Economics fails the Philosophy of Science Test

The Perverse Survival of Homo Economicus in the age of Evolutionary Psychology

1.   The Scandal Nobody Talks About

The central organising concept of modern macroeconomics – the rational optimising agent – is fundamentally incompatible with science.

That is an extraordinary statement. It is also, I will argue, well supported by evidence.

Imagine a science of ecology built around a theory of metabolism rejected by biologists. Imagine a branch of psychology that is incompatible with fundamental knowledge of neurology and brain mapping. Imagine a branch of chemistry that is fundamentally incompatible with and contradicted by our best theories of physics. Such a discipline would be considered to be in crisis, and in need of an urgent paradigm shift.

Yet economics has somehow escaped this scrutiny.

For more than a century, economists have built increasingly sophisticated theories upon a particular conception of the human mind – Homo Economicus – the rational optimising agent. This agent is assumed to process information, evaluate alternatives, form stable preferences, compare future possibilities, and choose actions that maximise utility.

Upon this conceptual agent, economists have elaborated an ever-more sophisticated system of invisible and unobservable hypothesised “shocks”, “preferences”, “revealed preferences”, and “frictions”. The role of these unobservable (and therefore unscientific) elaborations is to protect and insulate the central theory of utility optimisation from empirical evidence. Each test that shows the theory fails results in a new invisible, unobservable tweak, each having no role other than preserving the central sanctified hypothesis.

Homo Economicus sits at the centre of almost everything that passes for knowledge in mainstream macroeconomics. It is the foundation upon which theories of consumption, saving, investment, expectations, growth, welfare and public policy are constructed. It is the foundational assumption upon which the economic impacts of proposed public policy interventions are measured and assessed. It acts as an effective filter on the policies that government is able to justify in the language of accepted economics.

It is difficult to overstate its importance.

The problem is that outside economics, almost no other science believes such a cognitive architecture exists. Far from it – this model of human cognition is a complete “non-starter” in what I will unashamedly call “real science”. The theory of cognition adopted by mainstream economics simply contradicts too much of the hard empirical evidence that has developed over the last 100 years.

Homo Economicus would not be taken seriously in any serious scientific discussion.

Psychology has spent more than a century documenting the heuristic and bounded nature of human reasoning. Neuroscience has revealed a brain that is adaptive, approximate and resource constrained. A structure that is ‘hard-wired’ with adaptive evolved, parcellated functions, rather than a blank-slate rational information processor.

Evolutionary biology has provided increasingly sophisticated accounts of how cognitive systems emerge through natural selection, where each functional component is an adaptive solution to real-world information processing challenges faced by the organism within its information-bounded local environment.

Behavioural economics has repeatedly demonstrated systematic departures from the behaviour predicted by rational utility optimisation.

The sciences studying cognition have moved in one direction. Economics has largely remained fixed with the assumptions it adopted 80 or 100 years ago.

This essay asks a simple question:

Why has economics refused to update its assumptions based on evidence?

The answer matters because economics is not merely an academic exercise. Economic theories shape public policy. They influence taxation, welfare, labour markets, regulation, education, retirement systems and responses to inequality. When economic models are wrong, their consequences do not remain confined to universities. They affect societies. They impoverish real people.

At first glance, the problem appears to concern technical disputes about particular hypothesised theoretical constructs – utility functions, representative agents, equilibrium conditions and growth theory.

In reality, the problem is far deeper. It is a fundamental problem of philosophy of science in economics. The question is not whether economists occasionally make mistakes. Every science makes mistakes. It is not whether economics makes simplifying assumptions. Every science does that too. The question is whether economics has increasingly exempted its foundational assumptions from the evidentiary standards routinely applied elsewhere in science, and used this as a filter of which avenues of investigation are permissible for research. It is whether economics, and therefore public policy, is currently being dictated by a sphere of knowledge that owes more to faith and dogma than to science and evidence.

That is the scandal nobody talks about.

This essay will argue that modern macroeconomics has become an intellectual oxbow lake (or epistemic enclave) – once connected to the broader river of scientific inquiry, it gradually became isolated as neighbouring disciplines transformed their understanding of cognition, behaviour, information and adaptation.

Figure 1  Oxbow Lake formation

A river changes course, leaving behind a body of water that remains active yet no longer connected to the main flow. Modern macroeconomics increasingly resembles such a formation: internally sophisticated, highly productive, but progressively detached from the scientific developments occurring around it.

The economics discipline remained active. It remained mathematically sophisticated. It remained internally productive. But it became progressively disconnected from the scientific discoveries occurring around it, and it proceeded to shelter itself from the mountains of evidence discovered in adjacent sciences.

The result is a body of theory that is now more committed to preserving its assumptions than testing them.

