Key Takeaways
Treat failure as the baseline, not an exception to explain away
Failure is ordinary. Nearly 99.99 percent of species that have ever existed are extinct, and the book reports that more than 10 percent of American companies disappear each year. The same pattern spans firms, species, and government policies: persistence is the exception, not a guarantee of merit or good management.
Survival is temporary. Failure is not just a final mishap. It is a recurring feature of systems where conditions and relationships change. Ormerod’s central puzzle is that businesses and governments can plan deliberately, unlike species, yet still fail on a similarly pervasive scale. The book asks what this reveals about the limits of prediction, and whether failure can sometimes help the larger system adapt.
What's striking is the change in the question. Instead of asking only how winners succeed, the book asks why success so often expires. That guards against survivorship bias, the tendency to study the firms still standing and mistake their history for a repeatable recipe. Failure rates, however, do not prove that every collapse is useful or unavoidable. A preventable accounting fraud and a firm displaced by a better technology are both failures, but they carry different lessons. Treating failure as normal makes diagnosis more serious, not less: it directs attention to what systems learn, lose, and replace when a component disappears.
A century of scale did not save the corporate giants
New forms took over. Around 1900, anonymous outside investment and limited liability helped companies grow far beyond the scale of earlier businesses. The merger wave from 1895 to 1905 absorbed roughly 3,012 firms. US Steel, with more than 200,000 employees by the early twentieth century, showed what this new organizational form could achieve. The expansion helped raise living standards, but size did not mean permanence.
The survivors were vulnerable. Of the 100 largest industrial companies in the world in 1912, 48 had disappeared as independent firms by 1995, and 29 had gone bankrupt. Only 28 were larger than in 1912. Across US corporate leaders from 1919 to 1979, just 33 of the original top 100 remained in the top 100. Today’s celebrated firms may be tomorrow’s footnotes.
The Edwardian corporate boom makes institutional design visible: legal rules can unlock investment and experimentation, while leaving the resulting firms exposed to change. This is a useful correction to the idea that longevity follows automatically from scale, brand recognition, or accumulated expertise. Still, disappearance is not always equivalent to failure. A merger can end a company’s independent identity while preserving much of its productive activity. The relevant measure therefore depends on the question: bankruptcy, loss of market position, and absorption are different outcomes. The book’s evidence is strongest as a warning against treating today’s corporate leaders as permanent fixtures.
Do not mistake a tidy equilibrium for a forecast of real life
Textbook certainty has a price. Standard economics often describes firms as choosing the output where price equals marginal cost, the cost of producing one additional unit. That rule assumes firms know how demand changes at different prices and how their costs change with output. In practice, the future demand and competitors’ reactions are uncertain, and production costs can change as firms learn or invest.
The path matters. Equilibrium models describe a settled outcome but can say little about how long it takes to get there or what happens along the way. In one growth-model analysis cited by Ormerod, adjustment to a new equilibrium could take more than a century. Meanwhile, a study of business pricing found that firms commonly used cost-plus rules rather than calculating the textbook optimum. Real businesses operate amid change, not at rest.
Equilibrium models can still be useful as maps of incentives or as benchmarks for comparison. The danger comes when a benchmark is mistaken for a practical description of decision-making. Forecasting research and engineering both distinguish a model’s internal elegance from its performance under changing conditions. Ormerod is persuasive that uncertainty, learning, and strategic response belong inside economic explanations, rather than being dismissed as inconvenient exceptions. Yet simple pricing rules also have limits: a cost-plus markup can ignore demand shifts and competitive threats. The practical lesson is not to discard theory, but to ask which assumptions hold in the decision at hand.
Treat social policy as experimentation, not an engineering blueprint
Good intentions do not ensure control. The book compares British unemployment before and after the Second World War. The public sector became more than twice as large relative to the economy, yet average unemployment was similar: 4.9 percent from 1870 to 1938, and 4.5 percent from 1946 onward. This does not establish that public spending had no benefits. It does show that larger government did not automatically deliver lower unemployment.
Social outcomes resist simple stories. The book also describes declining social mobility in Britain and the United States, persistent residential segregation, and changing income inequality. Meanwhile, global inequality fell in the late twentieth century as Asian economies grew, complicating claims that worldwide inequality only rises. Ormerod argues that these patterns are difficult to predict because many interacting influences obscure any reliable signal.
