The complexity trap

This is article 1 in The Complexity Trap series.

Civilizations rarely become complex because their leaders decide that complexity is desirable. Complexity develops because it solves problems.

A growing population requires more reliable food production. More intensive agriculture requires irrigation, storage, transportation, and systems for allocating land and water. Infrastructure requires labor, engineering, administration, taxation, and political authority. Expanding trade requires standards, contracts, security, ports, roads, currencies, and institutions capable of resolving disputes. Larger territories require bureaucracies, communication networks, armies, and increasingly sophisticated systems of governance. As societies become wealthier, more urbanized, and more technologically capable, specialization increases and institutions multiply.

Each layer of complexity can create extraordinary advantages. A village cannot construct an aqueduct stretching across a region, maintain a professional army, coordinate trade across continents, support universities, operate sophisticated financial markets, or manage infrastructure serving millions of people. Complex societies can mobilize resources on scales that simpler societies cannot approach. They can generate knowledge, increase agricultural productivity, create technological innovation, protect populations, expand trade, and produce levels of prosperity that would otherwise be impossible.

This is why complexity should not be understood primarily as a weakness. It is one of humanity's most successful problem-solving strategies.

Yet the same process creates a paradox. Every new system requires resources to maintain it. Institutions need personnel and funding. Infrastructure must be repaired. Bureaucracies require administration. Armies must be supplied. Information must be collected and processed. Regulations require enforcement. Trade networks depend upon transportation, security, and political relationships. Technologies create new dependencies that themselves require additional technologies and institutions.

The solutions accumulate.

Eventually, a society can reach a point where maintaining the architecture created to solve previous problems consumes an increasing share of the resources available to solve new ones. Institutions may become more difficult to reform precisely because so many other systems depend upon them. Infrastructure can become simultaneously indispensable and increasingly expensive to maintain. Political structures can develop constituencies whose interests become tied to preserving arrangements that once generated prosperity but are no longer well suited to changing conditions.

The danger is not complexity itself. The danger emerges when the complexity of a system begins growing faster than its capacity to finance, understand, coordinate, and adapt that system.

This is the complexity trap.

It is an old problem, but it has extraordinary relevance to the modern world. Anthropology, archaeology, sociology, political science, economics, systems theory, and global development practice have approached it from different directions. Each discipline offers a different explanation for why societies prosper, stagnate, transform, or collapse. Taken together, they reveal something that conventional risk management often misses: some of the greatest vulnerabilities facing a system can originate not outside the system, but within the architecture that made the system successful.

Complexity works

Joseph Tainter's work on the collapse of complex societies provides one of the most useful starting points for understanding this process. His argument is frequently simplified into the claim that complexity causes collapse, but that misses the more important insight. Tainter's theory begins from the observation that societies invest in complexity because complexity produces returns.

A government creates an administrative institution because some problem requires coordination. Agricultural systems become more elaborate because greater production is needed. Territories expand because new resources, land, labor, or security can generate benefits. Information systems become more sophisticated because larger societies require greater knowledge about taxation, populations, resources, and threats.

Initially, these investments can produce substantial returns. Irrigation increases food production. Roads expand commerce. Political organization reduces conflict and enables collective action. Specialization allows people to become more productive. Taxation finances public works and defense. Education expands knowledge. Trade connects societies to resources they do not possess themselves.

The development of complexity is therefore not evidence that a society has made a mistake. Quite the opposite. It frequently represents evidence that a society has become extraordinarily good at solving problems.

The difficulty arises because the returns from additional complexity do not necessarily increase forever. Once the most productive agricultural land is already being cultivated, expanding production may require irrigation in increasingly difficult environments. Once the most valuable trade routes are established, expanding commerce may require longer and more expensive networks. Once easily defended borders have been secured, territorial expansion may require greater military expenditure for progressively smaller strategic gains.

Institutions can experience the same pattern. The first administrative structures may dramatically improve governance. Later layers of administration may produce smaller improvements while requiring more resources. Eventually, additional complexity may become necessary simply to coordinate the complexity that already exists.

Tainter describes this as declining marginal returns to complexity. The society continues investing because the problems have not disappeared, but each additional investment produces less benefit relative to its cost.

This distinction matters enormously. Societies do not generally wake up one morning and choose between simplicity and complexity. They inherit systems created by previous generations. Those systems have infrastructure, institutions, political interests, cultural expectations, and economic dependencies attached to them. Abandoning complexity can itself be extraordinarily expensive.

