OMXUS Press — Paper No. 5
Two Monkey Theory
2026
This paper began with a monkey throwing a cucumber.
8,936 words
~35 min read
7 chapters
OMXUS Press — Paper No. 5
2026
This paper began with a monkey throwing a cucumber.
Why do numerical majorities routinely accept institutional arrangements that demonstrably harm them? This paper introduces the Two Monkey Theory, a behavioural-institutional framework that synthesises findings from primatology, behavioural economics, cognitive psychology, game theory, and institutional analysis to explain the persistence of extractive equilibria and the conditions under which they collapse. The framework takes its name from the capuchin fairness experiments of Brosnan and de Waal (2003), which demonstrated that inequity aversion is phylogenetically ancient, not a cultural invention. If fairness sensitivity is innate, the puzzle is not why people object to inequality but why objection so rarely translates into coordinated reform. We identify four stabilising mechanisms — coordination frictions, narrative locks, selective attention, and cost-of-defection asymmetries — that maintain extractive arrangements despite majority dissatisfaction. We term the latent power of numerical majorities the "Ants Principle," after the insight dramatised in Pixar's A Bug's Life: dominated populations possess overwhelming numerical advantage but are prevented from exercising it through barriers to coordination rather than through genuine powerlessness. Drawing on over 500 studies from the Cooperative Learning Center, recent virtual-reality physiological research (Dal Monte et al., 2024), and extensive cross-domain evidence, we demonstrate that cooperative frameworks achieve equivalent or superior performance outcomes to competitive ones while imposing substantially lower physiological, psychological, and social costs. The paper presents historical case studies as natural experiments confirming the framework's predictions, derives testable hypotheses about tipping-point dynamics, and discusses implications for institutional design, social movements, and public policy. We conclude that the apparent stability of extractive equilibria is fragile — sustained not by genuine consent but by coordination barriers that modern information technologies are progressively eroding.
Keywords: collective action, inequity aversion, system justification, preference falsification, cooperation, competition, institutional design, social movements, coordination failure, game theory, extractive equilibria
Alex Applebee and L. N. Combe
OMXUS Research Series — Paper No. 5
This paper began with a monkey throwing a cucumber.
Not at us. At a researcher. The capuchin had watched its partner receive a grape for the same work and decided — wordlessly, instantly, without a degree in political science — that the arrangement was unacceptable.
That monkey knew something that 8 billion humans appear to have forgotten: unfairness is not a policy debate. It is a biological insult. You feel it before you think it. You reject it before you can articulate why.
The question that drives this paper is not why people object to unfairness. They do. The question is why objection so rarely translates into action — why 99 people will sit in a room being robbed by 1 person, and nobody moves.
This is a coordination failure paper. Not a moral failure paper. The distinction matters.
The 14 Goals that anchor the OMXUS Research Series — direct democracy, 22-hour work weeks, community emergency response, legalised drugs, safe food, play-based schools, free prisoners, no police, no courts, no foreign housing speculation, free internet, climbing infrastructure, cancer prevention, the $29 ring — are not utopian. They are the logically inevitable conclusions of the evidence. Every one of them has been implemented somewhere, by someone, and worked. The only reason they have not been implemented everywhere is that the people who would benefit from them — the overwhelming numerical majority — cannot coordinate.
That is the thesis.
Good people produce bad systems not because they are stupid, weak, or complicit, but because the architecture of coordination has been designed to prevent them from acting on what they already know. The cucumber-throwing monkey had one advantage over you: it could see the other monkey getting the grape. You cannot see the 99 other people in the room who agree with you, because every institution between you and them is optimised to make sure you never count.
This paper is the count.
We are not neutral observers. The evidence reviewed here — over 500 studies on cooperation versus competition, cross-cultural fairness research spanning 15 societies, physiological data showing cooperation achieves equal performance at lower biological cost, historical case studies from the labour movement to Indian independence — converges on a single conclusion: the systems we live under are extractive equilibria maintained by coordination barriers, not by consent. They are fragile. They have always been fragile. And every technology that lowers the cost of coordination makes them more so.
The ants outnumber the grasshoppers a hundred to one. They always have.
The only question is when they remember to count.
— A.A. & L.N.C.