The strongest policy implication is epistemic humility: ambitious goals need feedback, revision, and evidence of actual effects. That aligns with adaptive management, an approach in which institutions treat interventions as hypotheses and adjust them as results emerge. The comparisons in the book are informative but cannot, by themselves, isolate the causes of unemployment or mobility. A before-and-after comparison may conceal changes in demographics, technology, or labor markets. Nor does difficulty forecasting mean policy is pointless. It means that measurable aims, comparison groups, and willingness to revise are safer than assuming a large intervention will produce its intended result.
Small preferences can create stark social divisions
Local choices can scale up. Thomas Schelling’s model begins with two kinds of residents scattered across a grid, with a few empty homes. Each resident is willing to live among many people of the other type, moving only when their immediate neighborhood feels too unlike them. Despite this mild threshold, repeated individual moves produce sharply separated clusters.
The group outcome surprises. No resident needs to want a segregated city. Each person responds only to nearby conditions, but one move can change a neighbor’s choices, which can trigger further moves elsewhere. The model helps explain why strong segregation can coexist with weak individual preferences for separation. It also shows why complete knowledge of each person’s simple rule does not reveal the exact final pattern: interactions generate outcomes that no single resident designed.
This is a vivid example of emergence: a system-level pattern arising from many local decisions, without a central plan. Similar reasoning appears in traffic flow, where individually sensible lane changes can create a jam, and in ecology, where small local interactions shape population patterns. The model’s clarity is its strength, but its limits matter. Real neighborhoods involve unequal wealth, housing supply, discrimination, and unequal ability to move. A simple preference model cannot explain those structural forces by itself. Its key contribution is narrower and valuable: strong segregation does not require every individual to hold strong segregationist preferences.
Choose robust moves, not imaginary optimal strategies
Information is not omniscience. Conventional economic models often assume that people can gather and process everything relevant, then select the best option. Ormerod contrasts this with bounded rationality: decision-makers have limited information or limited capacity to calculate what it implies. A firm may know its current costs and customers yet still be unable to anticipate competitors or future demand.
Even clear games resist mastery. In the Prisoner’s Dilemma, two players each do better individually by betraying the other, although mutual cooperation would leave both better off. In repeated versions, the best strategy depends on how long play continues and how behavior changes. Chess has fixed rules and visible moves, yet even elite players cannot calculate every possible line. In complex situations, people rely on workable rules of thumb, not guaranteed optima.
The distinction between uncertainty and calculable risk is crucial. Risk has known probabilities; uncertainty means the probabilities themselves are unclear. Many real decisions are closer to the second condition, so optimization can imply more knowledge than decision-makers possess. Robust strategies, which avoid catastrophic outcomes across several plausible futures, can be more useful than a supposedly perfect choice based on fragile assumptions. Still, heuristics are not automatically wise: a shortcut can preserve bias or miss a new threat. The valuable practice is to make assumptions explicit, use simple rules where appropriate, and remain prepared to revise them.
A tiny shock can trigger a system-wide collapse
Extinction has a pattern. Ormerod describes a power law, a statistical relationship in which large events occur less often than small ones, but their frequency declines more slowly than a normal bell-shaped model would suggest. In the fossil record analyzed in the book, an extinction event twice as large is about four times less frequent. Similar patterns appear in the data on firm failures.
Financial models missed the tail. In 1998, Long-Term Capital Management lost about $4.6 billion, and a rescue was organized to reduce the danger of wider financial collapse. The fund’s models treated very large market movements as far rarer than they proved to be. The 1987 stock-market crash also showed that enormous consequences need not have an obvious, equally enormous cause. Small shocks can spread through connected systems.
Power-law thinking is valuable in risk management because it resists the comforting assumption that extreme events are practically impossible. A system can look stable for years while remaining exposed to rare cascades. But a power law describes a pattern in event sizes; it does not identify the cause of a particular crash or predict its date. The book’s historical comparisons make a compelling case for caution, though estimates of scaling relationships can depend on data and measurement choices. The practical implication is to stress-test plans against outsized losses, rather than treating past calm as proof that the tails are harmless.
Track how agents affect one another, not only their individual choices
Connections shape outcomes. The book contrasts external shocks, such as a meteorite impact, with endogenous failure, which arises from interactions inside a system. In a model where species or firms are connected, each agent’s strategy can help or harm others. Some relationships are cooperative, others competitive, and many systems combine both. When these connections change, a local shift can spread beyond its starting point.
The pattern tests the explanation. A model based only on external shocks can reproduce biological extinction patterns, but it struggles to match firm failures when businesses can change strategy quickly. Models that include changing interactions do better at reproducing the observed waves of failure and the relationship between event size and frequency. The point is not that outside shocks never matter. It is that internal dependencies can make a system fragile even without a dramatic external trigger.