The evidence for this claim can be found throughout the key issues facing the field:

• in economics’ theory of mind.
• in its missing evolutionary origin story.
• in the “as if” defence of rationality.
• in representative agents, equilibrium models and exogenous growth.
• in the profession’s response to behavioural economics.
• And in the growing accumulation of anomalies that Thomas Kuhn would immediately recognise as the warning signs of a paradigm under strain.

The purpose of this essay is not to argue that economics should be abandoned.

The purpose is to argue that economics must once again become scientific.

To do that, it must begin where every successful science begins. Not with assumptions about how people ought to behave. But with evidence about how they actually do behave.

2.   Economics Adopted a Theory of Mind

Economists often present utility maximisation as a mathematical convenience. A modelling assumption. A simplifying abstraction. Even a form of ‘null hypothesis’ from which any ‘behaviour’ must be judged as a deviation, abstraction or perversion.

But this radically understates what is actually being claimed.

Utility maximisation is not merely a mathematical framework, and is far from a neutral ‘null hypothesis’. It is a theory of cognition.

In order that the rational agent of mainstream economics performs its assumed functions whether consciously, subconsciously, using a single algorithm or a collection of behaviours that behave “as if” they were a conscious protocol, there is a collection of properties of such a mind that are indispensable. These include:

• store and retrieve information about the world in general, not limited to the local environment it directly experiences;
• estimate the likelihood of future states of the world through some sophisticated model of causation or frequency of states of unobserved future data;
• construct preferences over those future states;
• compare alternative courses of action, relying on a model of how the behaviour of the individual and its fellow individuals impacts future states of the world at large;
• evaluate costs and benefits across time based on a single, unified, consistent base of value applicable to all experiences, which economists call ‘utility’;
• model the consequences of possible choices not only in terms of causation of future states of the world at large, but also in terms of the utility-value capturable by the organism;
• make decisions that maximise expected future captured utility;
• do all of this under conditions of uncertainty that result from bounded access to information.

These are not mathematical operations. They are cognitive operations that must be performed in the ‘wetware’ (the biological equivalent of hardware) of actual living organisms.

The moment a theory attempts to explain how people make choices, it leaves the realm of pure mathematics and enters the realm of psychology and cognition, neurology and information processing, biology and evolution. Economics may not have intended to become a science of the mind, but it did so the moment it adopted the rational optimising agent as its foundational assumption.

This fact has surprisingly profound implications.

If economics is making claims about cognition, then those claims must be evaluated according to the same standards applied everywhere else in the ‘hard’ sciences. A theory of human decision-making cannot be exempted from evidentiary standards simply because it appears inside an economic model.

A neuroscientist proposing a new model of cognition would be expected to provide evidence. A psychologist proposing a new theory of decision-making would be expected to test it experimentally. An evolutionary biologist proposing a cognitive adaptation would be expected to explain how it arose through natural selection, and to show evidence that it exists in the species being studied. It is not enough to propose it as a mathematical abstraction.

Why should economics be exempted from the evidentiary standards of the broader scientific project?

This question becomes increasingly uncomfortable when we compare economics with the sciences that have developed alongside it.

Over the past century, psychology, neuroscience, evolutionary biology, cognitive science and information theory have transformed our understanding of how minds work.

These disciplines have repeatedly discovered:

• bounded cognition;
• limited information;
• heuristic reasoning;
• social learning;
• status sensitivity;
• computational constraints;
• adaptive shortcuts.

The resulting picture is messy, imperfect and provides a naturalistic explanation of cognition.

Yet at the centre of modern macroeconomics sits a very different character:

• An agent with stable preferences.
• An agent capable of evaluating alternative future states.
• An agent capable of comparing lifetime trajectories.
• An agent capable of solving optimisation problems under uncertainty.
• An agent whose behaviour can be represented by the maximisation of a utility function.

The question is not whether such a construct is mathematically elegant.

The question is whether it bears any meaningful resemblance to the cognitive architecture revealed by the sciences that study cognition based on evidence in the empirical world.

A further question arises when we remember that humans are actually biological in nature, and our brains developed through an adaptive process of evolution by natural selection. Economists begin with a theory of cognition that is simply assumed to exist a priori. No attention is paid by the economics profession to how such a cognitive architecture could have arisen through the same evolutionary processes that produced the organism within which it is asserted to exist.

This essay argues that this core assumption of economics is increasingly difficult to defend, in the face of evidence in adjacent sciences.

The problem is not that economics adopted a theory of mind. Every science requires theories. The problem is that economics adopted one theory of mind incompatible with evidence and more rigorous theories of cognition in adjacent sciences, and then largely insulated it from the scientific discoveries that should have forced its revision.

That is how a discipline becomes an intellectual oxbow lake. It is how a useful simplification hardens into dogma. And it is why modern macroeconomics increasingly finds itself in conflict with the sciences that actually reveal how human cognition works.

