascades into the next, because there’s nothing in between to absorb the shock.
The signature of Thinness is how far the damage travels. Not how thin the buffer is, but what happens when it fails. A well-buffered system absorbs a disruption locally. A thin system transmits it everywhere.
At infrastructure scale: Most supermarkets carry three days of inventory. A single pharmaceutical factory pauses, and a continent loses access to a medication. One region produces over 90% of the world’s most advanced semiconductors. The optimization that created these concentrations also removed the cushion between normal operations and crisis.
At organizational scale: Your team runs at 98% utilization. It just settled there, one efficiency initiative at a time. Then someone resigns. Not your weakest performer. Your most versatile one. Three projects stop the same week, because every person was already allocated and there was no one to shift. The backlog starts compounding on day two. By week three, the downstream teams are missing deadlines they didn’t know were connected to a single departure.
That ripple, when the disruption has traveled three levels further than the original event, is Thinness making itself visible. The departure was one person. The damage ran through the whole operation.
In the SVB analysis, the bank’s concentration of unhedged long-duration securities created a condition where a single market shift — rising interest rates — produced consequences that overwhelmed every other buffer in the system. Read the analysis →
Permission
Who controls the gate?
Permission appears wherever access to essential decisions, services, or resources depends on approval from systems or authorities that can delay, deny, or revoke that access, often without explanation or appeal.
Every organization has a permission architecture: the web of approvals, authorities, dependencies, and controls that determines who can make decisions and how quickly. When that architecture works, it protects the organization. When it calcifies, it becomes the thing the organization needs protection from.
The signature of Permission is revocability: the distance between having access and being guaranteed access. When that distance widens, everything that depends on the access becomes contingent on someone else’s decision, someone else’s timeline, someone else’s priorities.
At infrastructure scale: Your bank app shows “account restricted.” A fraud detection algorithm flagged a routine transaction. Until someone reviews the flag, you cannot buy groceries. After a wildfire season, your insurance carrier declines to renew. The gate that granted coverage has quietly closed. The house hasn’t changed. The algorithm has.
At organizational scale: A frontline manager spots a customer problem on Monday. She knows the fix. It’s not complicated. But it requires a budget exception, which requires her director’s sign-off, and her director is traveling until Wednesday. On Wednesday, the director asks for a one-page justification. By Thursday afternoon, when the approval comes through, the customer has already left. The whole sequence took four days. The fix itself would have taken twenty minutes.
That wait, the accumulated distance between seeing a problem and being allowed to solve it, is Permission at its most recognizable. You feel it at work, at the pharmacy counter, on hold with an airline. The gate is the same. The scale is different.
In the Boeing analysis, the permission architecture had been restructured so that engineering authority was subordinated to production schedule authority. The people closest to the technical risk could document concerns but could not stop the process those concerns were about. Read the analysis →
Management
Who knows what, who decides what, and is the gap visible?
Management appears wherever the information connecting what’s actually happening to the people making decisions has degraded, creating a gap between what the organization measures and what is real.
Organizations make decisions based on what they believe is true. When that belief is accurate, decisions have a chance. When it isn’t, the organization is flying blind and often doesn’t know it. The gap shows up in metrics that have drifted from what they were supposed to measure, in feedback that gets filtered or softened as it moves up the hierarchy, until the version that reaches leadership no longer resembles what was said on the ground.
The signature of Management is the distance between the instrument and reality. When that distance widens, every decision downstream is made on distorted information. And the organization loses its ability to self-correct, because the instruments it relies on to detect problems are themselves part of the problem.
At infrastructure scale: During a heatwave, your smart thermostat adjusts itself. The utility needed to shed load. Your comfort was optimized away, and no one asked. When you search for something online, the results are sorted by an algorithm whose priorities are not yours. The system is being managed, but not for you and not with your knowledge.
At organizational scale: Your quarterly dashboard shows green across every performance metric. In the same week, your operations lead (the person who knows what the numbers actually mean) updates her LinkedIn and starts taking calls from recruiters. She hasn’t raised a concern. She stopped raising concerns six months ago, around the time the last one was received with silence. The dashboard hasn’t changed. But the person who could tell you what the dashboard is missing is already halfway out the door.
That silence, the moment when the people who know stop telling the people who decide, is Management at its most dangerous. Not because anyone lied. Because the channel between reality and authority quietly closed.
In the East Palestine analysis, the temperature data that could have predicted the railcar failure existed in the monitoring system. But the information flow filtered it away from the person making the decision about how to respond. The instrument had the reading. The decision-maker never saw it. Read the analysis →
Absence
What knowledge has walked away?
Absence appears wherever critical institutional knowledge is concentrated in a small number of people — undocumented, untransferable, and irreplaceable — creating a hidden dependency that becomes visible only when those people leave.
Every organization carries knowledge that isn’t written down: the institutional memory of how things actually work, the undocumented expertise that keeps complex operations running, the relationships and contextual understanding that can’t be captured in a procedures manual. When that knowledge is distributed across many people, the departure of any one individual is manageable. When it’s concentrated in a few, the organization is one retirement away from discovering that someone who left was quietly holding something essential together.
The signature of Absence is concentration: how much critical knowledge lives in how few heads. When concentration is high, the organization’s capability depends on specific individuals staying, on what they carry that no one else has, regardless of their titles or documented responsibilities.
At infrastructure scale: The experienced water treatment operator retired. His replacement has a dashboard. When the dashboard fails, there is no manual override, because no one was trained on the manual process. The average age of power line technicians, water system operators, and diesel mechanics continues to climb. The pipeline of people behind them narrows every year. The skill of running these systems manually is disappearing because the digital systems have been reliable enough that no one practiced the fallback.