Network analysis shifts attention from isolated actors to dependencies: who supplies whom, which failures propagate, and where cooperation sustains the whole. This perspective is useful in supply chains, public health, and financial oversight, where a weakly visible connection can become a transmission route for disruption. It also suggests that resilience may depend on diversity and alternative pathways, not simply on making every component stronger. A caution is necessary: mapping connections does not by itself explain the system. The direction and strength of links, and how they change over time, must also be established from evidence.
Let innovation earn temporary power, then keep entry open
Competition can mean two things. Schumpeter argued that the most important pressure on a firm may come not from a near-identical rival, but from a new product, technology, or way of organizing business. A successful innovation can create temporary market power and profits, which reward experimentation. The book points to aluminum: its price fell dramatically as production and output expanded, despite one company’s strong position.
The system needs both pressure and room. Ormerod’s model suggests that a demanding external environment can raise the fitness of the system, while forcing ever more competition among firms inside an industry can weaken it. The implication is not to protect incumbents indefinitely. It is to allow cooperation and experimentation where useful, while preserving the possibility that new entrants and ideas can displace established firms.
This argument offers a more dynamic view of competition than the textbook picture of many identical sellers. Temporary rewards can finance research and make risky innovation worthwhile, while the credible threat of entry pressures dominant firms to keep improving. The tension is real: market power can support innovation, but it can also be used to block challengers or extract rents without improving products. The best policy question is therefore not simply whether a firm is large, but whether the surrounding rules preserve contestability, experimentation, and consumer choice. The book’s framework is strongest when it distinguishes a competitive environment from competition as an end in itself.
Analysis
Why Most Things Fail is a sustained challenge to the idea that social and economic systems behave like machines whose components follow known rules toward predictable outcomes. Its argument advances through a sequence of increasingly difficult examples: corporate lifetimes, policy outcomes, residential segregation, strategic games, biological extinction, and finally models of networked failure. The recurring lesson is that intention at the level of an individual does not determine the outcome at the level of the system.
The book’s most durable contribution is not simply the claim that failure is common. It is the method of explanation: test theories against observed patterns, distinguish local decisions from system outcomes, and treat uncertainty as more than a missing spreadsheet cell. Schelling’s segregation model and the comparison of firm and species extinctions make abstract complexity graspable. The historical stories, from the rise of limited-liability corporations to the collapse of Long-Term Capital Management, give the argument narrative force.
The central empirical result, that firm and biological extinctions display similar size-frequency patterns, is suggestive rather than a complete causal explanation. Similar statistical distributions can arise from different mechanisms, a point the book itself demonstrates by presenting competing models. The network account gains plausibility because it fits firms’ capacity to change strategies, but a model’s ability to reproduce a pattern does not prove that its mechanism dominates in every industry or period. The book’s policy conclusions also require care: showing that detailed planning often misses its target does not establish that public action cannot reduce harms or expand opportunity.
Read most productively, the book is not a case for fatalism or laissez-faire reflex. It is a warning against false precision. Use forecasts as provisional, design institutions that can adapt, watch for dependencies that transmit shocks, and avoid confusing a surviving champion with a guaranteed formula for success. Failure is not proof that effort is futile. It is evidence that systems must be judged by how they respond to change, not by how confidently they describe a desired equilibrium.
Review Summary
Why Most Things Fail explores why failure is common in economics, business, and biology. Ormerod argues that traditional economic models are flawed and that failure is often random and unpredictable in complex systems. He draws parallels between species extinctions and business failures, emphasizing the role of competition and innovation. While some readers found the book insightful, others criticized its structure and repetitiveness. The book challenges conventional economic thinking and promotes a more evolutionary view of markets and economies.
People Also Read
Glossary
Iron Law of Failure
pervasive organizational failureOrmerod’s label for the recurring disappearance of species, firms, and policies, despite the efforts of organisms or decision-makers to survive and succeed. It emphasizes failure as a widespread feature of evolving systems, not as a rare anomaly. The term frames the book’s central question: why purposeful human organizations fail at rates that resemble biological extinction.
Power law
size-frequency relationshipA mathematical relationship in which the frequency of an event falls as its size increases, but not as rapidly as in a normal, bell-shaped distribution. Ormerod uses it to describe extinction data for biological species and firms, and discusses similar patterns in financial markets. It can describe the relative frequency of event sizes, but it does not by itself explain their causes or predict when a large event will occur.
