Why Companies Die and Cities Don't: What HR Gets Wrong About Human Capital

Apple turned 50 this year. That fact alone is worth pausing on. The average publicly traded company in the United States does not survive a decade. Apple has survived five of them, through multiple near-deaths, radical product pivots, and at least two complete reinventions of its business model.

How? The standard answer is leadership, vision, and culture. Those explanations are not wrong, but they are incomplete. They describe the outcome without explaining the structural mechanism that made the outcome possible.

This article argues that Apple's longevity is best understood not as a leadership story, but as a structural anomaly rooted in how the company treats its people. And it argues that most organizations, through their standard HR practices, are systematically building the conditions for their own decline.

The evidence for this comes from an unexpected source: physics and evolutionary biology.

Companies Scale Like Organisms. Cities Scale Like Ecosystems.

When biological, city, and company data are plotted against size, consistent mathematical patterns emerge across all three (West, 2017).

For biological organisms, most metabolic and physiological quantities scale sublinearly with body mass. A creature twice the size of another needs roughly 75 percent as much energy per unit of body weight. Efficiency increases with size, but so does rigidity.

Cities scale differently. Across hundreds of cities in the United States, China, Japan, India, and Europe, socioeconomic output including wages, patents, new businesses, and creative production scales superlinearly with population. Double a city's population and you get more than double (roughly 115%) more economic and creative output. The bigger the city, the more productive each person inside it becomes.

Companies sit closer to organisms than to cities. Analysis of more than 25,000 publicly traded North American companies from 1950 to 2009 found a scaling exponent of approximately 0.9, meaning growth slows relative to size. The half-life of a publicly traded company is roughly 10.5 years, and mortality risk does not decline with age. A company operating for 50 years faces roughly the same annual probability of death as one operating for 5 (Daepp et al., 2015; West, 2017).

Cities are nearly impossible to kill because they preserve and recombine the diversity of human talent. They tolerate redundancy, support experimentation, and allow failing functions to be replaced without the whole system collapsing. Companies optimize existing processes, narrow their functional scope, and progressively reduce internal diversity. The result is that companies behave like aging organisms: increasingly rigid, decreasingly adaptive, and ultimately mortal.

Slime Mold Can Solve Exponentially Complicated Problems in Linear Time https://www.sci.news/biology/slime-mold-problems-linear-time-06759.html

The Specialization Trap: What Biology Tells Us About Organizational Fragility

Evolutionary biology offers a precise mechanism for what West describes mathematically. Species that evolve narrow, highly efficient adaptations to a specific environment perform exceptionally well in stable conditions, but face disproportionate extinction risk when conditions shift (Simpson, 1944).

Species with narrow geographic ranges suffer approximately 70 percent extinction rates during major disruptions, while widely distributed generalist species suffer closer to 20 percent. Specialist species face roughly three times the extinction risk of cosmopolitan ones, and the pattern holds across taxa and extinction events (Jablonski, 2008; Colles et al., 2009).

The physical principle underlying this is the principle of least action (Castillo & Vera-Cruz, 2011; Kaila & Annila, 2008). Natural systems tend toward local energy minima, the path of least resistance. Specialization is precisely this process applied to organisms: highly efficient, locally optimal, but globally fragile. Once a system has descended into a narrow energy minimum, it cannot easily climb out when the landscape changes.

Organizations follow the same pattern. The same capabilities that drive success progressively filter out variation and constrain exploration (Leonard-Barton, 1992). Organizations that over-invest in exploiting existing capabilities at the expense of exploring new ones paint themselves into increasingly narrow corners (March, 1991). This failure mode is most acute in successful companies, precisely because their systems are optimized to protect what already works (Christensen, 1997).

Two Logics: Market Logic and Home Logic

My own research offers a framework for understanding why this pattern persists despite its costs. Drawing on a cross-disciplinary analysis of more than 72,000 scholarly articles using deep learning and natural language processing (NLP), I identified two dominant institutional logics that shape how organizations think about people and value creation (Dunn, 2025).

The first is market logic. This is the dominant logic in most business and economics frameworks. It treats relationships as competitive and transactional, assumes that opportunism is a baseline feature of human behavior, and evaluates people primarily as inputs to production. Efficiency is the core metric. In this logic, the ideal employee is a well-specified component: hire to a job description, deploy against a defined function, replace when the function changes.

