Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed

City Journal

Close Nav

Not Quite Rational Man

from the magazine

Not Quite Rational Man

A new paradigm in economics recognizes the complexity in human behavior. Spring 2017
Economy, finance, and budgets

For more than a century, neoclassical theory dominated economic thinking. Neoclassical economics is a theory based on three key assumptions: individuals have rational preferences; individuals maximize utility, while firms maximize profits; and people choose independently, based on available information. As with any widely adopted theory, neoclassical economics has huge merits, but it also suffers from important shortcomings.

One increasingly acknowledged flaw of neoclassical theory is its oversimplified model of human nature, known by academics as “homo-economicus.” Homo-economicus is an efficient calculating machine, someone who always knows what he wants and how to get it (that is, he knows his utility and how to maximize it). But people don’t always know what they want, and if they do, they don’t know why they want it or how to get it. Humans are not cold, rational calculators. They are emotional beings, tricked easily with math; but they are also incredibly creative and fantastic social learners. Is it possible to build an economic theory that takes humans as they are? Or is the complexity of the economy too great for there ever to be a theory that includes the more esoteric aspects of human behavior, such as social learning, emotions, and imagination?

The good news is that, when it comes to building such a theory, economists do not have to work alone. For decades, scholars from a variety of disciplines have been exploring the consequences of these “less rational” aspects of human behavior. As the ideas of these outsiders have begun to penetrate the economic community, they have given rise to what I call “post-neoclassical economics.” This is a body of knowledge that incorporates not only the findings of psychiatrists and behavioral scientists but also those of evolutionary geographers, sociologists, anthropologists, political scientists, historians, development experts, and even some physicists. These nontraditional thinkers have explored the role of social networks and political institutions, as well as innovation, imagination, and collective learning in our study of the economy.

Neoclassical economics succeeded by translating the world into an accepted paradigm, which was delineated by some foundational assumptions. Post-neoclassical economics, by contrast, is a more methodologically agnostic approach, which considers rational agency as just one of many possible models of human behavior. Indeed, what unifies the post-neoclassical approach is the desire to understand economic behavior using any empirically valid methods, no matter in what field they originated. Evolutionary geographers, for example, will borrow methods from, say, network science, if that helps them improve their grasp of regional economic diversification. Behavioral economists don’t hesitate to draw on psychology. These interdisciplinary dialogues create bridges that promote learning and advance our knowledge of how economies operate.

In recent decades, some post-neoclassical work has emerged as a new economic mainstream. The research of psychologists like Daniel Kahneman and of political scientists like Elinor Ostrom has been validated with the highest honor in the economics field: the Nobel Prize. Of course, describing all of post-neoclassical economics is beyond the scope of a single essay. So here I will explore just a few examples of the emerging post-neoclassical paradigm. First, I’ll focus on the study of the economic consequences of emotion, now a relatively mature field. Then I will venture into more uncharted territory, which includes the study of imagination and of collective learning.

Illustrations by Ryan Peltier

The study of emotions and their impact on decision making was pioneered by psychologists, including Kahneman, Amos Tversky, Dan Ariely, Jonathan Haidt, and Daniel Gilbert. Their theories initially met resistance from economists, despite being empirically valid. Part of this resistance is explained in an essay by Milton Friedman, who argues that the assumptions of economic theory don’t have to be empirically valid, as long as they predict the behaviors observed. The analogy he uses is that of an expert billiards player, who performs as if he were skilled at calculating the trajectories of balls using the laws of physics—even if he knows nothing about physics. A physicist could do a reasonable job of explaining the player’s actions in the game. Economic models, Friedman says, can similarly be empirically wrong about the actual motivations of economic actors, but justifiable if they predict their behavior correctly. Some neoclassical theorists use Friedman’s analogy to defend the use of empirically invalid models of human behavior. Yet for that argument to be right, we would have to reject economics as a science. I think that would be too much to lose.