3.   The Missing Evolutionary Origin Story

In every serious science, explanations have origins.

Biologists do not merely observe that eyes exist and then declare the problem solved. They seek to understand how eyes evolved. They identify intermediate stages, reconstruct evolutionary pathways, and explain why each stage conferred an adaptive advantage. The extraordinary explanatory power of evolutionary biology comes not simply from describing what exists, but from explaining how it came to exist.

The economic profession has resisted any attempt to subject its foundational assumptions to the same scrutiny.

The rational optimising agent is generally introduced as though its existence were self‑evident. Yet if this agent is more than a mathematical convenience, if it is genuinely a claim about human cognition, then it requires exactly the same kind of explanation that we would demand of any other proposed cognitive adaptation.

Where did it come from?

Taking Homo Economicus Seriously

Let us assume, for the sake of argument, that mainstream economics is correct.

The rational optimising agent is not a metaphor. It is a reasonably accurate description of human cognition (including, potentially, “as if” aggregate unconscious behaviour).

What follows?

Such an agent would require a remarkable collection of cognitive capabilities stated above. This is an extraordinarily sophisticated cognitive architecture.

We need to explain how such a sophisticated capability came to exist.

The Evolutionary Standard

Evolutionary biology imposes a demanding standard on explanations. Complex adaptations do not appear suddenly, fully formed. They emerge gradually, and they leave evidence of each intermediate form.

The theory of natural selection requires that every intermediate stage must be both viable and useful.

Figure 2                      Evolution of the Eye

This is why the evolution of the eye is considered one of the triumphs of modern biology. We do not merely possess a theory. We understand a documented, evidenced pathway of development that provides an adaptive advantage at each step, and so explains the process by which organisms move from simple adaptations to complexity through a gradual process of natural selection of adaptive change. We can observe each step in existing species, providing an observable evolutionary pathway.

What is the equivalent pathway for the rational optimising agent?

Which component evolved first?

• The utility function?
• The information gathering system?
• The forecasting mechanism?
• The optimisation process?
• The representation of future states?
• At what point did our ancestors begin solving intertemporal optimisation problems?

These questions are rarely asked within economics. They are unavoidable in any genuine science of cognition or biology.

The Irreducible Complexity Problem

The challenge becomes even more difficult when we deconstruct the asserted components of Homo Economicus’ cognitive abilities.

• An optimisation algorithm has little value without information.
• A forecasting system has little value without a way of evaluating alternative futures.
• A utility function has little value without a mechanism for constructing and comparing future states.
• An information-gathering system has limited value if the organism lacks the capacity to process and use that information.

Each component is logically dependent upon the existence of the others. Their value arises primarily from their integration into a complete system.

This creates a profound evolutionary puzzle.

What adaptive advantage is provided by a partially completed utility-maximising architecture?

If we give serious consideration to these questions, they appear to present an insurmountable barrier to the evolution of lifetime utility maximisation. The utility maximising algorithm appears irreducibly complex – each part depends on each of the others in order that it is able to confer any value whatsoever. As a result, there does not appear to be a process of gradual evolutionary development that can provide a benefit at each step. Homo Economicus must appear in a single step, fully formed, as if by magic. The cognitive equivalent of the cyclone in a junkyard spontaneously creating a jumbo jet.

Consider the information requirements alone. A hominid attempting to gather sufficient information to estimate long-run opportunities, risks, and future resource states would incur substantial costs in time, attention, and exposure to danger. The adaptive value of exhaustive information gathering is therefore far from obvious, particularly in highly uncertain environments where much of the required information is fundamentally difficult and expensive to collect.

A hominid attempting to gather information for lifetime utility optimisation would have been eaten by a lion long before moving from information gathering to decision making. He would have failed to pass on his genes, nullifying any asserted benefit from rationality. Evolution rewards simple adaptive solutions, not rigorous rationality.

If, on the other hand, lifetime utility maximisation evolved in a step-by-step process, like the eye, then what are the useful intermediate steps?

Biologists can answer these questions for each evolutionary pathway and increasing for cognitive abilities.

Economists have yet to even pose the question in relation to Homo Economicus.

Where Are the Transitional Forms?

There is a second problem.

In evolutionary biology, pathways are not merely hypothetical. They leave traces.

We can observe simpler forms of vision in living species. We can identify intermediate structures. We can compare related organisms and reconstruct how complexity emerged through time.

There exist extant species with all of the transitional forms of primitive eyes that provide the explanatory pathway of evolution – a light sensitive patch, a primitive concave hollow, a camera obscura, a primitive lens (see Figure 2           Evolution of the Eye). In fact, this process of evolution can be observed to have occurred several times in the fossil record independently.

If a sophisticated utility-maximising architecture of Homo Economicus evolved gradually, we should expect to observe transitional forms. Where are they?