A large city cannot simply decide that its water system has become too complicated. An empire cannot instantly abandon its borders without political and security consequences. A population dependent upon intensive agriculture cannot easily return to less productive farming systems. Once complexity supports millions of lives and livelihoods, maintaining it becomes a necessity.

Success can therefore create path dependence. The solutions of one generation become the operating requirements of the next.

Rome and the cost of maintaining success

Few historical cases illustrate the paradox better than Rome.

Roman power was built through complexity. Roads, ports, aqueducts, taxation, law, military organization, provincial administration, urban infrastructure, agricultural production, and trade allowed Rome to coordinate an enormous territory. Its institutions enabled resources to move across extraordinary distances. Grain could support cities far from the fields where it was produced. Taxes collected in the provinces could finance armies stationed on distant frontiers. Roads designed partly for military movement also facilitated trade and administration.

The Roman system did not become vulnerable because these institutions were useless. They were the source of Roman power.

But scale creates obligations as well as advantages. Borders must be defended. Roads require maintenance. Armies require salaries, food, equipment, and logistics. Administrative systems require personnel. Political legitimacy must be maintained across culturally and geographically diverse populations. Urban populations depend upon continuing flows of food, water, and economic activity.

As the empire matured, maintaining the existing system demanded substantial resources even before new challenges appeared. External military pressure increased, political instability periodically disrupted governance, fiscal pressures intensified, and disease and climatic variability added further stresses. Historians continue debating the relative importance of these forces, and no serious explanation of Rome's transformation can be reduced to a single variable.

That is precisely what makes Rome useful for thinking about systemic risk.

The question is not which individual shock "caused" Rome to fall. The more revealing question is why a system capable of surviving enormous challenges for centuries eventually became less capable of absorbing additional ones.

Complexity changes the consequences of shocks because systems become dependent upon one another. Fiscal weakness affects military capacity. Military demands affect taxation. Political instability affects administration. Disease affects population and revenues. Trade disruptions affect cities and state finances. External threats require resources that may already be committed elsewhere.

A resilient system can absorb several of these pressures without fundamental transformation. A system operating closer to its institutional, fiscal, or logistical limits has less room for error.

The Roman experience therefore suggests a principle that extends far beyond antiquity: the capacity of a system cannot be measured only by what it can accomplish under favorable conditions. It must also be measured by how much additional stress it can absorb while continuing to perform its essential functions.

This is the beginning of the connection between complexity and risk management.

The Maya and the danger of single explanations

The Classic Maya provide a different but equally important lesson.

Popular explanations of societal collapse often search for a single cause. Drought becomes the explanation. Or warfare. Or environmental degradation. Or political instability. Yet archaeological evidence increasingly demonstrates how misleading these categories can become when separated from one another.

Maya societies developed sophisticated political institutions, agricultural systems, trade networks, water-management systems, monumental architecture, religious institutions, and highly organized urban centers. These systems supported substantial populations in environments that could be challenging, particularly where rainfall was seasonal and surface water limited.

That sophistication created capacity, but it also created dependencies. Large populations required reliable agricultural production. Political authority depended partly upon the ability of elites and institutions to maintain social order and manage resources. Cities depended upon surrounding landscapes and networks of exchange. Competition among polities required resources, alliances, and military capacity.

Periods of drought therefore did not act upon empty landscapes. They acted upon already complex political and economic systems.

This is why drought alone cannot explain the Maya transformations. Some political centers declined before the most severe climatic conditions. Different regions experienced different trajectories, while some northern centers remained prosperous after major southern centers had declined. Warfare, political fragmentation, environmental degradation, demographic pressures, elite competition, and climatic stress interacted differently across time and geography.

The important analytical shift is from asking whether climate caused collapse to asking how environmental pressure interacted with the structure of society.

A drought affecting a small, mobile population with diverse subsistence strategies is fundamentally different from the same drought affecting a large urban population dependent upon intensive agriculture, political institutions, water infrastructure, and established patterns of settlement. The physical hazard may be similar. The social consequences are not.

This is a foundational principle of modern risk analysis as well as global development practice: hazards do not automatically become disasters. Vulnerability determines how shocks are transmitted through societies.

The Maya case also reminds us that collapse does not necessarily mean disappearance. Political centers can fail while populations continue living, trading, adapting, and reorganizing. Institutions may fragment while other forms of social organization persist. A state can collapse without a culture vanishing.

This matters because our language often imposes an artificial binary between success and collapse. History is usually more complicated. Societies decentralize, reorganize, migrate, lose political complexity, develop new institutions, and sometimes recover under different forms.