Instead, I suggest that we borrow from the epistemology of physics, psychology, and computer science, and reinterpret the billiards analogy in that light. A physicist modeling the trajectory of a billiard ball would not claim to have a model of the player but rather, one only of the ball. A psychologist or computer scientist, on the other hand, would probably argue that the billiards player performed the calculations implicitly, by an intuitive system that is accurate but nonnarrative—like the neural networks involved in deep learning, if you are a computer scientist, or using what Kahneman calls “system one,” if you are a psychologist.

Most scientists, in my experience, agree that no theory, including economic theory, can be excused from empirical testing of its underlying assumptions (even if Friedman says that it should). Think of the Higgs boson (a.k.a. “the God particle”), which high-energy physicists used for half a century as a theoretical construct to perform calculations. Yet they never accepted the Higgs boson as real until it was confirmed in experiments at CERN, the European Organization for Nuclear Research, in a feat that required a 27-kilometer-long tunnel at the border of France and Switzerland. What makes science science is not the use of mathematical theories but the experiments and observations that validate the theories.

Economics is a science—a beautiful science—and is thus subject to this principle. (See “Economics Does Not Lie,” Summer 2008.) One of the unifying ideas of post-neoclassical economists has been to prioritize empirical findings over theories. If the theory does not match the empirical finding, the theory has to go.

Psychologists have been at the forefront of deepening our understanding of human decision making and how emotions shape our choices. In his 2006 book Stumbling on Happiness, Daniel Gilbert considers both emotions and imagination to explain how our thoughts are distorted when we think about the future and the past. By citing a range of experiments and crafting clever analogies, Gilbert shows that when humans think about the past or the future, they “fill in the blanks” automatically and unconsciously. We suffer from “presentism,” a cognitive bias that limits our ability to imagine ourselves as hungry when we’re full or as happy when we’re sad. Ultimately, the way we see the future or evaluate the past is based on hedonic assessments, where our present feelings are powerful factors that we cannot ignore. Our choices do not represent coldhearted rational calculations; we’re decision-making agents whose choices are inevitably influenced by our present emotions.

“One of the unifying ideas of post-neoclassical economists has been to prioritize empirical findings over theories.”

Jonathan Haidt’s work also explores the role of emotions in decision making, though he focuses sharply on moral choices—decisions in which the answer is good or bad, instead of true or false, or a number. Making moral choices requires performing mental acts that are quite different from, say, calculating the cost of a 5 percent interest rate on a 20-year mortgage. Moral choices are complex computations that scholars have tried to explain for centuries, using one of two hypotheses. The first, the “rationality-first” hypothesis, assumes that humans assess the consequences of their moral choices by anticipating whom an action will harm and how bad the damage will be. This resembles how a neoclassical economic model would operate: here, humans are harm-minimizers who construct behavioral heuristics encoding their rational decision making.

The second hypothesis of moral choice is that humans do not think rationally first but that they make quick emotional decisions that their brains later rationalize, composing a narrative in support of the choice. In this “emotions-first” view, rationality is like a lawyer hired to justify decisions made by our feelings. So when psychologists ask one of their cleverly crafted, albeit sometimes weird, moral questions—is it wrong to have sex with a frozen chicken (if nobody sees you)?—we get a gut feeling justifying our answer first (yes, it’s wrong!) and a stream of words justifying it later. In his book The Righteous Mind, Haidt presents evidence that the emotions-first mechanism is the dominant way by which we make moral choices.