At organizational scale: Your three longest-tenured engineers carry the institutional knowledge of why things are built the way they are. Not the documentation of what was built, but the reasoning behind the decisions. One of them has been mentioning retirement for two years. Everyone nods. No one has sat down with her to map what she knows, because what she knows isn’t a system or a process. It’s thirty years of judgment about which shortcuts are safe and which ones aren’t. When she leaves, the documentation will still describe the architecture. It won’t describe why certain decisions were made, which alternatives were tried and failed, or which components are fragile in ways that only show up under load. The team will discover those gaps one emergency at a time.
That slow discovery, learning what someone knew by finding out what you can no longer do, is how Absence announces itself.
In the Drug Shortage analysis, an entire generation of quality-manufacturing expertise was economically rationalized out of the generic pharmaceutical supply chain. The knowledge didn’t walk out the door in a retirement; it was never replaced because the economics didn’t justify replacing it. The absence became visible only when quality failures revealed that no one remaining understood the manufacturing processes well enough to prevent them. Read the analysis →
How They Connect
An organization’s supply chain is running lean — not crisis-level, but thin enough that a disruption would require fast, expert rerouting to avoid delays. Separately, three senior logistics specialists are approaching retirement, and their supplier relationships and contingency knowledge have never been documented.
Assessed individually, each condition is moderate risk. Assessed together, they’re something different entirely. The supply chain disruption that would normally be recoverable becomes unrecoverable, because the only people who know how to reroute are the same people about to leave. The thinness made the knowledge-concentration problem catastrophic. The knowledge concentration made the thinness problem unfixable.
This is what most assessments miss. These four frequencies don’t operate independently. When two of them are both under stress, the combined effect isn’t additive; it’s compounding. Each one makes the other worse. At WeWork, the compounding locked the entire organization: governance concentration prevented the board from checking the CEO, the information environment prevented the valuation disconnect from being challenged, and the funding dependence on a single investor reinforced the CEO’s authority because that investor’s position depended on maintaining the growth narrative.
A strong frequency can also absorb stress that weaker ones create, and that compensatory load is its own kind of risk. The organization that looks stable because one area of strength is quietly absorbing pressure from two areas of weakness is one disruption away from discovering how much that strength was carrying.
The combination across all four frequencies creates a readable pattern. It reveals what the overall structural condition looks like and how it behaves under stress, a fundamentally different question from which individual areas are weakest. Standard assessments score dimensions, average them, and hand you a ranked list. This framework maps the connections: which pairs are compounding, which strengths are compensating, where the cascade runs if something gives way. The difference between a list of problems and a map of how they interact is the difference between knowing what’s wrong and understanding why it isn’t getting better.
SVB’s collapse was driven by the interaction between concentrated asset exposure and risk information that never reached decision-makers. Boeing’s 737 MAX failures were driven by restructured engineering authority compounding with information flow that couldn’t override production priorities. The CrowdStrike outage demonstrated how these patterns bridge scales: architectural decisions at the organizational level predetermined the blast radius at infrastructure level, pushing a single update simultaneously to 8.5 million devices across airlines, hospitals, and banking systems with no staged rollout and no customer override. For more on how AI is amplifying these interactions, see AI & The Four Frequencies.
What the Framework Reveals
The analytical vocabulary above describes the landscape: The Four Frequencies, compounding interactions, compensatory load. The framework also maps specific features within that landscape.
Governance gaps. Every post-mortem documents what went wrong. This framework measures the window between “fixable” and “feasible”: the period during which a failure is avoidable but the decision-making architecture cannot execute the fix. At SVB, the interest rate risk was visible and the hedging solution was straightforward. But the way information flowed through the organization prevented that risk assessment from reaching the people with authority to act. The fix was known. The path to implementing it was blocked. That gap, between seeing the problem and being able to address it, consistently marks the point where recoverable conditions become irreversible ones. In the six cases the framework has examined, governance gaps ranged from 4.5 months to more than two decades.
Recovery windows. Some vulnerabilities are still in territory where the right intervention can reverse the trajectory. Some are approaching a threshold, still addressable but with a narrowing window. And some have crossed into conditions that can’t be fully reversed; the organization can improve from that point forward, but the state before the crossing can’t be restored. In the SVB analysis, the window for portfolio rebalancing was open for roughly eighteen months before market conditions and deposit concentration made the position irreversible. In the Drug Shortage analysis, the window has been narrowing for two decades, and the governance gap that would need to close before intervention is even feasible remains open. The question is not whether things are bad. The question is whether the window to fix them is still open.
Cascade pathways and intervention sequence. When multiple frequencies are degraded, the order in which they’re addressed changes the total outcome. Fixing the wrong thing first can accelerate the cascade rather than interrupt it, because the first intervention changes the landscape that the second one operates within. A repair that deactivates a compounding interaction between two frequencies produces a different result than the same repair applied after the compounding has advanced further. The path forward is a sequence, and the sequence matters.
The evidence base. Every pattern described on this page is documented across verified citations from independent organizations across sectors. The evidence is searchable, filterable, and public. Explore the Evidence Library →
A note on boundaries. The framework reads the present. It identifies conditions that exist now and maps how they interact. It does not predict specific outcomes or timelines. Every published analysis includes an explicit section on where the framework’s explanatory power reaches its limits, because a methodology that won’t name its own boundaries isn’t one worth trusting. The claim is that in every failure examined so far, these four patterns account for the mechanics, and that observation is testable against the evidence.