From the perspective of people living through these transitions, however, the distinction is hardly academic. Political transformation can involve violence, displacement, hunger, inequality, and enormous losses of accumulated capability. The fact that society continues in another form does not mean institutional failure is inconsequential.

The more useful question is therefore not whether societies survive. Human beings are remarkably capable of surviving. The question is what capabilities, institutions, knowledge, infrastructure, and standards of living survive with them.

The first globalized systems

The Late Bronze Age eastern Mediterranean provides perhaps the most striking ancient parallel to modern concerns about interconnected systems.

By the second millennium BCE, societies across the eastern Mediterranean and Near East were connected through extensive networks of diplomacy, trade, warfare, migration, and resource exchange. Mycenaean kingdoms, the Hittite Empire, Egypt, Ugarit, Cyprus, and other political centers participated in networks moving metals, agricultural commodities, luxury goods, technologies, information, and people across considerable distances.

Bronze itself embodied interdependence. Producing it required access to copper and tin, resources whose geographic distribution encouraged long-distance exchange. Political elites developed diplomatic relationships and commercial systems capable of securing materials that were unavailable locally. Specialists, merchants, sailors, soldiers, administrators, and rulers became participants in a network whose prosperity depended partly upon continued connectivity.

It worked extraordinarily well.

Connectivity expanded access to resources, supported specialization, increased wealth, facilitated technological exchange, and allowed political centers to participate in economic systems extending far beyond their immediate territories. In some respects, the Late Bronze Age Mediterranean demonstrates the same principle that later made modern globalization so powerful: societies do not need to possess every resource themselves if networks allow them to obtain what they need from elsewhere.

But connectivity changes the architecture of risk.

A society dependent upon external metals becomes vulnerable to interruptions in metal supplies. Political centers dependent upon trade become exposed to instability occurring along trade routes. Specialized systems can become vulnerable when the inputs supporting specialization disappear. Political alliances can transmit conflict just as commercial networks transmit goods.

Around the end of the Late Bronze Age, many eastern Mediterranean political systems experienced severe disruption. Palatial centers were destroyed or abandoned, trade networks changed, political structures disappeared, and several major powers weakened or fragmented. Explanations include drought, famine, warfare, migration, internal political instability, earthquakes, and disruptions to trade.

Again, searching for a single cause misses the point.

The significance of the Late Bronze Age is precisely that multiple pressures appear to have interacted within an interconnected system. A drought does not need to destroy a civilization directly if it reduces agricultural production, increases food prices, weakens state revenues, encourages migration, intensifies political conflict, and reduces the government's capacity to defend trade routes. Warfare does not need to destroy every city if it interrupts the networks upon which specialized economies depend.

Interdependence allows prosperity to travel farther.

It also allows disruption to travel farther.

This is one of the most important lessons ancient societies offer the modern world. Networks are neither inherently resilient nor inherently fragile. Their effect depends upon their structure. A diversified network can provide alternatives when one region fails. A tightly concentrated network can transmit disruption throughout the system.

The question is therefore not whether a society is interconnected. It is whether its interconnections create options or dependencies.

Collapse is not an event

Archaeology has increasingly complicated the popular image of societal collapse because the archaeological record rarely provides the clean endings imagined in modern narratives.

Empires fragment. Cities decline while others expand. Political institutions disappear while religious practices continue. Trade routes shift. Populations migrate. Technologies survive. Local communities reorganize. New political structures emerge from the remains of older ones.

What looks like collapse when viewed from the palace can look very different when viewed from the village.

This distinction is important because complex political systems often concentrate benefits and costs unevenly. Maintaining an empire may be extraordinarily valuable to elites, administrators, merchants, soldiers, or populations dependent upon imperial infrastructure while imposing substantial taxation or labor obligations on others. Political simplification can therefore represent catastrophic loss for some groups while creating greater autonomy for others.

Tainter's framework is especially interesting here because collapse can be understood as a reduction in sociopolitical complexity. From this perspective, collapse does not necessarily require extinction, demographic catastrophe, or the disappearance of culture. It can mean that the institutional architecture required to maintain a particular level of organization can no longer be sustained.

That definition allows us to move beyond dramatic images of abandoned cities.

A system can undergo meaningful collapse while many of its components continue functioning.

This is familiar in the modern world as well. A government can continue existing while losing administrative capacity. Infrastructure can remain physically present while becoming unreliable. A corporation can remain solvent while becoming incapable of innovation. An institution can continue producing rules and reports long after it has lost the ability to solve the problem it was created to address.

