Is Economics a Pseudoscience?

Sorting through the criteria for scientific status, including explanatory modelling, mathematical rigour, and a culture of critical thinking

As to whether economics is genuinely scientific, the easy answer is that you can define “science” as you like to include different disciplines under that heading. The German word for “science” includes the humanities, not just the social sciences. In English, the meaning of “science” has been influenced by positivists who credited the hard, natural sciences as being the model for all knowledge, not just for other sciences.

But there’s a deeper problem than just the possibility that our labels are unclear or flexible. We can’t confront the question of what economics is without acknowledging that the Information Age might better be known as an age of disinformation. For example, sociologists speak of “agnotology,” of the need for studying merchants of doubt who manipulate public opinion with misleading data and “fake news” for commercial gain.

Pseudoscience abounds, therefore, and the discipline of economics is especially susceptible to political and business influences because its subject matters — money, wealth, poverty, taxes, business, employment, buying and selling, inflation, societal stability — are of great interest to the public.

At first glance, mind you, economics is the most scientific of all the social sciences since economists devise models to explain statistical patterns, and their discourse is overtly mathematical, which seems to indicate that economists are rigorously objective and precise in their judgments. There’s even a Nobel Prize in economics. Indeed, surveys before and after the 2008 crash on Wall Street show that most economists agree that “economics is the most scientific of the social sciences.”

Yet if we just look at a relevant definition of “science,” we see some important characteristics that economics lacks. Science, then, is “systematic knowledge of the physical or material world gained through observation and experimentation.” This sense of the word excludes economics on several grounds since economists deal with society rather than nature, and macroeconomists, at least, don’t make observations in a controlled, experimental setting.

Science and pseudoscience

That standard definition of “science,” though, would also exclude the social sciences, such as anthropology, psychology, political science, and sociology. The question, then, is whether there’s a worthy sense of “social science” that includes economics but distinguishes those sciences from the humanities or arts, let alone from pseudosciences.

The social sciences are supposed to be close enough to the natural sciences (to physics, chemistry, geology, biology, and astronomy) to be considered scientific, meaning that social sciences use the scientific method to generate objective knowledge of how society or the mind works. What, then, is that coveted method?

The heart of science is the testing of hypotheses with repeated experiments in which variables are controlled to enable patterns in the data to emerge, patterns which scientists compete to explain.

Scientific explanations are evaluated according to humanistic rather than more narrowly cultural or personal values. For instance, as implicit progressive humanists, scientists value logic and explanatory simplicity and fruitfulness since those values generate knowledge which is supposed to help our whole species. In short, the culture of science prizes critical thinking.

Finally, those tested explanations typically lend themselves to technological applications, which indirectly confirm the explanations’ reliability and contact with reality.

That culture of objectivity seems key to distinguishing science from the arts since artists are encouraged, rather, to express themselves, including their emotions, intuitions, and subjective opinions. Artists search for meaning and criticize society on moral grounds, whereas scientists aim to understand the facts, not to protect how anyone feels about them.

Thus, the heroic act in science is to falsify your cherished presumption, to submit to the best explanation and to sacrifice your fragile ego and personal agenda to advance the human endeavour. Indeed, in science there’s typically a conflict between the facts and our preferences. As the modern period emerged from the medieval one in the West, early modern scientists realized that nature might not be so intuitive as we’d long thought, that the Aristotelian framework had been sustained for centuries based on little more than dogmatic assent to Church authorities, which retarded human progress.

Indeed, outside of science there may be no progress. For thousands of years, the aim in society was stability, as we deferred to traditions and to our natural inclinations, and demonized deviations from those norms. The Scientific Revolution overcame that complacency, dispensing with religious dogmas on empirical grounds because scientists discovered ways for the facts to speak as loudly as possible for themselves.

In so far as the social sciences adopt that experimental method and that humanistic culture of self-sacrifice on the altar of critical thinking, and have some technological applications or at least broad utility, they might be deemed close enough to real science. Artists don’t conduct rigorous experiments to make sense of data. Moreover, the arts have a more expressive culture, one that explores subjectivity rather than objectivity, and they have more intangible benefits such as the inculcation of civility and good taste.