Do gorillas systematically survey the future distribution of resources across their territory before determining optimal foraging effort? Do chimpanzees formulate plans stretching decades into the future? Do bonobos compare alternative lifetime trajectories and select the one with the highest expected utility? Do any of our close evolutionary relatives display primitive forms of the cognitive machinery assumed by modern macroeconomics?

There is simply no evidence for this.

What we observe instead are behaviours that look strikingly different:

• social learning;
• imitation;
• status sensitivity;
• local information gathering;
• bounded planning horizons;
• adaptive heuristics;
• satisficing behaviour.

These look very much like the sort of cognition evolutionary theory would predict if human cognition were characterised by simple evolved heuristics and satisficing.

They do not resemble primitive versions of lifetime utility maximisation.

The Asymmetry

This creates a remarkable asymmetry.

Heuristic cognition possesses:

• a plausible evolutionary pathway;
• observable intermediate forms;
• comparative evidence across species;
• extensive experimental support.

Lifetime utility maximisation possesses:

• no widely accepted evolutionary pathway;
• no obvious transitional forms;
• no comparative evidence among related species;
• substantial contradictory evidence from behavioural research.

Yet mainstream economics treats lifetime utility maximisation as the default account of human decision-making. Why?

Natural selection does not optimise for rationality. It optimises for reproductive success. Any claim that evolution produced lifetime utility maximisation therefore requires an explanation of why such a costly cognitive architecture would have been favoured over simpler heuristic alternatives.

The challenge facing economics is therefore not:

Why don’t people optimise perfectly?

The real challenge is: “What evolutionary process produced a mind capable of lifetime utility optimisation?” Until a credible answer is proposed, utility maximisation should be regarded not as an established scientific description of human cognition, but as a heroic assumption.

4.     The “As If” Defence

The “As If” defence does not reference a monosyllabic guttural utterance of a defiant teenager.  It is a defence offered by economists, initially popularised by Milton Friedman, that has similar intellectual value to its namesake.

Economists in Friedman’s tradition point out that economics does not require people to literally solve optimisation problems. Utility maximisation, they will argue, is merely an “as if” model. People need not consciously calculate utility. They merely behave as if they do.

This is perhaps the strongest defence available to mainstream economics, so let’s give it the attention it deserves – a few paragraphs. We will find that this defence, despite being relied upon by the economics profession to protects its sacred hypothesis, actually fails in two different ways.

Problem 1: The Empirical Test

The purpose of an “as if” model is not to be literally true.

Its purpose is to be useful.

Physicists routinely employ idealised models. No physicist believes gases consciously solve differential equations. No evolutionary biologist believes animals perform explicit optimisation calculations. These models are accepted because they generate useful explanations and successful predictions.

The legitimacy of an “as if” model therefore depends entirely upon performance.

Unfortunately for economics, the empirical record is increasingly difficult to reconcile with the claim that people behave as if they are utility maximisers.

Behavioural economics is an entire field whose name belies the fact that economists separate actual observed behaviour into a subfield that they hold separate to economic theory. That’s one way to ignore evidence – name it a sub-field that can be conveniently ignored!! Behavioural economics has spent decades documenting systematic departures from the behaviour predicted by rational optimisation, which has been systematically ignored by keepers of the hegemonic mainstream economic theory.

Behavioural economics demonstrates clearly that human beings exhibit:

• framing effects;
• anchoring;
• loss aversion;
• reference dependence;
• social conformity;
• status competition;
• bounded rationality;
• satisficing rather than optimisation.

These are not rare anomalies. They are among the most robust findings in the behavioural sciences.

The standard response within economics has been to treat such findings as deviations from an underlying rational benchmark. The profession did not reject behavioural economics, nor counter it with contradictory evidence. It compartmentalised it. Treated it as a departure from orthodoxy.

But science requires us to ask a different question: “At what point do enough deviations cease to be deviations?” “At what point do the exceptions become the rule?”

The “as if” defence only works if people actually behave approximately as if they optimise.

I’m calling it. The evidence is now in. Homo Economicus fails the empirical “as if” test.

Problem 2: The Model Became a Gatekeeper

The deeper problem is not that utility optimisation survives despite contradictory evidence. The deeper problem is that optimisation has become the gatekeeper through which alternative theories are expected to pass.

Behavioural economics is not new. Herbert Simon introduced bounded rationality in the 1950s. Decades of research have since explored:

• satisficing;
• heuristics;
• social learning;
• adaptive behaviour;
• non-optimising decision processes.

Yet the foundations of mainstream macroeconomics remain largely untouched by the mountain of evidence that contradicts their central organising principle.

Why?

Because alternative behavioural theories are rarely permitted to stand on their own foundations.

Instead, alternatives to rational utility optimisation are routinely required to explain how they can be incorporated into optimisation frameworks.