The second is what I call home logic. This logic, which appears consistently in anthropological and sociological research, treats organizations more like households or communities. The goal is not to extract maximum output from each component, but to sustain the system's capacity to function over time. People are treated as participants with developing capabilities, not interchangeable inputs. Value creation is understood as cumulative and relational, not transactional.

These two logics map directly onto West's distinction between companies and cities. Market logic produces organizational behavior that resembles organism biology: efficient, specialized, and ultimately finite. Home logic produces behavior that resembles city dynamics: recombinant, generative, and more durable.

The irony is that most organizations publicly claim home logic while operationally running market logic. Leaders say things like "our people are our greatest asset" and "we're a family here." But when the market shifts or a quarter disappoints, their systems respond with workforce reductions, role eliminations, and hiring freezes, all of which are market logic decisions dressed in home logic language. My research identifies this gap as a structural source of organizational risk: the language signals one thing to employees, stakeholders, and the public, while the systems deliver another.

What Standard HR Practice Actually Optimizes For

Most HR systems are built around the job description as the fundamental unit of organization. The process is familiar: define a role, write a list of required qualifications, post the opening, screen for credential match, hire to spec.

This is efficient. It is also, from a systems perspective, a specialization machine.

Each hire sourced to a static job description narrows the organization's internal diversity. Each performance system built around role-specific outputs reinforces existing functions rather than developing transferable capacity. Each workforce reduction triggered by a role becoming obsolete discards accumulated knowledge and capability rather than redeploying it.

The implicit assumption underneath all of this is that people are interchangeable inputs relative to a fixed set of functions. If you need a function performed, you find a person with the matching credential. If the function disappears, you remove the part/person. If the market shifts, you rehire with a new specification.

Corning Incorporated's experience during the fiber optic collapse of 2001 and 2002 illustrates the downstream cost of this logic. The company had built a factory and developed highly specialized equipment to manufacture fiber optic cable at a massive scale. When the market collapsed, the equipment had no transferable value. Corning's stock dropped from roughly $113 per share at its 2000 peak to under $2 by mid-2002. The specialized assets, built for one context, were simply destroyed.

Human capital treated as specialized assets follows the same logic. Organizations that hire narrowly, develop people only within defined role boundaries, and make no investment in broader capability building are accumulating a form of organizational fragility that does not show up on a balance sheet until a disruption forces it to.

Why Cities Don't Die: The Fungibility Argument

West's answer to the city-versus-company asymmetry centers on what he calls the expanding dimensionality of cities. As cities grow, the space of possible functions, occupations, and recombinations grows faster than the population. A person who loses a job in a declining industry does not disappear from the city's productive capacity. They carry skills, relationships, and knowledge that can be recombined into new configurations as the economy shifts.

Urban economic development is fundamentally about diversification and recombination. New work grows out of old work not by replacing it wholesale, but by adding divisions of labor and activities onto existing foundations (Jacobs, 1969). The substrate of human capability remains, and new structures grow from it.

In my framework, this is home logic operating at scale. The city does not ask whether a person matches a current job description. It asks what a person can do, and it provides a sufficiently diverse environment that most people can find productive combinations for their capabilities, even as specific industries rise and fall.

The organizational implication is direct. Companies that treat human capital as fungible, meaning capable of being recombined and redeployed rather than consumed and replaced, preserve more of the adaptive capacity that allows cities to survive context change. Companies that treat human capital as specialized components do the opposite. They build efficiently for today's environment and sacrifice resilience for tomorrow's.

Apple as Structural Anomaly: A Test of the Framework

In a 2021 analysis of more than 31,000 companies over 70 years, Apple appeared as a confirmed over-performer relative to the predicted sublinear scaling trajectory (Zhang, Kempes, Hamilton, and West, 2021).

The structural reason, viewed through this framework, is that Apple has repeatedly behaved more like a city than a company in how it treats internal human capital. The engineers who built the Macintosh contributed to the iPod. The iPod team fed into the iPhone. The iPhone platform supported the iPad, the Watch, and Apple Silicon. Each transition involved redeploying accumulated capability into new configurations rather than discarding it and rehiring to a new specification.

This is precisely what home logic looks like at an organizational scale: investment in people's developing capacity rather than consumption of their current credential match.

West's theory predicts, however, that even Apple faces the gravitational pull of sublinear scaling over time. As bureaucratic overhead grows and systems optimize for existing products, the conditions that enabled repeated reinvention become harder to sustain. Apple's longevity confirms the argument by showing what it takes to escape the normal pattern, not by suggesting the pattern is easy to escape.