And pseudosciences such as astrology, numerology, and faith-healing likewise depart from science in those respects: these subjects don’t abide by experimental falsifications of hypotheses, don’t adhere to principles of critical thinking, and mislead the public rather than improving our living standard.

The difference between art and pseudoscience, of course, is that the artist isn’t pretending to be scientific.

Where, then, does economics fit into this analysis?

Economics as a social science

Again, macroeconomists can’t conduct experiments to test their assumptions. Instead, they rely on statistics, the historical record, and computer modelling. Statistics are notorious for their liability to be misused. Corporate advertisements routinely present statistics to sell products, and those facts and figures are invariably misleading. The problem is that instead of bypassing or dispensing with bias, statistics can merely bury the subjective element. Statistics can easily obfuscate an issue rather than clarifying it.

As for historical data, the problem is that while history can be construed as informally sorting through ways of organizing society, the variables are hardly controlled. If capitalism coincides with a boom-and-bust cycle, who’s to say that capitalism causes that cycle? There would have been numerous other factors that might have contributed to that pattern. Unlike repeated, controlled scientific experiments, then, history doesn’t enable the facts to speak for themselves. We understand history by interpreting it, which is as much an artistic endeavour as a scientific one.

There are policy implications of economic models, but no technological applications, which means that the models aren’t sufficiently comprehensive or reliable (or that society is loathe to apply them). Mind you, there are economic institutions such as corporations, banks, and the stock market which may be made more efficient with trading apps and the like. But those institutions are subject matters of economics, the rudiments of which pre-existed economic modelling; they’re not technological applications of economic knowledge or demonstrations of the rigor of economic methods.

So far, then, the scientific status of economics is questionable, at best, but that can be said about all the social sciences. There are, though, three other scientific merits economics might have: useful modelling, mathematical formulations, and an epistemic culture of critical thinking and problem-solving. Let’s look at each of those in turn.

Is economic modelling scientific?

Scientific modelling is analytical, meaning that instead of trying to explain everything at once, scientists divide the environment into its apparent parts. Special sciences take on a domain, period, or level of complexity, and they simplify by treating that subject as though it were perfectly isolated. That is, they abstract from how a system or process might interact with other systems or processes, and this simplification of the subject matter is best supported by experiments which isolate the variables in a controlled, artificial setting.

To know how a natural system, force, or element behaves on its own terms, with its inherent properties, minus the complexity of how the rest of the environment might interfere with that pattern, scientists isolate the system in the lab, or they perform a thought experiment (as Albert Einstein famously did) to imagine how the system would respond under extreme conditions. A scientific model, then, is a simplified representation that’s based on observations of what happens in the lab, or on abstract thinking.

Although macroeconomists can’t efficiently or morally experiment on societies to test their models, early modern economists attempted to emulate natural scientists by simplifying their subject matter. Classical and neoclassical economists abstracted from (or held as being equally irrelevant) the irrational aspects of human behaviour and imagined how an economy would work if everyone were perfectly rational, dedicated to maximizing their personal utility (as a corporation seeks to maximize its profits), and to act independently on full and relevant information in deciding what to sell or to buy.

The problem with that simplification is that it seems motivated less by an interest in understanding how economies work than in boosting the scientific credentials of economics by making the math tractable. Logic is formalizable, so if people were rational, you could use math to formalize that behaviour. Emotions, insights, fallacies, unconscious longings, and so forth aren’t so subject to a mathematical treatment because that side of our behaviour isn’t consistent. Arguably, then, that kind of economic modelling was self-serving for economists, rather than being justified by the humanistic ideal of critical thinking.

Indeed, there’s a difference between a simplification and an absurdity, the latter being an oversimplification, we might say. A model simplifies if the abstraction is justified. The best justification would be an experiment that isolates the variable in question, eliminating the others in a controlled setting. In that case, the model would reflect what’s observed in that artificial scenario that uncovers the essence of some phenomenon, such as a causal relation between properties, which the scientist represents with a law. The law of nature might be ceteris paribus, meaning that it holds for what the system would do by itself, were the system not embedded in its natural environment.