The question becomes: “How can bounded rationality be added to utility maximisation?” rather than: “Why should utility maximisation remain the foundation at all?”

Or:

“How can behavioural findings be reconciled with rationality?” rather than: “Does the evidence support rationality in the first place?”

Notice what has happened.

The optimisation model has ceased to function as a hypothesis. A hypothesis is something tested against evidence. Instead, optimisation increasingly functions as a constraint on admissible theories.

Evidence is filtered through the assumption, rather than the assumption being tested against the evidence.

Mainstream economics has quietly, but emphatically inverted the scientific process.

What would Poseidon do?

Imagine that oceanography had developed around the assumption that the oceans behave as if they are governed by Poseidon – a capricious demi-God demanding libations and calling up tempests when unsatisfied.

Over time, researchers discover that:

• tides are influenced by the gravitational attraction of the Moon and Sun;
• ocean currents emerge from temperature gradients;
• circulation patterns depend upon salinity, wind systems, and planetary rotation.

A scientific discipline should ask: “Does the evidence support these explanations?”

Instead, imagine being told:

Fascinating work. But before publication, could you explain how your theory accords with Poseidon’s dominion over the seas?

Or:

Which tine of Poseidon’s trident is responsible for the gravitational affects you describe?

The problem would not be that Poseidon began as a useful metaphor.

The problem is when the metaphor is imposed as a gatekeeper on scientific development.

At that point, the originally useful “as if” hypothesis is no longer serving scientific inquiry. It is constraining it.

Yet this is uncomfortably close to the role utility optimisation now occupies in macroeconomics. Behavioural evidence must explain how it accords with rationality. Satisficing must explain how it accords with utility maximisation.

Alternative theories of decision-making must explain how they accord with rationality in order to be admitted for publication in economics journals.

The original hypothesis is serving as gatekeeper over scientific progress.

5.    The Great Averaging Machine

For all its mathematical sophistication, modern macroeconomics possesses a curious habit.

When reality becomes complicated, economists average it away.

People differ enormously in income, wealth, aspirations, risk tolerance, access to information, access to credit, social circumstances and countless other dimensions. These differences are not a minor feature of the economy. They are the economy.

Yet mainstream macroeconomics has spent decades treating heterogeneity not as a phenomenon requiring explanation, but as an inconvenience requiring simplification. This tendency reaches its most extreme form in the representative agent.

The representative agent is one of the most remarkable inventions in economics. An economy containing millions of individuals with different circumstances, incentives and opportunities is replaced by a single fictional individual who earns average income, consumes average consumption, saves average savings and behaves according to average preferences.

The result is a model that is easier to solve mathematically. The question is whether it still describes anything useful.

To appreciate the problem, consider its application elsewhere. No biologist would attempt to understand biodiversity by studying a representative organism. No ecologist would attempt to understand a rainforest by studying a representative tree. No chemist would attempt to understand chemistry by assuming all elements were identical and then modelling a representative atom.

Complex systems derive many of their most important properties not merely from their components, but from the interactions between heterogeneous components. Chemistry exists because elements behave differently. Ecology exists because organisms behave differently. Heterogeneity is not noise. It is often the phenomenon requiring explanation.

Yet economics repeatedly treats variation as noise and the average as reality.

This tendency extends well beyond representative agents. Income distributions become average incomes. Wealth distributions become average wealth. Consumer behaviour becomes access to a common consumption function. Business decisions become representative firms using a common production function. Expectations become representative expectations, with heteroskedastic noise.

The economy is repeatedly passed through what might be called a great averaging machine.

The output is mathematically elegant. The policy prescriptions are simple and reductive. The question is whether anything important survives the process. Because averaging does not merely simplify. It destroys information.

If one person earns $20,000 a year and another earns $2,000,000, their average income tells us almost nothing about either of them. The average conceals the very heterogeneity that may be driving the behaviour we wish to understand.

The preference for representative agents is not accidental. It emerges naturally from another commitment at the heart of mainstream economics: general equilibrium. Equilibrium is easier to analyse when agents are similar. Aggregation is easier when behaviour is uniform. Distributional differences are easier to ignore when they are assumed not to matter.

Reality, unfortunately, does not cooperate.

Real economies are heterogeneous, adaptive, path-dependent, networked and evolutionary systems. Financial crises, business cycles and asset bubbles do not emerge from the behaviour of representative households or representative firms behaving rationally with perfect foresight. They emerge from interactions between heterogeneous participants operating within complex systems, making decisions using heuristics founded on bounded access to partial information.

Economics has long justified its simplifications on grounds of tractability. Solvable models are undoubtedly valuable. But there is a danger in confusing tractability with explanation.

A model that achieves elegance by removing the very phenomenon requiring explanation is not a scientific success. It is merely mathematical elegance elevated over empirical reality. Over scientific accuracy.