Failure is not always disappearance.

Sometimes failure is the gradual loss of adaptive capacity.

The problem of institutional accumulation

This brings us to an aspect of complexity that becomes particularly important when connecting anthropology with modern organizations.

Institutions rarely begin as pointless bureaucracy. They are generally created because someone needed to solve a problem.

A tax authority exists because governments require revenue. Environmental regulations exist because markets can impose costs on people who are not participating in a transaction. Financial regulations exist because banking failures can affect entire economies. Procurement systems exist because organizations need mechanisms for controlling expenditure and corruption. Reporting systems exist because leaders require information.

Each institution can therefore be rational in isolation.

The difficulty emerges through accumulation.

New problems generate new rules, departments, reporting requirements, technologies, oversight mechanisms, and professional specializations. Existing structures rarely disappear at the same rate. Over time, organizations accumulate layers representing solutions to problems encountered across different periods.

This creates what might be called institutional sediment. Yesterday's solution remains underneath today's solution, which becomes the foundation for tomorrow's.

The result can be an organization that possesses enormous capability but also increasing coordination costs. Information must move through more layers. Decisions require approval from more stakeholders. Specialized departments develop different vocabularies, incentives, and definitions of success. Leadership receives more information while potentially understanding less about the relationships among the systems generating it.

The organization becomes better at managing individual domains while becoming less capable of understanding the whole.

This is not simply bureaucracy. It is a systems problem.

And it appears at multiple scales. Corporations experience it. Governments experience it. International institutions experience it. Development programs experience it. Even risk-management systems can experience it when organizations respond to every newly identified risk by adding another control without examining how the controls interact.

Eventually, the architecture designed to reduce risk can become a source of risk itself.

Development and the accumulation of dependency

Global development practice adds another dimension because development is, in many respects, the deliberate creation of complexity.

Development expands infrastructure, institutions, markets, education, healthcare, financial systems, technology, urbanization, governance capacity, and economic specialization. These are precisely the systems that allow societies to improve human welfare. But every improvement also changes the structure of dependency.

A rural community using local energy sources may have limited productive capacity but little dependence upon a national electricity grid. Electrification dramatically improves economic opportunity, healthcare, education, communications, and quality of life, but it also means electricity failure now affects many more essential functions.

Irrigation can transform agricultural productivity but creates dependence upon water infrastructure, maintenance, energy, governance, and long-term hydrological conditions. Digital financial services can dramatically expand inclusion while creating dependence upon telecommunications, electricity, cybersecurity, identification systems, and financial institutions.

Urbanization creates enormous productivity advantages because people, businesses, infrastructure, knowledge, and capital become concentrated. Yet concentration also means failures can affect millions of people simultaneously.

None of this is an argument against development. It is an argument for understanding development as a transformation of risk rather than simply a reduction of it. Successful development eliminates certain vulnerabilities while creating others.

The objective therefore cannot be to avoid complexity. A society without complex infrastructure, healthcare, education, finance, communications, and governance would sacrifice many of the achievements that have dramatically improved human welfare.

The challenge is to build complexity without allowing dependency to outrun adaptive capacity.

When solutions become constraints

Across these historical cases, a common pattern begins to emerge.

Societies encounter problems. They develop institutions, technologies, infrastructure, and networks to solve them. Successful solutions allow populations and economies to grow. Growth creates new requirements, which produce additional complexity. The expanding system becomes increasingly capable, but it also becomes increasingly dependent upon the continued functioning of its components.

Eventually, the society may encounter a different class of problem.

The challenge is no longer simply solving an external threat. It is maintaining the enormous architecture of solutions accumulated over time while simultaneously adapting that architecture to new conditions.

This is where diminishing returns become dangerous.

Resources committed to maintaining existing systems cannot simultaneously be used to create new capabilities. Political institutions designed around established interests may resist reforms that threaten those interests. Infrastructure built for one environmental or economic regime may become increasingly expensive to operate under another. Specialization may reduce the ability to substitute when critical inputs disappear.

The system does not necessarily collapse. More commonly, its room for maneuver narrows.

That is perhaps the most useful way to think about the complexity trap. Vulnerability emerges when a system loses optionality. It becomes dependent upon maintaining arrangements that are increasingly difficult to finance or change.

A resilient complex system can reorganize. A fragile complex system must continue functioning in essentially the same way because too many other systems depend upon it.