The model shouldn’t be so simplistic, however, that it no longer does justice to what’s being explained. Beyond that point, the model becomes a fantasy and its simplifications absurdities. That is, the model would be a distortion rather than a representation of reality. The difference between the two lies in the fact that a model should reflect norms observed in nature or in the laboratory, so that what’s abstracted away are exceptions to the rule, as it were, not the rule itself. Ignoring the system’s normal behaviour would be an Orwellian reversal of the scientific aim.

Arguably, though, the neoclassical model that underlies much current economic cheerleading for capitalism is Orwellian in that sense. Neoclassical assumptions are typically preposterous by design, and there’s no reason to think they uncover the essence of economic relations.

On the contrary, the abstractness and pseudo-exactitude of these models leaves out the historical and political contexts that vastly increase our understanding of what’s happening in any economy. Consider who could explain the state of the economy better, the economist with her equations or the journalist or historian who talks about the history of the relevant class struggles, businesses, and political parties.

As one critic points out, “Perhaps the most pernicious effect of the status of economics in public life has been the hegemony of technocratic thinking. Political questions about how to run society have come to be framed as technical issues, fatally diminishing politics as the arena where society debates means and ends.”

Moreover, pretending or insinuating that people are fundamentally rational and selfish, in line with the model of Homo economicus, is like saying Jews are greedy and they control the world. Either stereotype makes for a loaded, ideological simplification, not for a good-faith scientific one.

Elsewhere, I speculate that the observations that inspired the early modern economists’ individualism and utilitarianism were of none other than the leaders of capitalism, namely the robber barons who would have been more sociopathic than the average person, having been corrupted by their outsized wealth and power over society. It was the sociopathic captain of industry who fit the economist’s model of hyperrational selfishness. What the economists did, in effect, was project that special case onto all buyers and sellers in a capitalist market.

Even if that’s not so, however, economists rely more on thought experiments than on real-world ones, which means economists’ abstractions are liable to go astray since we can freely imagine fantasies and mistake them for an underlying reality. That would be fine if the fantastic hypotheses could be tested, yet that’s just what can’t be done in the macroeconomic case, and we’ll see in a moment some further problems with the falsifiability of economic models.

Even the IMF concedes that “An important feature of an economic model is that it is necessarily subjective in design because there are no objective measures of economic outcomes. Different economists will make different judgments about what is needed to explain their interpretations of reality.” Those subjective judgments are needed because there are few experiments to settle the issue.

Also, says the IMF, “All economic models, no matter how complicated, are subjective approximations of reality designed to explain observed phenomena. It follows that the model’s predictions must be tempered by the randomness of the underlying data it seeks to explain and by the validity of the theories used to derive its equations.”

Indeed, the IMF points out that economists are split between two camps on how to interpret economic models. The first insists “that the equations must assume maximizing behavior (for example, an agent chooses its future consumption to maximize its level of satisfaction subject to its budget), efficient markets, and forward-looking behavior.” These economists ignore how a policy change “itself alters agents’ behavior,” which means they abstract from history.

Other, more behavioural rather than pseudo-physical economists

favor a more nuanced approach. Their preferred equations reflect, in part, what their own experience has taught them about observed data. Economists that build models this way are, in essence, questioning the realism of the behavioral constructs in the more formally derived models. Incorporating experience, however, often means it’s impossible to untangle the effect of specific shocks or predict the impact of a policy change because the underlying equations do not explicitly account for changes in agent behavior. The gain, these same economists would argue, is that they do a better job of prediction (especially for the near term).

In a nutshell, these latter, renegade economists aren’t as susceptible to physics envy.

Keynesians and Neoclassicals

Most mainstream economists are divided between Neoclassical and Keynesian approaches. Neoclassicals assume the free market is the most efficient way of distributing resources and wealth, in that this arrangement reaches a natural equilibrium so that there would be a buyer for every seller and a seller for every buyer.