The central challenge for a new scientific economics is not to explain the behaviour of an average person who does not exist. It is to explain the behaviour of real people who do. That requires understanding heterogeneity, not averaging it away.

6.   The Miracle Step

Every discipline develops simplifying assumptions. Sometimes they are useful. Sometimes they illuminate deeper truths.

But there is a difference between simplifying a problem, versus ignoring it.

Modern growth theory provides a striking example.

Figure 3                      The Exogenesis Problem

“I think you need to be more explicit in Step Two.”

Economic growth is one of the most important phenomena in human history. Why do living standards rise? Why does productivity increase? Why do societies become wealthier over time?

These are scientific questions.

Yet modern growth theory has a remarkable tendency to assume away the very thing it claims to explain.

The classic example is the treatment of productivity growth. Solow demonstrated that capital accumulation alone could not explain long-run growth. Much of the remaining growth was attributed to a residual, later termed total factor productivity. The question then shifted from “Why does growth occur?” to “Why does productivity increase?”

Later theories shifted attention again, towards innovation, knowledge and human capital. Useful concepts, certainly. Yet they largely relocate the mystery rather than resolve it.

What drives innovation? What drives knowledge creation? What drives the continual expansion of productive capacity?

The deeper question remains unanswered.

In this respect, growth theory reveals a methodological habit that we have already encountered elsewhere in economics. When faced with difficult phenomena, economists often transform the phenomenon requiring explanation into an assumption. The evolved human mind becomes an assumption of utility maximisation. Empirical human diversity becomes an assumed representative agent. Economic growth is “explained” by productivity growth – the same rose by another name.

Again and again, the difficult question disappears and a mathematically tractable assumption takes its place to make the model tractable.

There is nothing wrong with simplifying assumptions. Every science relies upon them. The problem arises when the assumption begins to occupy the very space where explanation should exist.

Science advances by explaining phenomena. Renaming them is not the same thing as explaining them.

Until economics can explain why productive capacity continually expands, the miracle remains at the centre of the theory. It may have acquired respectable names and elegant mathematics, but it remains a mystery, nonetheless.

7.     How Economics Became an Intellectual Oxbow Lake

At this point a reasonable question arises. How did economics arrive here?

How did an intellectual tradition that began with moral philosophy, study of productive capacity and methods, careful observation of markets, institutions and human behaviour become committed to a theory of cognition with no accepted scientific origin story, insulated from contradictory evidence, and protected by methodological defences that would fail the test of scientific credibility in any real science?

The answer is certainly not that economists are stupid or dishonest. There is no secret cabal of economists with Machiavellian desires to impose poor public policy on the world, despite that being the appearance.

The answer is more subtle, and perhaps more troubling.

Economics became trapped by its own history. Ultimately, it’s an historical anomaly that owes its origin to Cold War era paranoia and an accident of historical timing.

The Questions of an Earlier Era

Scientific theories do not emerge in a vacuum. They are shaped by the questions their age regards as important.

In the nineteenth century, economics was still deeply connected to philosophy, politics and moral inquiry. Adam Smith, David Ricardo and John Stuart Mill saw little distinction between economics and the broader moral enquiry of how societies should organise themselves.

The early twentieth century changed everything. The dominant political conflict of the time became the struggle between markets and central planning. This was often characterised as a struggle between freedom and authoritarian rule.

Economists found themselves confronting an enormous intellectual challenge.

Could decentralised markets coordinate the activities of millions of people? Could individual choice generate social order? Could capitalism survive competition from centrally planned systems?

These were not unreasonable questions. Indeed, they were among the defining questions of the early twentieth century. And western-centric economics was motivated to find answers in the affirmative. Economics became the front line of the intellectual battleground where ‘freedom’ was defended from authoritarianism.

The tragedy is that the tools developed to answer questions of that era gradually became the foundations of an entire discipline, and remained ossified long after the Cold War politics had been fought out.

In the trenches, the defence of markets against central planning rewarded theories built upon:

• methodological individualism;
• rational decision-making;
• equilibrium;
• spontaneous order.

Questions of behaviour, social influence, class, status and distribution increasingly migrated elsewhere, into sociology, psychology and political science. Within the economics profession, even asking such questions was viewed with suspicion. A form of intellectual McCarthyism took hold in economics.

The political environment of the Cold War created strong incentives for theories emphasising markets, optimisation and methodological individualism, while making alternative approaches centred on behaviour, institutions and distribution intellectually suspect.

The result was not ideological conspiracy. Not a banality of evil, so much as an ideological reward structure that increasing rewarded ideological purity over scientific accuracy.

A discipline responding to the pressures, incentives and intellectual battles of its time through an ideological lens. Economists pursued the questions that seemed most important. The assumptions that proved useful in those battles survived. Economics became dominated by the theories that could push back against communism because that was the task society required of them.