Schelling model
local choices, global segregationA model of residential sorting in which two types of residents occupy a grid and move when too few nearby residents are like themselves. Each person can tolerate a substantial number of neighbors from the other group, yet local moves produce pronounced segregation across the whole grid. The model demonstrates how system-level patterns can arise without strong individual preferences or a coordinated plan.
Prisoner's Dilemma
cooperation versus self-interestA strategic game in which each of two players can cooperate or defect. Each player has an incentive to defect regardless of the other player’s choice, even though both would receive a better joint outcome if they cooperated. The game illustrates a conflict between individual and collective interests. In repeated, uncertain settings, the best strategy depends on the rules and duration of interaction.
Bounded rationality
decision-making with limited knowledgeA condition in which decision-makers lack complete information, or cannot fully process the information available to them. Ormerod uses the term to contrast real people and firms with economic models that assume perfect knowledge and effortless calculation. Decisions may still be purposeful, but they are not guaranteed to be optimal, especially when outcomes depend on other agents and future conditions.
Exogenous extinction
failure from outside shocksExtinction caused by forces outside the system being studied. In biological models, an external environmental stress can exceed a species’ tolerance and cause it to disappear. Applied to firms, the idea would attribute failure chiefly to external shocks rather than to interactions among firms. Ormerod contrasts this explanation with endogenous extinction, while allowing that real systems may experience both.
Endogenous extinction
failure from internal interactionsExtinction generated within a system by changing relationships among its components. In the book’s models, species or firms influence one another’s ability to survive through cooperative or competitive connections. A failure can alter the prospects of connected agents and trigger further changes. This mechanism can produce irregular waves of failure without requiring a major external shock.
Fitness
capacity to surviveThe relative ability of a species or firm to persist in its environment. In biological evolution, fitness concerns survival and reproduction; in the book’s organizational models, it is represented through the effects of other agents’ strategies and the surrounding conditions. Fitness is not a guarantee of permanence: even highly adapted species or successful firms can become extinct as environments and connections change.
Creative destruction
innovation replaces old structuresJoseph Schumpeter’s term for the process through which new products, technologies, and forms of organization displace existing ones. This disruption can destroy particular firms, skills, or investments while improving the productive capacity of the wider economy. Ormerod connects the idea to the possibility that failures at the level of individual firms can contribute to renewal at the level of the system.
FAQ
What is "Why Most Things Fail: Evolution, Extinction and Economics" by Paul Ormerod about?
- Core theme of failure: The book investigates why failure is the dominant outcome in both biological evolution and economic systems, arguing that understanding failure is crucial to understanding success.
- Biology and economics analogy: Ormerod draws parallels between species extinction and firm failure, showing that both are governed by similar patterns and dynamics.
- Critique of traditional economics: The book challenges the equilibrium-focused, rational-agent models of mainstream economics, emphasizing complexity, uncertainty, and constant change.
- Interdisciplinary approach: It integrates insights from biology, economics, game theory, and statistical physics to explain why most things fail.
Why should I read "Why Most Things Fail" by Paul Ormerod?
- Fresh perspective on failure: The book offers a new lens for understanding why failure is so common and why success is rare in both nature and business.
- Challenges orthodox economics: Ormerod critiques mainstream economic assumptions, providing a more realistic view of how firms and policies actually behave.
- Practical relevance: Readers gain insights into strategy, innovation, and policy-making in uncertain, complex environments.
- Accessible and empirical: The book uses real-world examples and empirical evidence to make complex theories understandable and actionable.
What are the key takeaways from "Why Most Things Fail" by Paul Ormerod?
- Uncertainty dominates outcomes: The future is inherently unpredictable, and intent does not guarantee success.
- Complex interactions shape results: Networks of positive and negative interactions largely determine survival and extinction patterns.
- Power laws and failure: Both biological and economic failures follow power-law distributions, meaning small failures are common and large failures, though rare, can be catastrophic.
- Bounded rationality: Agents operate with limited knowledge and cognitive ability, which constrains their ability to adapt and learn optimally.
What is the "Iron Law of Failure" in "Why Most Things Fail" by Paul Ormerod?
- Definition: The Iron Law of Failure states that failure is endemic and unavoidable in complex systems, affecting species, companies, and policies alike.
- Biological and economic parallels: Patterns of extinction in nature closely mirror those of firm failures, differing mainly in timescale.
- Implications for success: Success is rare and often temporary; understanding failure is essential for understanding how systems evolve.