Why do numerical majorities routinely accept institutional arrangements that demonstrably harm them? This paper introduces the Two Monkey Theory, a behavioural-institutional framework that synthesises findings from primatology, behavioural economics, cognitive psychology, game theory, and institutional analysis to explain the persistence of extractive equilibria and the conditions under which they collapse. The framework takes its name from the capuchin fairness experiments of Brosnan and de Waal (2003), which demonstrated that inequity aversion is phylogenetically ancient, not a cultural invention. If fairness sensitivity is innate, the puzzle is not why people object to inequality but why objection so rarely translates into coordinated reform. We identify four stabilising mechanisms — coordination frictions, narrative locks, selective attention, and cost-of-defection asymmetries — that maintain extractive arrangements despite majority dissatisfaction. We term the latent power of numerical majorities the "Ants Principle," after the insight dramatised in Pixar's A Bug's Life: dominated populations possess overwhelming numerical advantage but are prevented from exercising it through barriers to coordination rather than through genuine powerlessness. Drawing on over 500 studies from the Cooperative Learning Center, recent virtual-reality physiological research (Dal Monte et al., 2024), and extensive cross-domain evidence, we demonstrate that cooperative frameworks achieve equivalent or superior performance outcomes to competitive ones while imposing substantially lower physiological, psychological, and social costs. The paper presents historical case studies as natural experiments confirming the framework's predictions, derives testable hypotheses about tipping-point dynamics, and discusses implications for institutional design, social movements, and public policy. We conclude that the apparent stability of extractive equilibria is fragile — sustained not by genuine consent but by coordination barriers that modern information technologies are progressively eroding.
Keywords: collective action, inequity aversion, system justification, preference falsification, cooperation, competition, institutional design, social movements, coordination failure, game theory, extractive equilibria
In 2003, primatologists Sarah Brosnan and Frans de Waal published a study in Nature that would become one of the most-cited experiments in behavioural science. Two capuchin monkeys performed identical tasks — handing a stone token to a researcher. One received a cucumber slice; the other received a grape. The cucumber-receiving monkey, upon witnessing this disparity, refused to continue participating, frequently hurling the cucumber back at the researcher (Brosnan & de Waal, 2003). The experiment demonstrated something profound: inequity aversion is not a product of human culture, moral philosophy, or political ideology. It is phylogenetically ancient, present in species that diverged from the human lineage approximately 35 million years ago.
This finding generates a puzzle. If the rejection of unfairness is wired into primate cognition, why do human societies — populated by beings with far greater cognitive, communicative, and organizational capacities than capuchins — sustain levels of inequality that would provoke immediate revolt in a capuchin colony? The top 1% of wealth holders in the United States own approximately 38% of all wealth, while the bottom 90% collectively hold roughly 23% (Federal Reserve, 2023). The ratio of average CEO compensation to median worker pay exceeds 350:1 (Economic Policy Institute, 2023). In Australia, the top 20% of households hold 63% of total wealth while the bottom 20% hold less than 1% (ABS, 2022). The working population outnumbers economic elites by at least 99 to 1. The mathematics of power are overwhelmingly on the side of the majority. Yet the majority acquiesces.
This paper introduces the Two Monkey Theory as a framework for resolving this puzzle. The framework is not a single hypothesis but an integrated explanatory architecture that draws on behavioural economics, evolutionary psychology, cognitive science, game theory, and institutional analysis. Its core claim is tripartite:
The paper proceeds as follows. Section 2 reviews the relevant literatures. Section 3 develops the Two Monkey Framework in detail. Section 4 formalises the game-theoretic structure of coordination failure. Section 5 synthesises the extensive evidence on cooperation versus competition. Section 6 examines historical case studies as natural experiments. Section 7 applies the framework to Australian governance as a contemporary case study. Section 8 analyses tipping-point dynamics. Section 9 presents falsifiable predictions. Section 10 discusses implications and limitations. Section 11 concludes.
The behavioural economics of fairness extends well beyond the capuchin experiments. Ultimatum game research consistently demonstrates that human participants reject offers they perceive as unfair, even at personal cost — a behaviour that is irrational under classical expected utility theory but robust across cultures (Henrich et al., 2005; Guth, Schmittberger, & Schwarze, 1982). Dictator games reveal that most people voluntarily share resources even when anonymity eliminates reputational incentives (Engel, 2011). Public goods games show that people contribute to collective welfare and punish free-riders, even when punishment is costly (Fehr & Gachter, 2000).
Cross-cultural research conducted by Henrich and colleagues (2005) across fifteen small-scale societies demonstrated that fairness norms vary in their specific expression but are universally present. No society studied exhibited the purely self-interested behaviour predicted by classical rational choice theory. These findings converge with the primate evidence to suggest that fairness sensitivity is a deep feature of social cognition, not a cultural artifact.
Critically, Bohnet and colleagues (2021) have shown that people accept unfairness from systems more readily than from identifiable individuals — a finding with direct implications for how institutional inequality persists. When unfairness is mediated through impersonal structures, the emotional response that drives rejection in face-to-face encounters is attenuated. The capuchin can see the other monkey getting the grape. You cannot see the CEO getting 350 times your salary — the institution stands between you and the disparity, and the institution has no face to throw a cucumber at.
System justification theory (Jost & Banaji, 1994; Jost, Banaji, & Nosek, 2004) provides a complementary psychological account. The theory proposes that people possess a motivated tendency to defend, bolster, and justify existing social arrangements, even when those arrangements are personally disadvantageous. This tendency intensifies as inequality becomes more extreme (Jost et al., 2019), creating a perverse dynamic in which the systems most in need of reform generate the strongest psychological resistance to reform.