By contrast, Keynesians assume that capitalism has inherent inefficiencies so that the government needs to intervene to stabilize the market, such as by bailing out the economy when monopolies form and when they act recklessly and threaten national security by plunging the economy into a depression.

Is that division between economists like the difference between classical and quantum physics? Do these models merely have different scopes, forming a suboptimal patchwork of knowledge before the theorists figure out how those respective domains relate to each other? That is, just as theoretical physicists are searching for a theory of quantum gravity, might Keynesians and Neoclassicals be reconciled with a model that will one day subsume those rival perspectives?

Let’s return to that question after we consider the other two remaining scientific criteria.

Mathematical precision and obfuscation

The reasons for mathematically precise formulations are clear in the natural sciences. It’s not just that precision is always better than the ambiguity of natural language expressions. Precision may make for greater efficiency, but there’s a more fundamental basis for this resort to math in the natural, physical, or “hard” sciences, which is that their subject matters are objective.

Before modern science, mind you, the assumption was the opposite, that God made the universe as an artifact (as in the monotheistic traditions) or that natural processes are rife with purpose (as in the Aristotelian and Thomistic frameworks). Early modern scientists distinguished themselves by abandoning those presumptions and treating nature with more neutrality. What if natural processes were impersonal, amoral, and unplanned, and what if they operated according to brute physical regularity? If so, scientists could tame nature without fear of divine reprisal. Such was the Faustian or “Luciferian” agenda that’s inherent to scientific humanism.

In any case, extreme precision is useful especially if you’re dealing with an object in that sense, with something mindless and lacking in feeling or purpose. You model the object with math that cuts to the object’s physical structure, leaving out normative considerations (aside from those presupposed by the pragmatic, progressive agenda of humanism). That’s what math in the sciences is supposed to do, to bypass how we might feel about some natural system, or how we’d prefer to deal with it, and to represent what the system is as far as we can objectively tell. Natural language includes too many emotional connotations, archaic superstitions, and creative ambiguities to do justice to nature’s objective order.

Yet again, when we ask whether the economist’s use of math is scientific, we’re met with a problem. Certainly, the resort to mathematical precision is what we’d expect from a science, but pseudoscientists might likewise employ mathematical abstractions to seem more scientific than they are. Mathematical precision isn’t sufficient for scientific status, as the pseudosciences of astrology and numerology demonstrate.

And suspicion mounts as we observe that economists lack the basic scientific justification for objectifying their subject matter, since economists deal with society, not with nature. Ironically, then, the very characteristic that’s supposed to single out economics as the most scientific of the social sciences might have the opposite effect of establishing that economics is a pseudoscience.

After all, if society is filled with people and it expresses and embodies our collective emotions, plans, ideals, purposes, and meanings (as well as our rational calculations and knowledge), what could be the scientific point of objectifying all of that, of abstracting away from that subjectivity and treating the economy like a physical object with an exact, permanent, mindless structure that can be mathematically captured?

Early modern economists like Adam Smith, Karl Marx, and Alfred Marshall didn’t suffer from physics envy. They understood that as a social science, economics is closer to philosophy than to physics, so they didn’t pretend that economic questions are amoral; they didn’t hide the prescriptive side of their work with impenetrable math. Certainly, mathematical exactitude can clarify as it resolves ambiguity, but math can also obfuscate, as it does on Wall Street, hiding corporate frauds from the government regulators who are bereft of highly paid “quants.”

The principle of GIGO (garbage in, garbage out) applies here, as numerous critics of economists’ “mathiness” have pointed out. Mathematical formulations may ensure that the conclusion follows from the premises, avoiding fallacies of equivocation, for instance. But mathematical rigor can also deter an examination of the model’s assumptions, by burying them. However sophisticated and intimidating it might be, a mathematical demonstration is worthless if the model’s inputs are absurd, as they often are in economics. And again, their absurdity is by design because it makes the math tractable, which means the ideality makes economics seem like prestigious science.

For that matter, math can obfuscate because math is a useful fiction. Mathematical idealizations are abstract fantasies we evoke with our imagination. Some of those rules and other idealizations turn out to be useful in understanding natural patterns, and lots of them don’t. Thus, we can dream up mathematical formalisms to explore fantasy worlds.