Eventually those theories became orthodoxy. And questioning them became a particular kind of blasphemy.

The Quest to Become a Hard Science

Economics also faced another challenge. It wanted scientific legitimacy.

Physics had become the benchmark against which all sciences were judged. Physicists where household names – Einstein personified the concept of genius. Physics possessed mathematical elegance, predictive precision and universal laws. Naturally, economists sought the same status.

Mathematics increasingly became the language of economics. This was not necessarily a mistake. Formal reasoning often improves clarity. But over time a subtle shift occurred. The objective ceased to be understanding human behaviour or the behaviour of economic systems in all of their complexity.

Constructing mathematically elegant models became the end in itself.

Questions that were difficult to formalise began to disappear from view, or be simplified away in the service of tractability.

Behaviour became less important than optimisation. Institutions became less important than equilibrium. Distribution became less important than efficiency. History became less important than generalised equilibrium.

The discipline gradually learned to prefer questions that mathematics could answer over questions that reflect reality.

The River Changed Course

The problem is that science did not stand still.

Psychology changed. Neuroscience changed. Evolutionary biology changed. Cognitive science changed. Information theory changed. Behavioural science changed.

Across all of these scientific disciplines, a new understanding of human cognition backed by empirical evidence took hold.

A new science of evolutionary psychology developed, and was able to explain human cognition with a high degree of fidelity, but it arrived too late to inform economics. By the time it existed, economic orthodoxy had already ossified into a dogma that could not admit the new science of human cognition.

Human beings were revealed to be:

• bounded rather than omniscient;
• adaptive rather than optimal;
• heuristic rather than rational;
• social rather than isolated;
• satisficing rather than maximising;
• endowed by evolution with ‘hard-wired’ evolved heuristics, rather than ‘blank‑slate’ rational information processors.

The river of scientific discovery had changed course, leaving economics as an intellectual Oxbow Lake. Economics was largely left isolated on a former bend, separated from the main stream of scientific progress.

Instead of rebuilding its foundations in response to these discoveries, the discipline insulated itself from them. The mathematics became more sophisticated. The discipline turned inward, and began to treat adjacent disciplines with scepticism or suspicion.

The assumptions underpinning economics remained rock solid, and impervious to empirical facts.

8.      Kuhn’s Warning

Defenders of modern macroeconomics will often point out that all sciences possess anomalies. No theory is perfect. Physics contains unanswered questions. Biology contains unresolved disputes. Climate science contains uncertainty.

Why should economics be held to a higher standard?

The answer is that it should not.

Economics should be held to the same standard.

It is precisely when we apply that standard that the work of Thomas Kuhn becomes relevant.

In The Structure of Scientific Revolutions, Kuhn argued that science progresses through paradigms: shared frameworks that define acceptable assumptions, methods and research questions. Most scientific work takes place within these frameworks. Scientists do not constantly revisit foundational assumptions. They solve problems and extend knowledge within the accepted paradigm. Kuhn called this normal science.

Problems arise when anomalies begin to accumulate.

An anomaly is an observation that does not fit comfortably within the prevailing paradigm. Initially, anomalies are often ignored or explained away. But over time a question emerges: “Are these isolated exceptions, or symptoms of a deeper problem?”

A paradigm enters crisis when its anomalies begin challenging its foundations rather than merely its details. Modern macroeconomics is increasingly surrounded by such anomalies.

Behavioural science reveals persistent departures from rational optimisation. Evolutionary theory offers no support for the cognitive architecture assumed by utility maximisation. Representative-agent models struggle to account for heterogeneity. Financial crises emerge from complex interactions that equilibrium models incapable of representing mathematically. Distributional dynamics continue to matter despite repeated attempts to minimise their importance.

Individually, none of these observations is fatal.

Collectively, they tell a story.

Many of the anomalies concern the foundational assumptions of economics.

Kuhn observed that disciplines approaching crisis often exhibit growing fragmentation as alternative approaches emerge around the edges of the dominant paradigm. This pattern is clearly visible in economics.

Behavioural economics, complexity economics, agent-based modelling, evolutionary economics, institutional economics and network economics all represent attempts to move beyond the traditional assumptions of rational optimisation, representative agents and equilibrium. These approaches differ substantially from one another, but they share a common dissatisfaction with the foundations of mainstream macroeconomics.

At the same time, the core assumptions of the discipline remain remarkably stable. Rational optimisation, utility maximisation, representative agents and equilibrium continue to occupy a privileged position despite mounting contradictory evidence. Rather than reconsidering these assumptions, economists often incorporate anomalies through additional frictions, shocks, preferences and qualifications.

This does not necessarily mean the framework is wrong. But that’s one possibility. And it is precisely the pattern Kuhn and Lakatos identified in paradigms under strain.