System justification manifests through several pathways: the just-world hypothesis, in which people believe that outcomes are deserved (Lerner, 1980); the belief in meritocracy, which attributes differential outcomes to differential effort and talent rather than structural advantage (McNamee & Miller, 2009); and false consciousness, in which dominated groups internalise the ideological frameworks of dominant groups against their own material interests (Jost, 1995).
The perversity of this mechanism cannot be overstated. It means that the worse things get, the harder people work to convince themselves that things are fine. The more extractive the arrangement, the more psychological energy the exploited invest in defending it. This is not stupidity. It is a cognitive coping mechanism operating at scale — and it is one of the four pillars holding up every unfair system on earth.
Timur Kuran's (1997) theory of preference falsification addresses the gap between private dissatisfaction and public compliance. Kuran argues that individuals publicly express support for systems they privately question because the perceived costs of dissent exceed the perceived benefits. This creates an information problem: because each individual observes primarily public expressions rather than private preferences, the apparent level of support for the status quo is systematically inflated. Everyone believes they are in a minority of dissenters when they may in fact be in a supermajority.
This informational distortion creates the conditions for sudden, unexpected regime change. When some exogenous shock reduces the cost of expressing dissent — or when a critical mass of individuals begin expressing their true preferences — the resulting cascade can be explosive. Kuran uses this framework to explain events ranging from the fall of the Soviet Union to the Arab Spring: systems that appeared stable for decades collapsed in weeks once preference falsification ended.
The implication is devastating for every politician who claims a mandate: the votes they count are not preferences. They are performances. The real preferences are spoken at kitchen tables, in private messages, in the silence after the news broadcast. And those preferences, if they could be aggregated honestly, would produce a very different world.
Olson's (1965) classic analysis of collective action demonstrates that shared interests do not automatically produce coordinated behaviour. Large groups face free-rider problems: each individual benefits from collective action whether or not they personally contribute, creating incentives to defect. Coordination games add further complexity: even when all parties prefer collective action, uncertainty about others' participation can prevent anyone from moving first (Myerson, 2017).
First-mover disadvantage compounds the problem. Early actors in any reform movement bear disproportionate costs — social sanctions, economic retaliation, physical danger — while the benefits of successful reform are distributed broadly. This asymmetry between concentrated costs and diffuse benefits creates a structural bias toward inaction, even when the aggregate benefits of reform vastly exceed the aggregate costs.
Elinor Ostrom's (1990) work on governing the commons demonstrated that these problems are not insoluble — communities worldwide have developed institutional solutions for managing shared resources without either privatisation or centralised control. Her eight design principles for successful commons governance (clearly defined boundaries, proportional equivalence between benefits and costs, collective-choice arrangements, monitoring, graduated sanctions, conflict-resolution mechanisms, minimal recognition of rights to organise, and nested enterprises) provide a concrete institutional blueprint for overcoming coordination failures. The principles are not theoretical. They describe what actual communities actually do when the state and the market both fail them.
Sapolsky's (2017) synthesis of primate research in Behave documents how dominance hierarchies in primate societies are maintained through a combination of coalition building, resource control, threat displays, selective benefit distribution, and divide-and-rule strategies. Human societies exhibit precise parallels: elite networks and political alliances mirror coalition building; concentrated capital ownership mirrors resource control; military and police displays of force mirror threat displays; preferential economic policies mirror selective benefits; and identity-based divisions mirror divide-and-rule tactics.
Critically, neuroscience research demonstrates that perceiving oneself as powerless alters brain function, reducing executive control and increasing threat response (Sapolsky, 2017). This creates a neurological feedback loop: structural powerlessness produces cognitive states that further reduce the capacity for coordinated resistance. The system does not merely oppress you. It rewires your brain to make you less capable of recognising the oppression.
The selective attention literature, anchored by the famous "Invisible Gorilla" experiment of Simons and Chabris (1999), reveals that human attention is severely constrained. The brain processes approximately 11 million bits of information per second, but conscious awareness handles only about 50 bits per second (Zimmermann, 1989). This creates massive filtering requirements. Critically, what passes through the filter is determined not by objective importance but by expectation, cognitive load, and perceptual set.
This has direct implications for the persistence of inequality. As Kahneman (2011) has shown, cognitive systems are strongly biased toward confirming existing frameworks and screening out disconfirming evidence. Information that contradicts prevailing narratives about how economic and social systems work is systematically filtered before reaching conscious awareness — not through deliberate censorship but through the basic architecture of human cognition.