More examples of the importance of emotions in human behavior can be found in the work of Daniel Kahneman and Amos Tversky, the famous duo who begot the field of behavioral economics. Here, I will focus on just one of their contributions: prospect theory, which explains some common, yet extreme, situations, where neoclassical economic theory fails—for example, when people buy lottery tickets or pay large settlements for frivolous lawsuits. Neoclassical economics fails to explain these situations because it assumes that, in uncertain situations, people will pay the expected value of an item. For instance, in a lottery with 1 million tickets, and a prize of $1 million, neoclassical theory predicts that people should buy tickets only when they cost less than one dollar. In a frivolous lawsuit, where people have only a 1 percent chance of losing $1 million, neoclassical theory predicts that settlements should not be larger than $10,000. But the fact that people buy lottery tickets at prices higher than their expected value, and settle frivolous lawsuits by paying more than their expected loss, tells us that we do not weigh our decisions using their expected economic value, at least in these extreme situations.

Prospect theory says that the connection between decision weight and expected values is not linear, as neoclassical economics would have predicted, but S-shaped. That means that people deviate from “rationality” when the cost of a decision is small and the potential benefit is large (lottery tickets), or when the loss is unlikely but could be substantial (the frivolous lawsuit). The frivolous-lawsuit calculation is an example of loss-aversion, the psychological bias that makes people value the things they have at roughly 2.5 times the value of those same things when they don’t have them. Emotional attachment is pricey and real.

Cognitive biases like those embodied in Kahneman and Tversky’s prospect theory, or the presentism described by Gilbert, are so numerous that some people see their sheer number as discrediting behavioral economics. In other words, the existence of such a large number of biases prevents the development of a single coherent theory. Yet for me, this embarrassment of empirical riches is a sign of progress. Consider particle physics. Decades ago, a myriad of particles had been discovered, but physicists at first didn’t know how to assimilate them into a single model. Now, all these particles are seen as a manifestation of a few quarks, leptons, and bosons. Psychology today faces a similar abundance of findings, but the wealth of new evidence is not reason for despair; rather, it provides the fertile empirical ground that post-neoclassical economics needs to model human behavior more accurately. As some have suggested, these biases could be manifestations of shortcuts that evolved to help us make quick decisions in an information-deprived social environment. I’d bet that, over the next few decades, some plausible unifying theories will be proposed in this field.

In neoclassical economics, agents use their imagination to make purchasing and production decisions. In reality, people use their imagination for far more than just commercial strategic choices. In fact, one could argue that the main contribution that imagination makes to the economy is creative instead of strategic: imagination is more important to help us design products than to help us decide what products to exchange.

The creative aspects of imagination, however, are not the bread and butter of neoclassical theory. Creativity and imagination can seem flimsy and hard to define. Nevertheless, three recent books have examined the role of imagination in the economy: Yuval-Noah Harari’s Sapiens (2014), Joseph Henrich’s The Secret of Our Success (2015), and my own Why Information Grows (2015). The authors of these three books are all outsiders to economics: Harari is trained as a historian, Henrich is an anthropologist, and I am a physicist. One could see this as a limitation. But others may value the fact that people trained in wide-ranging disciplines are making an effort to contribute to economics.

In Sapiens, Harari examines the imagination-based origins of human institutions, from religions to corporations. This is an important topic, since institutions have been a difficult nut to crack (though many attempts have been made, including the research of Douglas North and the institutional economics of Ronald Coase and Oliver Williamson). Harari notes that shared beliefs play an important role in society because they facilitate cooperation among strangers. Take religion, one of his chief examples. People who believe in the same God share expectations about moral choices and agree on rituals and behaviors. God is, in an empirical sense, imaginary, but the concept of a deity serves a powerful coordination purpose nevertheless. Similarly, Harari sees institutions as shared imaginings that humans construct collectively, thanks to the inventive capacities of human language—an important feature differentiating human languages from animal communication systems. By creating common worlds through narratives and stories, we can coordinate our activities more effectively. In Harari’s view, institutions were born during the cognitive revolution, some 70,000 years ago, and humans developed imaginative language and could begin sharing worldviews. Imagination is thus a precondition for the emergence of human institutions.