For decades economists treated departures from rationality as anomalies. Perhaps it is time to reverse the question.

What if the anomaly is no longer the behaviour? What if the anomaly is Homo Economicus itself?

What if the persistent failure of reality to conform to the assumptions of macroeconomics is not evidence that reality is unusually complicated, but evidence that the field is in need of a new central organising theory?

That possibility is deeply uncomfortable. Paradigms do not surrender easily. Careers, journals, departments and entire intellectual traditions become invested in their preservation.

Which is precisely why scientific revolutions are so difficult.

Kuhn’s warning was that paradigms rarely collapse because their supporters abandon them voluntarily. They collapse when the accumulation of anomalies eventually makes the existing framework impossible to defend.

The uncomfortable question facing economics is whether that process has already arrived.

9.     Back to Adam Smith

If the preceding arguments are correct, then economics faces a deeper problem than utility functions, representative agents or equilibrium models.

It faces a problem of purpose.

Modern economics often behaves as though growth, efficiency, utility and equilibrium are ends in themselves. Yet these concepts were never meant to be ends. They were meant to be tools.

The question that originally gave economics its importance was far simpler:

What is the economy actually for?

Adam Smith would have recognised the question immediately. Smith was not attempting to maximise utility functions or prove welfare theorems. He was a moral philosopher concerned with prosperity, institutions and human flourishing. The economy was a means. Human wellbeing was the end.

Somewhere along the way that distinction became blurred.

Growth became a goal rather than a phenomenon. Efficiency became a goal rather than an analytical result. Utility became a foundational belief rather than a hypothesis.

The question, “What promotes human flourishing?” was gradually replaced by the question, “What maximises the assumptions embedded within the model?“

The distinction matters because economics cannot determine society’s goals. Science can tell us what is. It cannot tell us what ought to be.

Questions of prosperity, sustainability, equity, resilience and freedom remain political and moral questions. Economics can help us evaluate the consequences of pursuing different goals, but it cannot relieve society of the responsibility of choosing its goals.

A more scientific economics would therefore be a more modest economics.

Its task would not be to provide moral truth or ideological certainty. Its task would be to understand how economic systems actually function, forecast the consequences of alternative policies, and revise its theories when confronted by contradictory evidence.

That is not a revolutionary vision.

It is a return to the tradition from which economics emerged. A return to the idea that markets are institutions designed to serve people. A return to the view that economies are social systems rather than optimisation problems. A return, ultimately, to Adam Smith’s question:

How can human societies organise themselves in ways that promote flourishing?

That question remains as important today as it was when economics began.

10.   Rejoining the River

The argument of this essay has not been that economics is useless.

Nor has it been that generations of economists were unintelligent, irrational, or acting in bad faith. The problem is more subtle.

Economics has progressively isolated itself from the wider scientific project.

A theory of cognition was settled on by economists before modern sciences had matured. Before psychology, neuroscience, evolutionary biology and information theory. As those adjacent sciences transformed our understanding of how minds work, economics largely retained its own foundational assumptions. The resulting framework became increasingly internally sophisticated, while becoming increasingly disconnected from the sciences studying the very phenomena upon which it relied.

But economics became progressively detached from the main current of scientific discovery. It turned inward rather than embracing and incorporating these new scientific discoveries.

The Choice Ahead

Economics now stands at a crossroads.

One path leads toward ever more elaborate versions of the existing paradigm:

• more optimisation;
• more calibration;
• more frictions;
• more representative abstractions;
• more mathematical refinement;
• more assumptions of invisible, unfalsifiable shocks and preferences.

The other path leads back toward the scientific mainstream. It asks economists to engage seriously with:

• behavioural evidence;
• evolutionary theory;
• complexity science;
• network effects;
• heterogeneity;
• institutions;
• empirical reality.

One path seeks to preserve a familiar framework. The other seeks to understand the real world.

Only one of those paths deserves to be described as scientific.

Back into the Current

The central failure of modern macroeconomics is not mathematical.

It is philosophical.

The discipline gradually exempted its foundational assumptions from the evidentiary standards routinely applied elsewhere in science. It built an increasingly elaborate intellectual structure upon those assumptions. And then it mistook the structure of its assumptions for reality.

The remedy is not complicated.

Economics must rejoin the scientific enterprise. It must once again allow its theories of human behaviour to be constrained by the sciences studying human behaviour. It must seek to explain heterogeneity rather than average it away. It must explain growth rather than assume it. It must test assumptions rather than defending them.

It must treat evidence as master and theory as servant.

Above all, it must rediscover the humility that lies at the heart of every successful science.

The river of scientific discovery has continued to flow. The question for economics is whether it wishes to remain an intellectual oxbow lake, or whether it is finally prepared to rejoin the broader scientific endeavour, and take up a paddle and row.