The fact that neoclassical economics, for instance, is highly mathematical doesn’t mean the math corresponds to patterns in real economies. And if there’s no such correspondence, the economist’s ongoing reliance on math looks like pseudoscientific obfuscation. Far from dispelling ambiguity, the economist’s technicalities might hide an awkward truth, which is that the economist’s pro-capitalist discourse is more like a social Darwinian prescription than a scientific model of an objective structure.

Is the economist’s professional culture humanistic?

The crucial determinant, however, is surely the third criterion, which has to do with the epistemic culture in question. Again, modelling and mathematics don’t clinch economics’ scientific bona fides because pseudosciences can simplify and get lost in abstractions too. Apart from their technological applications, what distinguishes scientific models and precision and what preserves their authority is the humanistic culture of critical thinking that anchors the endeavour.

And here we encounter perhaps the biggest hurdle for economists, which is that their discipline is highly politicized. Thus, we can’t readily credit the leading economists, at least, with having scientific authority because we can’t be sure they haven’t been bought off by banks, hedge funds, and large corporations that benefit from — and that often help fund — their analyses.

To be sure, this is a much larger problem. The social sciences in general face a replication crisis, meaning that their results aren’t as easily replicated as they once were. There’s a rush to judgment and to publish because sensationalism furthers the academic’s career, and much social science is funded by corporations that want an immediate return on their investment. Certainly, these corporations lack the scientist’s humanistic values; what the corporate world’s after isn’t knowledge but propaganda to help sell its products such as pharmaceuticals, social media apps, and financial instruments.

Economists face immense pressure from the private sector because their subject is largely that very sector. This isn’t to suggest that lobbyists write economists’ papers the way they sometimes write laws for the government. The institutional capture needn’t be so comprehensive and conspiratorial. But there’s a revolving door in the most powerful capitalist societies, and the leading economists know there are lucrative opportunities to serve as consultants for banks, corporations, think tanks, and the like if they push a pro-corporate narrative with their work.

Here, finally, is why we shouldn’t expect much reconciliation between Keynesians and Neoclassicals anytime soon. True, these two approaches are partly objective in that they spell out the effects of certain economic policies. If you lower or raise taxes, for example, such and such will likely happen as a result, as best as can be determined from history and from the mathematical models. But these two approaches are far from being purely or scientifically objective.

What they are, rather, are selections of different social ends or goals, plus formulations of the efficient means of achieving them. The choice of those ends is subjective and normative, meaning that it expresses the economist’s values which are supposed to be irrelevant to science. Real scientists have values, too, of course. Indeed, real science is guided by humanistic values of critical thinking and perhaps by a loftier, progressive agenda, as I’ve said. But scientific methods filter out the more idiosyncratic biases to let the facts speak for themselves.

By contrast, the neo-Keynesian (socially progressive) and neoclassical values — namely socialist redistribution of wealth and socially Darwinian authoritarianism, respectively — are indispensable to these economic approaches in that they orient them, and the latter approach, at least, is obviously at odds with modern humanism. Cutting through the jargon, the real political divide between economists is just that between liberals and conservatives, which I argue elsewhere reduces to the conflict between humanists and premodern authoritarians.

In short, functionally modern economists want to help sustain a middle class with government intervention, because left to its devices, capitalism divides the rich from the poor, generating plutocracies and kleptocracies that endanger the social benefits of modernity. Thus, these economists question or reject the laissez-faire narrative that’s a large part of mainstream economics.

By contrast, the functionally premodern (“conservative”) economists accept that narrative because their values or vision of society lies elsewhere, in the distant past with the dominance hierarchies and brutal power pyramids that prevailed in the animal kingdom and in human societies for thousands of years. Capitalism, for them, is just a modern styling of that natural necessity.

Note that that’s not a conflict that can be settled with any mere empirical discovery. Free-market predictions have been falsified many times over, and the neoclassical, “conservative,” or “libertarian” economists still rush to defend the model not just because of their financial conflicts of interest, but because these economists are more like ideologues or theologians than scientists.