Fifty percent of participants in the Invisible Gorilla experiment failed to notice a person in a gorilla costume walking through the middle of a basketball game. The parallel to politics is not metaphorical. It is neurological. Citizens focused on GDP growth miss the gorilla of wealth inequality. Consumers focused on price miss the gorilla of planned obsolescence. Voters focused on party politics miss the gorilla of structural power that transcends every party. The information reaches the retina. It does not reach consciousness. And the filtering is not random — it is shaped by the same institutions that benefit from your inattention.
The first pillar of the framework establishes that inequity aversion is a biological endowment, not a cultural construction. The evidence for this claim comes from three converging lines of research.
Primate evidence. Brosnan and de Waal's (2003) capuchin experiments have been replicated and extended across multiple primate species. Chimpanzees show similar inequity aversion in cooperative contexts (Brosnan, Schiff, & de Waal, 2005). Notably, when both monkeys in the original experiment received grapes — high-value, equal rewards — cooperation continued and productivity increased. This finding is theoretically significant: it demonstrates that equality does not merely satisfy a fairness constraint but actively enhances cooperative performance.
Cross-cultural evidence. The large-scale cross-cultural experiments conducted by Henrich and colleagues (2005) across fifteen societies on five continents demonstrated that fairness norms are universal, though their specific expression varies with social ecology. Nowhere did researchers find the purely self-interested homo economicus predicted by neoclassical theory. The Ultimatum Game rejection rates — in which participants sacrifice real money to punish unfair offers — averaged around 40-50% for offers below 20% of the stake across all societies studied.
Developmental evidence. Research on children's fairness intuitions shows that inequity aversion emerges early in development, well before explicit moral instruction. By age three, children protest unequal distributions; by age eight, they sacrifice personal gain to establish equality (Fehr, Bernhard, & Rockenbach, 2008). These developmental findings are consistent with a biological basis for fairness sensitivity.
The convergence of primate, cross-cultural, and developmental evidence establishes a strong case that humans enter the world equipped with cognitive-emotional machinery that detects and responds to inequity. The puzzle, therefore, is not why people object to unfairness but why this objection is so often suppressed, redirected, or rendered ineffective at the collective level.
The framework identifies four interlocking mechanisms that stabilise extractive equilibria despite majority dissatisfaction. These mechanisms do not operate independently; they reinforce one another in a self-sustaining architecture of inertia.
The most fundamental barrier to collective action is not the absence of shared preferences but the inability to coordinate simultaneous action. Even when every individual in a population would prefer a different equilibrium, each individual faces a coordination dilemma: "If I move alone, I bear all the costs and gain nothing. If everyone moves together, we all benefit. But I cannot know whether others will move."
This is a classic assurance game (Sen, 1967). Unlike the prisoner's dilemma, in which defection is individually rational regardless of what others do, the assurance game has two equilibria: one in which everyone cooperates (superior for all) and one in which no one cooperates (inferior for all but individually safe). The status quo persists not because anyone prefers it but because no one can be assured that others will participate in the transition.
Game-theoretic analysis confirms that uncertainty about others' participation creates stable suboptimal equilibria (Myerson, 2017). The larger the group, the more severe the coordination problem, because each individual's confidence that "enough others will participate" decreases with group size.
Coordination frictions explain why people fail to act on their preferences. Narrative locks explain why many people fail to form accurate preferences in the first place. A narrative lock is a coherent but misleading explanatory framework that makes extractive arrangements appear natural, inevitable, or beneficial.
Examples include:
Narrative locks are maintained through information control (media concentration, financial system complexity), institutional reinforcement (educational curricula, policy think tanks), and the psychological mechanisms described by system justification theory. They function as what Gramsci (1971) called cultural hegemony: the dominance of a particular worldview to the point where it appears not as ideology but as common sense.
The cognitive science of attention provides a mechanism through which narrative locks operate at the individual level. Selective attention does not merely filter sensory information; it filters conceptual information. People systematically fail to perceive phenomena that contradict their existing frameworks.
Simons and Chabris's (1999) Invisible Gorilla experiment provides the foundational demonstration: approximately 50% of participants failed to notice a person in a gorilla costume walking through the middle of a basketball-passing exercise. The parallel to social perception is direct. Citizens focused on GDP growth miss the "gorilla" of wealth inequality. Consumers focused on purchase price miss the "gorilla" of lifecycle cost and planned obsolescence. Voters focused on party politics miss the "gorilla" of structural power that transcends partisan divisions.
This is not metaphorical. The cognitive mechanisms are identical. As Mack and Rock (1998) demonstrated, inattentional blindness is not a failure of visual processing but a failure of attentional allocation. The information reaches the retina — or, in the social case, the information environment — but is filtered before reaching conscious awareness. Zerubavel (1997) extended this analysis to "social mindscapes," showing how entire communities can share patterns of attention and inattention that produce collective blindness to phenomena that would be obvious from a different attentional framework.