What ancient societies can teach modern risk management

It would be intellectually careless to argue that Rome, the Maya, or Bronze Age kingdoms provide direct predictions for twenty-first-century civilization. Modern societies possess technologies, scientific knowledge, communications, financial systems, energy resources, and institutional capabilities that ancient societies could not have imagined.

But modernity has not eliminated the underlying systems problem. If anything, we have increased its scale.

Today's global economy contains supply chains spanning continents, financial transactions occurring in milliseconds, cities containing tens of millions of people, electricity grids serving entire regions, digital infrastructure connecting billions of individuals, agricultural systems dependent upon global fertilizer and energy markets, and technological industries dependent upon extraordinarily specialized components produced in a limited number of locations.

Our capacity to solve problems is historically unprecedented.

So is our capacity to create dependencies.

This changes the meaning of risk management. Traditional approaches often begin with external threats: What could happen to the organization? What geopolitical event, climate hazard, cyberattack, market disruption, or regulatory change could affect us?

Complexity requires an additional question:

What has the organization built that makes the consequences of those events greater than they would otherwise be?

That question shifts attention from hazards toward architecture.

A drought is a hazard. Dependence upon one water source is architecture. A geopolitical conflict is a hazard. Dependence upon one supplier located in the affected region is architecture.

A cyberattack is a hazard. Connecting every essential function through systems lacking adequate segmentation is architecture.

An economic downturn is a hazard. Operating with financial structures that cannot tolerate a temporary decline in revenue is architecture. The distinction is fundamental because organizations cannot control many hazards, but they can influence how those hazards propagate through their systems.

Ancient societies could not control drought, disease, migration, or external threats either. Their vulnerability depended partly upon the systems through which those pressures traveled. The same remains true today.

The complexity paradox

The lesson of history is therefore not that complexity inevitably leads to collapse. That conclusion would be both historically questionable and practically useless.

Complexity creates civilization. It allows societies to coordinate knowledge, specialize labor, construct infrastructure, develop technology, organize healthcare, expand education, facilitate trade, govern large populations, and respond collectively to challenges that individuals or small communities could never manage alone. The problem emerges when complexity becomes rigid.

A system becomes vulnerable when the cost of maintaining it consumes the resources required to adapt it; when institutions protect their own continuity more effectively than the functions they were created to perform; when specialization creates dependencies without alternatives; when information increases faster than decision-makers' ability to interpret it; and when every new problem produces another layer of complexity without reconsidering the architecture underneath.

Collapse, when it occurs, may therefore represent less a failure to solve problems than the culmination of generations of successful problem solving under conditions that have changed.

That distinction matters. Rome did not become powerful despite its complexity. It became powerful because of it. Mayan societies did not flourish despite sophisticated political, agricultural, and water-management systems. Those systems helped make large populations and urban centers possible.

Late Bronze Age societies did not become prosperous despite long-distance interdependence. Their networks created access to resources, knowledge, technologies, and markets that no individual polity possessed alone.

Their vulnerabilities emerged partly from the same structures that created their capabilities.

That is the paradox.

And it leaves us with a question considerably more important than whether modern civilization is destined to collapse. There is little value in attempting to force ancient histories into predictions of contemporary catastrophe.

The more useful question is whether we can recognize the point at which successful complexity begins reducing the capacity of a system to adapt.

That question belongs as much to global development, corporate strategy, public administration, and risk management as it does to archaeology.

Because modern societies are not becoming simpler. Artificial intelligence, electrification, global finance, digital infrastructure, urbanization, climate adaptation, biotechnology, energy transitions, and geopolitical competition will require new institutions, new infrastructure, new regulations, new technologies, and new forms of coordination.

We will respond to complexity with more complexity because, much of the time, that is precisely what successful societies do.

The challenge is whether the next layer solves more problems than it creates.

Part II of this series will examine that challenge through global development: how the very processes we associate with progress — urbanization, infrastructure, technology, industrialization, agricultural modernization, and institutional expansion — can eliminate old vulnerabilities while simultaneously creating new ones. Development remains essential. Complexity remains essential.

The question is whether we can learn to distinguish complexity that expands human capability from complexity that quietly consumes the capacity to adapt. History suggests that distinction may determine far more than whether a society becomes prosperous. It may determine whether prosperity can endure.

illuminem Voices is a democratic space presenting the opinions of leading Sustainability Thought Leaders, their views do not necessarily represent those of illuminem.

The world needs sustainability knowledge. At illuminem, no interest group or shareholder can influence our work. Thank you for supporting our mission to make high-quality and independent sustainability information free for all. Every contribution helps. Thank you for donating today.