Granted, progressive social values may be more consistent with the scientist’s humanism, but that doesn’t make progressive (Keynesian or socialist) economics more scientific. Although scientists may be working implicitly to benefit humankind, that’s not to say their job is just the providing of propaganda to sell that message. Scientists need to understand how nature works to enable us to exploit opportunities to advance our interests.

True, by observing capitalism’s inefficiencies, progressive economists add to our knowledge of how that kind of society unfolds. But there’s still a culture war between “liberals” and “conservatives,” or between modern humanists and regressive authoritarians or populists, and that war leaves little objective, scientific work for any economist to do.

Economists will insist that they leave the explicit policymaking to government and only spell out how adopting such and such a policy will impact society. Often those predictions don’t hold because the models oversimplify. But even if that weren’t so, economists are still situated right in the middle of a culture or class war, whereas natural scientists are further removed from it in their professional capacities.

There are objective questions to answer in any instrumental project since once a goal is chosen, there are objective differences in efficiency between the possible means of achieving it. But economists all too easily find themselves caught up in some institution’s political or commercial agenda.

Due to these extraordinary social pressures, and to the lack of experimental testing in macroeconomics and the untethered abstractness of their discourse, economists aren’t likely to change their views. They may explore new ways of modelling, as the technology and other social sciences develop. But economics is far more ideological and implicitly normative than its mathematical veneer would lead us to believe.

Is economics a science?

To sum up, we’ve seen that economics lacks some features of science (experimentation and technological applications), and that while it has some others (modelling and mathematical precision), those don’t clinch the issue because pseudosciences could have them too.

The question of the scientific status of economics comes down to its culture. Are economists trained to change their mind based on a commitment to explain the facts? Do their institutions reward them for doing so? Are economists critical thinkers who aim to benefit humanity by understanding how economies work?

I expect that some economists are indeed scientific in that cultural sense, especially the rank-and-file ones, whereas the prominent economists are likely to be more ideological than skeptical. Either way, economics as a set of institutions is vulnerable to enormous political and corporate pressures that could interfere with its humanistic program. Science progresses by pursuing a humanistic agenda, whereas economics might tow the line to rationalize capitalism’s tendency to generate plutocracies, and the latter agenda isn’t obviously in humanity’s interest.

Instead of seeking the objective truth (which in this case is the paradoxical one that there is no purely objective truth in economics, contrary to the economist’s mathematical showmanship, since economies are filled with subjects, not objects), both rank-and-file and leading economists are liable to be caught up in societal power games. The difference is that the leading economists will be more overtly political propagandists, whereas the followers will be cogs in the machine.

The answer, then, as to whether economics is a pseudoscience isn’t likely to be a simple yes or no. There’s a scientific ideal that all sciences only approximate, including physics. Indeed, Lee Smolin, Sabine Hossenfelder, and other physicists have criticized theoretical physics for being more mathematical than empirical over the last several decades, which threatens the scientific status of even that discipline.

There are various branches of economics, and some may be more scientific than others; moreover, economics might be scientific in some ways and not in others. Overall, even if most economists were to strive to be as scientific as possible, they face challenges that other social sciences don’t face. It’s not that the subject of economics is intrinsically more difficult, but that economists are especially tempted to exaggerate their scientific authority to capitalize on their important position as arbiters in a perennial class war (between the rich and the poor).

What we can say with some certainty is that economics is less scientific than economists typically maintain, which means the scientific posturing with its intensive use of math is a case of overcompensation.

Indeed, we should be suspicious of the lack of technological applications of the economist’s purported hyper-rigorous models. Why didn’t the models enable more economists to predict the 2008 collapse of the American real estate market? Didn’t politics and the economists’ financial interests obviously have a hand in blinding them to that looming disaster? And why isn’t there an algorithm for getting rich on the stock market, if economists supposedly have such exact models of how economies work?

The more complex we say these economic phenomena are, the humbler we should be in speaking of how well we understand them, which makes the economist’s mathematical models seem weirdly misplaced.