Critically, power structures actively manage collective attention. As the selective attention literature demonstrates, focusing attention on one set of phenomena necessarily creates blindness to others. Systems of power exploit this by directing public attention toward proximate causes rather than structural ones, individual actors rather than systemic patterns, and short-term crises rather than long-term trends (Davenport & Beck, 2001; Wu, 2016). This is not always deliberate conspiracy; it can emerge from the structural incentives of media systems, political competition, and information markets.
Even when individuals see through narrative locks and overcome attentional biases, they face asymmetric costs for defection. Powerful incumbents can impose concentrated costs on individual dissenters while the benefits of successful reform are distributed broadly across the population.
Historical examples are abundant: early union organizers faced firing, blacklisting, physical violence, and imprisonment. Civil rights activists faced economic retaliation, social ostracism, and murder. Whistleblowers face career destruction and legal persecution. These costs fall on identifiable individuals, while the benefits of the reforms they seek would accrue to millions of anonymous beneficiaries.
This asymmetry creates a first-mover disadvantage that reinforces the coordination friction described above. Even individuals who are fully aware of systemic injustice and personally committed to reform face a rational deterrent: the concentrated costs of acting alone versus the diffuse benefits of successful collective action (Olson, 1965).
In game-theoretic terms, the cost-of-defection asymmetry transforms what would otherwise be a simple coordination game into a game of chicken: the first mover risks catastrophic personal loss, while the benefits of successful coordination are shared by all — including those who waited safely until the outcome was assured.
The mechanisms described above explain the persistence of extractive equilibria. The Ants Principle addresses the opposite question: what is the latent power of numerical majorities, and why should we expect coordination barriers to be ultimately unstable?
The principle takes its name from a scene in Pixar's A Bug's Life (1998) in which the antagonist Hopper articulates the fundamental vulnerability of every extractive system:
"You let one ant stand up to us, then they all might stand up! Those puny little ants outnumber us a hundred to one. And if they ever figure that out... there goes our way of life!"
This is not merely a children's film conceit. It is a precise statement of the power relationship in any extractive equilibrium. The dominated population produces the resources, provides the labour, and constitutes the overwhelming numerical majority. The extracting minority contributes primarily the coordination and control infrastructure that prevents the majority from exercising its latent power.
The Ants Principle holds that:
The historical record confirms this analysis. Throughout human history, general strikes — the simultaneous withdrawal of collective cooperation — have toppled governments, won labour rights, and ended wars. The withdrawal of cooperation is arguably the most powerful force in social change, and it requires no violence, no military capability, and no material resources. It requires only coordination.
The fourth component of the framework addresses a subtler mechanism through which extractive equilibria are maintained: the construction of social norms and cognitive biases that make inequality appear natural rather than constructed.
Research from Harvard's Project Implicit, using Implicit Association Tests (IATs), has demonstrated that unconscious biases are pervasive, operating below conscious awareness and often contradicting individuals' explicitly stated values (Greenwald & Banaji, 1995). These biases are not innate but learned — they reflect the associations present in the cultural environment (Devine, 1989).
The critical insight is that unconscious bias and constructed social norms form a self-reinforcing feedback loop:
The process of naturalization — through which contingent social arrangements come to appear as inevitable features of reality — operates through several mechanisms identified in the social construction literature (Berger & Luckmann, 1966):
As Roger Johnson has explained: "A typical classroom teacher is taught to keep students quiet and apart, indirectly fostering competition. Yet... people learn best when they work cooperatively with each other." The irony is substantial: educational systems designed around competition actively undermine their own stated objective of maximizing learning.
The professional evidence parallels the educational findings:
The open-source software movement provides a large-scale natural experiment. The Linux kernel, developed cooperatively by thousands of contributors, powers everything from Android smartphones to the world's 500 fastest supercomputers. Apache, developed cooperatively, runs approximately 35% of all websites worldwide. Wikipedia, built entirely through voluntary cooperation, has become the world's largest encyclopedia. These cooperative products consistently outperform or match their proprietary, competition-produced counterparts.
The health evidence is particularly significant because it reveals biological mechanisms through which cooperation and competition differentially affect human functioning.
A study published in the Journal of Psychology (Johnson, Johnson, & Krotee, 1986) examined 57 collegiate and semi-professional ice hockey players trying out for the 1980 U.S. Olympic team. Using personality measures and social-interaction scales, researchers found that more cooperative individuals were better adjusted psychologically and physically healthier than competitive colleagues. Competition created unhealthy physiological side effects; cooperation generated positive physiological responses comparable to "runner's high."
More recently, Dal Monte and colleagues (2024) conducted a study using virtual reality Stroop tests with physiological monitoring that provides what may be the cleanest demonstration of the cooperation-competition differential to date. Their key findings:
The researchers concluded: "Cooperation can be just as effective as competition in improving individuals' performance. However, cooperation does not carry the same level of stress and physiological burden as the competitive context, representing a healthier and more optimal way to boost individual performance."