What This Means for Leaders

None of this is an argument against hiring standards, role clarity, or performance accountability. Organizations need structure. The argument is about where the structural logic starts.

Starting from a static job description and asking "who fits this slot?" is market logic. Starting from a person's developing capabilities and asking "how can this person contribute as the organization grows and changes?" is home logic. The first builds toward fragility. The second builds toward resilience.

In practice, this means several things.

Hire for capacity, not just current credential match. A resume is a record of past value: what a person has done, in contexts that may no longer exist. A good hire decision asks about future value: what a person can do, can learn, and can contribute as the organization adapts to changing conditions. Skills and knowledge inventories, broader than a job description allows, give organizations better information about the adaptive capacity they are accumulating rather than just the functions they can fill today.

Invest in development as a strategic decision, not a benefit. Organizations that invest in developing people's capabilities beyond their current roles are widening their internal possibility space. This is not philanthropy. It is the organizational equivalent of what cities do when they support diverse industries: creating conditions for recombination when the context changes.

Treat workforce transitions as redeployment problems before treating them as elimination decisions. When a role becomes obsolete, the first question should be what the person in that role can contribute in a different configuration. Most organizations skip that question because their HR systems are not built to answer it, not because redeployment is inherently more difficult than the alternative. It is also what distinguishes organizations that accumulate institutional knowledge from those that keep discarding and rebuilding it.

Audit the gap between your stated values and your operational logic. If your organization says people are its greatest asset but its systems respond to market pressure by eliminating people first and developing people last, that gap is not just a culture problem. It is a structural signal that market logic is running the organization regardless of what the values statement says.

The companies that survive long enough to matter are not the ones that optimize most efficiently for today's environment. They are the ones that maintain enough internal diversity, adaptive capacity, and human capital depth to recombine when the environment changes. That is what West's physics describes, what evolutionary biology predicts, and what the history of Apple, as an outlier, illustrates.

Conclusion

West asked why cities persist while companies do not. The answer is not that cities have better leaders. It is that cities are structured to preserve and recombine the fungible capacity of their human populations, while companies are structured to optimize that capacity for current conditions and discard it when conditions change.

Standard HR practice, built around static job descriptions and credential matching, accelerates the company pattern rather than countering it. It converts potentially fungible human capital into specialized components, narrows internal diversity, and leaves organizations progressively less able to adapt when the environment shifts.

The alternative requires a different starting assumption: that people are not components to be sourced but developing systems to be invested in. That is not a soft idea. It is a structural one. And the evidence from physics, evolutionary biology, and organizational research suggests it is the structural logic most consistent with long-term organizational survival.

References

Castillo, L. F. del, & Vera-Cruz, P. (2011). Thermodynamic Formulation of Living Systems and Their Evolution. Journal of Modern Physics, 02(05), Article 05.

Christensen, C. M. (1997). The innovator's dilemma: When new technologies cause great firms to fail. Harvard Business School Press.

Colles, A., Liow, L. H., & Prinzing, A. (2009). Specialists at risk under environmental change? Neoecological, paleoecological and phylogenetic approaches. Ecology Letters, 12(8), 849-863.

Daepp, M. I. G., Hamilton, M. J., West, G. B., & Bettencourt, L. M. A. (2015). The mortality of companies. Journal of the Royal Society Interface, 12(106), 20150120.

Dunn, S. T. (2025). Deep learning, deeper insights: A multilevel exploration of 'value.' Doctoral Dissertation, University of Hawai'i at Manoa.

Jacobs, J. (1969). The economy of cities. Random House.

Jablonski, D. (2008). Extinction and the spatial dynamics of biodiversity. Proceedings of the National Academy of Sciences, 105(suppl. 1), 11528-11535.

Kaila, V. R. I., & Annila, A. (2008). Natural selection for least action. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 464(2099), 3055–3070.

Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in managing new product development. Strategic Management Journal, 13(S1), 111-125.

March, J. G. (1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71-87.

West, G. B. (2017). Scale: The universal laws of growth, innovation, sustainability, and the pace of life in organisms, cities, economies, and companies. Penguin Press.

Zhang, Z., Kempes, C. P., Hamilton, M. J., & West, G. B. (2021). Scaling laws and a general theory for the growth of companies. arXiv:2109.10379.