- Challenge to survival myths: The law contradicts the idea that only the "fittest" survive, showing that even well-adapted entities eventually fail.
How does Paul Ormerod critique conventional economic theory in "Why Most Things Fail"?
- Equilibrium focus: Ormerod criticizes economics for relying on static equilibrium models that ignore the dynamic, ever-changing nature of real economies.
- Neglect of failure: Traditional economics often overlooks the prevalence of failure, focusing instead on idealized success and optimal decision-making.
- Unrealistic assumptions: Economic models assume perfect knowledge and rationality, which is not reflective of actual market conditions.
- Inadequate treatment of time: The theory fails to address how economies adjust and evolve over time, missing the processes that lead to failure.
What is the significance of power-law distributions in "Why Most Things Fail" by Paul Ormerod?
- Definition and relevance: Power laws describe how the frequency of an event decreases as its size increases, with large failures being rare but possible.
- Biological and economic parallels: Both species extinctions and firm failures follow power-law distributions, indicating similar underlying dynamics.
- Implications for causality: Small shocks can sometimes cause massive effects, challenging the idea that big causes are needed for big effects.
- Practical understanding: Recognizing power laws helps explain phenomena like stock market crashes and mass extinctions as emergent from complex systems.
How does "Why Most Things Fail" by Paul Ormerod explain the gap between intent and outcome in decision-making?
- Bounded rationality: Agents have limited cognitive ability and incomplete information, making it hard to predict the impact of their actions.
- Complexity and uncertainty: The interconnectedness of agents means that even well-intentioned strategies can lead to unintended outcomes.
- Empirical findings: Even small increases in knowledge improve survival, but beyond a point, more knowledge leads to unrealistic predictions.
- Intent vs. outcome: The book shows that intent often fails to translate into desired results due to the unpredictable nature of complex systems.
What is the Hotelling location model/game and its relevance in "Why Most Things Fail" by Paul Ormerod?
- Model setup: The Hotelling model involves competing sellers choosing locations along a line to maximize market share, illustrating strategic decision-making.
- Key insight: Competition drives sellers to cluster in the middle, which is stable but suboptimal for consumers.
- Complexity with more firms: As the number of competitors increases, predicting optimal strategies becomes impossible due to the explosion of possible combinations.
- Implication: The model demonstrates the complexity and indeterminacy of real-world competition, reflecting the uncertainty firms face.
How does Paul Ormerod address the limitations of game theory in "Why Most Things Fail"?
- Promise and limits: Game theory provides a framework for analyzing strategic interactions but often assumes clear rules and rational agents.
- Real-world complexity: Many situations involve multiple players, changing rules, and incomplete information, making game-theoretic solutions impractical.
- Empirical deviations: Experiments show that people often do not play Nash equilibrium strategies, highlighting the gap between theory and practice.
- Unpredictability: The book emphasizes that uncertainty and bounded rationality limit the usefulness of game theory in predicting outcomes.
What are the endogenous and exogenous models of extinction in "Why Most Things Fail" by Paul Ormerod?
- Exogenous model: Extinctions are caused by external shocks that exceed the stress tolerance of species or firms, with agents not connected to each other.
- Endogenous model: Extinctions result from internal network interactions, where the failure of one agent can cascade through the system.
- Empirical fit: Both models can replicate observed extinction patterns, but the endogenous model better explains firm failures due to interconnectedness.
- Hybrid reality: In practice, both external shocks and internal dynamics contribute to extinction events.
What does "Why Most Things Fail" by Paul Ormerod say about the role of failure in economic and social systems?
- Failure as fundamental: Failure is not an anomaly but a pervasive and essential feature of biological, economic, and social systems.
- Systemic benefits: Failure at the individual level can enhance the overall fitness and adaptability of the system.
- Patterns and regularities: The book identifies mathematical regularities in failure events, suggesting that failure is structured, not random.
- Strategic implications: Embracing the inevitability of failure leads to strategies that focus on learning and adaptation.
What policy implications does Paul Ormerod draw in "Why Most Things Fail"?
- Limits of planning: Centralized, detailed planning is likely to fail in complex, evolving systems due to inherent uncertainty.
- Value of competition: A competitive environment fosters innovation and system fitness, but excessive internal competition can be harmful.
- Role of failure: Failure and extinction are necessary for creative destruction, which drives progress and improvement.
- Regulatory caution: Policymakers should balance competition and cooperation, avoiding over-regulation that can stifle economic vitality.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.