This finding has direct implications for institutional design. If cooperative arrangements achieve equal performance outcomes with lower physiological costs, then the massive stress-related health burden associated with competitive workplace cultures — estimated at $190 billion annually in U.S. healthcare costs and 120,000 excess deaths per year (Goh, Pfeffer, & Zenios, 2016) — represents a preventable consequence of institutional design choices. Not a cost of doing business. A cost of doing business wrong.
The Two Monkey Framework predicts that competitive systems will generate not only individual costs (stress, health effects) but systemic costs that are invisible under competitive attentional frames. The evidence confirms this prediction.
The "Two Washing Machine Theory" illustrates how competitive dynamics generate systematic waste. In a competition-based economy, product durability becomes a liability. A washing machine that lasts 20 years generates $800 in revenue; a machine designed to fail after 2 years generates $4,000 over the same period. The mathematics of competition actively discourage creating products that last.
The environmental costs are substantial: the average washing machine contains 30 kg of steel, requiring approximately 800 kWh of energy to produce — enough to power an average home for a year. When machines are discarded every 2-5 years instead of every 20+, these environmental impacts multiply by factors of 4 to 10. This calculation covers only steel, ignoring plastics, electronics, mining impacts, transportation emissions, and landfill costs.
This pattern is not limited to washing machines. It extends to smartphones designed for 2-3 year replacement cycles with sealed batteries; software operating systems that compete rather than complement each other, duplicating core functionalities across platforms; automobiles with styling changes prioritised over mechanical longevity; and fast fashion designed to last a season rather than years.
Competition produces massive duplication of effort that cooperative arrangements would eliminate. The software industry provides a clear example: the parallel development of Windows, macOS, and multiple Linux distributions duplicates core functionalities while creating compatibility barriers that consume developer time and user patience. The cumulative cost of this duplication — in developer hours, user frustration, and computing resources devoted to compatibility layers — is enormous and almost entirely invisible.
Competition-based systems systematically generate information asymmetries that undermine the theoretical efficiency justification for competition. Consumers cannot easily assess product longevity at purchase time. Marketing creates the impression of continuous improvement while actual quality may decline. Technical complexity obscures repairability. True lifecycle costs remain hidden behind point-of-sale pricing.
These asymmetries represent a fundamental market failure: competition is supposed to drive efficiency through informed consumer choice, but the competitive dynamics themselves generate the information distortions that prevent informed choice.
Against these systemic costs, cooperative models demonstrate viable alternatives:
These examples confirm the framework's prediction: cooperative arrangements are not merely idealistic alternatives but practically superior ones, held back not by their own limitations but by coordination frictions and narrative locks that privilege competitive arrangements.
The Two Monkey Framework generates specific predictions about the conditions under which extractive equilibria collapse. Historical cases of large-scale social transformation provide natural experiments for testing these predictions.
The labour movement of the late nineteenth and early twentieth centuries provides a paradigmatic example of the Ants Principle in action. Industrial workers outnumbered factory owners by orders of magnitude and were functionally indispensable to production. Yet for decades, workers accepted conditions that included 12-16 hour days, child labour, dangerous workplaces, and subsistence wages.
The stabilising mechanisms are clearly identifiable:
The breakthrough came when coordination costs dropped sufficiently to enable collective action: the formation of trade unions provided institutional infrastructure for coordination; the printing press and later the telegraph enabled information sharing; and early successful strikes demonstrated feasibility. The resulting transformations — the eight-hour day, the weekend, minimum wage laws, workplace safety regulations, the prohibition of child labour — represented a massive shift in the distribution of surplus from capital to labour.
The speed of transformation, once coordination barriers fell, is consistent with the preference falsification cascade predicted by the framework. Workers had privately opposed their conditions for decades; once collective expression became feasible, the apparent consensus in favour of existing arrangements evaporated rapidly.
The American Civil Rights Movement exhibits the same structural dynamics. African Americans constituted a majority in many Southern communities and were economically indispensable to the agricultural and service economies. Yet the Jim Crow system persisted for nearly a century after the formal end of slavery.
The stabilising mechanisms were:
The breakthrough came through a combination of institutional infrastructure (Black churches, the NAACP, student organizations), information technology (television broadcasting of police violence against peaceful demonstrators), and demonstration effects (early victories in Montgomery and elsewhere). The movement's success was fundamentally an exercise of the Ants Principle: the withdrawal of cooperation through boycotts, sit-ins, and marches by a numerical majority that had always possessed the power to disrupt the system but had been prevented from coordinating its exercise.
The women's suffrage movement provides perhaps the purest illustration of the framework's predictions. Women constituted approximately 50% of the population — an absolute numerical majority — yet were denied political voice for centuries. The absurdity of this arrangement is, in retrospect, staggering: a majority was excluded from governance by a minority, and the arrangement was sustained not by genuine incapacity but by narrative locks ("women are unfit for politics"), coordination frictions (women were isolated in domestic spheres), selective attention (women's political capabilities were invisible within the prevailing attentional framework), and cost-of-defection asymmetries (suffragists faced imprisonment, force-feeding, social ostracism, and family dissolution).
India's independence movement illustrates the Ants Principle at continental scale. Three hundred million Indians were governed by a British administration of a few thousand officials, supported by an Indian military and civil service that depended entirely on Indian cooperation. Gandhi's strategic genius lay in recognizing that the British Empire in India was sustained not by British power but by Indian cooperation — and that the withdrawal of that cooperation was irresistible.
The Salt March, the Quit India movement, and the broader campaign of noncooperation were all exercises of the Ants Principle: demonstrating that the numerical and functional reality made continued colonial governance impossible once coordination barriers were overcome.
Two contemporary examples deserve attention because they are not historical — they are happening now.
Rojava (Autonomous Administration of North and East Syria). Since 2012, approximately 4.6 million people in northern Syria have been governing themselves through a system of direct-democratic communes, councils, and cooperatives — without a state, without a traditional military hierarchy, and under active military threat. The system is built on principles of gender equality (mandatory co-leadership by one man and one woman at every level), ethnic pluralism (Kurdish, Arab, Assyrian, and Turkmen communities share governance), and ecological sustainability. Rojava is the Ants Principle in its most literal form: a population that was denied autonomy by every surrounding state simply began governing itself, and nobody could stop them because the coordination barriers had already collapsed through shared existential threat.
Zapatista Autonomous Municipalities (Chiapas, Mexico). Since 1994, approximately 360,000 indigenous Maya people in southeastern Mexico have maintained autonomous governance outside the Mexican state. Their system of rotating leadership (cargos), community assemblies, and autonomous education and health systems has persisted for over 30 years. The Zapatista motto — mandar obedeciendo ("to lead by obeying") — is a direct institutional expression of the Ants Principle: governance that cannot extract because the governed are the governors.
These are not theoretical possibilities. They are existence proofs. The narrative lock that says "this cannot work" is falsified by 30 years of it working.
Across all cases — historical and contemporary — the same pattern obtains:
The Two Monkey Framework predicts that representative democracies will exhibit the same extractive dynamics as other institutional arrangements when coordination barriers prevent citizens from exercising direct governance. Australia provides a particularly instructive case because it combines compulsory voting — which eliminates the participation problem — with institutional structures that nevertheless produce coordination failure.
Australia's political system concentrates decision-making authority in approximately 227 federal parliamentarians who make policy for 26 million people — a representation ratio of roughly 1:115,000. At the state level, the ratio is similarly extreme. This is not direct democracy. It is a hiring decision made once every three years, after which the hired representatives face minimal accountability for the decisions they make between elections.
The four mechanisms operate clearly:
Coordination frictions. Australia's preferential voting system and party-dominated preselection processes make it structurally difficult for citizens to coordinate around alternatives to the major parties. Independent candidates face prohibitive costs of name recognition, ballot positioning, and campaign financing. The two-party preferred system actively channels votes toward the Coalition or Labor, regardless of how many citizens would prefer neither.
Narrative locks. Australia's political discourse is shaped by an unusually concentrated media landscape. Two corporations — News Corp Australia and Nine Entertainment — control approximately 70% of print media circulation (ACCC, 2019). This concentration produces a narrow range of "acceptable" political positions that excludes most alternatives, including direct democracy, which has been practised successfully by Switzerland for 178 years. The narrative lock is not "we cannot have direct democracy" — it is that the question is never asked.
Selective attention. Australian political coverage focuses overwhelmingly on leadership contests, personality politics, and party strategy — the "horse race" — rather than on structural questions about whether representative democracy itself is the best available system. The question "should citizens vote on policy directly?" receives approximately zero attention in mainstream media, despite Switzerland demonstrating for 178 years that the answer is yes.
Cost-of-defection asymmetries. Australians who challenge the two-party system face social pressure ("wasting your vote"), institutional barriers (above-the-line Senate voting funnels preferences through party-controlled deals), and the practical reality that independent representatives have minimal power within a party-dominated parliament.
The framework identifies several areas where Australian governance produces outcomes that the majority of citizens privately oppose:
Housing affordability. Australia's median house price-to-income ratio is approximately 13:1 in Sydney and 10:1 in Melbourne (Demographia, 2024), among the highest in the world. The majority of Australians aged 25-40 cannot afford to purchase housing in the city where they work. Yet housing policy consistently favors existing property owners and investors through negative gearing, capital gains tax discounts, and foreign investment rules that are permissive by international standards. The coordination failure: the majority who need affordable housing cannot coordinate against the minority who profit from expensive housing, because the minority's interests are concentrated and well-represented while the majority's interests are diffuse and poorly coordinated.
Climate and energy policy. Australia is one of the world's largest per-capita emitters and largest fossil fuel exporters. Survey data consistently shows majority support for stronger climate action (Lowy Institute, 2023). Yet policy lags far behind public preference, because the fossil fuel industry's interests are concentrated, well-funded, and structurally embedded in the political donation system — while the public interest in a stable climate is diffuse and lacks institutional representation. This is the Volunteer's Dilemma at national scale.
Indigenous affairs. The 2023 Voice to Parliament referendum demonstrated all four mechanisms simultaneously. A proposal supported by the majority of Indigenous Australians was defeated 60-40 after a campaign characterised by narrative locks ("if you don't know, vote no"), selective attention (the campaign focused on procedural uncertainty rather than substantive justice), coordination frictions (supporters were fragmented across multiple messaging strategies), and cost-of-defection asymmetries (politicians who supported the Voice risked electoral backlash from constituencies primed by concentrated media opposition). The preference falsification dimension is evident in polling: support dropped 20 points between early polls and the vote, consistent with Kuran's model of publicly expressed preferences converging toward the perceived majority position under conditions of uncertainty.
Wage stagnation. Australian real wages have been stagnant or declining for over a decade despite productivity growth (ABS, 2023). The gap between productivity and wage growth represents a transfer of surplus from labour to capital — precisely the extractive dynamic described by the framework. The coordination failure is identical to the pre-union labour movement: individual workers cannot negotiate against structural power, and collective bargaining has been progressively weakened through legislative changes to the Fair Work Act that were themselves products of the coordination failure they perpetuate.
The comparison with Switzerland is devastating for the Australian model. Switzerland has conducted over 700 federal referendums since 1848 — 178 years of direct democracy — and consistently ranks at or near the top of global indices for quality of life, economic competitiveness, social stability, trust in institutions, and citizen satisfaction with government (World Happiness Report, 2024; IMF, 2024; World Values Survey).
Switzerland achieves this despite having no natural resources (no iron ore, no coal, no significant agriculture), four official languages, and a diverse religious and cultural landscape that Australian politicians routinely cite as barriers to direct democracy. The Swiss example falsifies every argument against citizen-led governance. It is not a thought experiment. It is 178 years of data.
The framework predicts that the absence of direct democracy in Australia is maintained not by practical impossibility but by the four mechanisms: coordination frictions (no institutional pathway to referendum), narrative locks ("Australians aren't informed enough to vote on policy"), selective attention (the Swiss model is virtually invisible in Australian public discourse), and cost-of-defection asymmetries (politicians who propose direct democracy are marginalised by party structures that would be rendered obsolete by it).
The Two Monkey Framework predicts that extractive equilibria, despite their apparent stability, are vulnerable to sudden collapse when coordination barriers erode past critical thresholds.
Kuran's (1997) model of preference falsification provides the formal mechanism. Each individual has a private threshold — the level of visible dissent they must observe before they are willing to express their own dissent publicly. These thresholds are heterogeneously distributed across the population. When the level of observed dissent reaches a given individual's threshold, that individual adds their voice to the public expression of dissent, which in turn may push other individuals past their thresholds.
The dynamics are nonlinear. A system can appear perfectly stable for decades — with public discourse reflecting near-unanimous support for existing arrangements — and then collapse in weeks when a small perturbation triggers a cascade through the threshold distribution. The key insight is that the apparently stable state was always fragile; the stability reflected coordination failure, not genuine consent.
The framework identifies three conditions that, when met simultaneously, are sufficient to initiate a cascade:
Modern information and coordination technologies are systematically eroding the barriers that have historically stabilised extractive equilibria. Specifically:
These principles define what we term a "trust-first" institutional posture — not naive trust, but efficiency-optimised trust backed by transparency and reversibility. Trust-first designs assume cooperation, make information public by default, and rely on reversibility plus observability to bound risk. The hypothesis is that such designs achieve equal or better compliance at lower administrative cost relative to control-first systems (see P3 above).
The framework suggests that successful social movements should prioritise:
Policy implications include:
Where b = benefit of reform, c = individual cost of participation, k = critical threshold.
Equilibria: Two pure-strategy Nash equilibria exist: (1) all cooperate (if b > c and n_c >= k), and (2) all defect. The all-defect equilibrium is risk-dominant when uncertainty about others' participation is high. The all-cooperate equilibrium is payoff-dominant. The coordination problem is: how to move from the risk-dominant equilibrium to the payoff-dominant one.
Key insight: The problem is not that defection is individually rational (as in the Prisoner's Dilemma). The problem is that cooperation is conditionally rational — rational if and only if enough others cooperate — and the condition cannot be verified in advance.
Setup: N individuals indexed by private dissatisfaction threshold t_i, uniformly distributed on [0,1]. Individual i publicly expresses dissent if and only if the observed fraction of dissenters d exceeds t_i.
Dynamics: Starting from d = 0, a small exogenous shock increases d to d_0 > 0. All individuals with t_i