Harari’s ideas resonate with Henrich’s in The Secret of Our Success, which emphasizes that human success is not a simple result of our species’ “superior” intelligence—especially since, in important ways, our intelligence isn’t superior to that of other primates. Our success, rather, hinges on our ability to learn from others and on our ability to accumulate knowledge through generations. Our success isn’t solely the result of individual intelligence but a consequence of collective forms of intelligence, powered by social learning. Humans, Henrich argues, accumulate “cultural packages” of adaptive behaviors. Groups with superior cultural packages, he explains, outcompete other groups, making social learning adaptive. But because cultural packages are hard to explain, their transmission usually involves mysterious or not fully understood rituals that people adopt: taboos, songs, and myths, for instance, which might be literally “inaccurate” but are evolutionarily useful because of the adaptive knowledge that they help convey. In Henrich’s view, the institutions emphasized by Harari are adaptive when they aid in the intergenerational transmission of knowledge.

Henrich teaches us that our ability to imagine solutions to adaptive problems, or to understand why these solutions work, is individually very limited, and therefore has evolved to be tacitly collaborative. As a species, we have not historically relied on our individual ingenuity or rationality but on wisdom, the accumulation of ingenuity developed through generations and transmitted through rituals, some of which seem bizarre—like adding ashes to corn before you eat it, or narratives about why people should share meat after hunting—but have proved decisive for the survival of some groups. Once again, imagination is crucial, since it not only helps provide the narratives that perpetuate the ritual across generations but also because over long periods, imagination is what our species truly accumulates. The growth that preceded the modern pecuniary expansion of economies is that of accumulated wisdom and imagination—what some would call “culture.”

Harari’s and Henrich’s books contribute to our understanding of imagination in the context of human institutions and adaptive culture. Why Information Grows, on the other hand, focuses on the role of imagination in the context of products and economic growth. Economists have habitually considered products as widgets that people exchange to create value, or mathematically, as points in a continuum. But products are far from abstract; they have specific uses (have you tried brushing your teeth with a shoe?). In Why Information Grows, I develop a more granular theory of what products are and how our ability to make them shapes the economy.

Comparing the world of early hominids with our modern world can help us understand the economic relevance of imagination and products. The atoms available to cavemen were the same that we have today, but our world looks extremely different from theirs. What changed? Two things: the way in which those atoms are arranged; and our ability to arrange atoms. Products are not actually made of those atoms but from the physical order that they embody. The same plastic can be used to create a spoon or a comb, just as the same tree can be used to create four chairs or one table. The homes, cars, subways, and refrigerators that we associate with prosperity are made of physical order, begot first as imagination. I refer to that physical order as information and to our capacity to create physical order as computation. Economies are computers that not only calculate prices, as Friedrich Hayek would have said, but that also rearrange atoms to create products.

But why create products? Because, by embodying imagination in matter, we can communicate the practical uses of our knowledge. We live in a world where we constantly use products that we do not know how to make but that make our lives easier. We can communicate at long distances, fly across the world, and enjoy quality entertainment, not because we ourselves know how to manufacture planes, build global communication networks, or make movies but because other people do. And that is true for all of us, since most of the time, we are consuming things made by people who know things that we don’t. By creating products, we multiply the number of people who can benefit from the knowledge and know-how embodied in only a few individuals. Products can communicate uses in ways that words cannot. They represent a different form of communication, essential to understanding economic growth. In this view, economic growth represents our ability to transform useful imagination into reality at scale.

Ultimately, then, a better conceptualization of the role of imagination in the economy involves thinking of imagination in the context of, first, shared beliefs that help us coordinate our activities with others; and, second, the embodied information that allows products to distribute the practical uses of knowledge and know-how.

Can we put these two ideas together? Since creating products is difficult, because making them requires more knowledge than what any single individual possesses, humans need to create networks to accumulate that knowledge and know-how. The creation of these networks is facilitated by the institutions and rituals described by Harari and Henrich but also by the products that we make, since many of these involve devices that augment our communication and transportation capacities. So by embodying imagination into the institutions that help us form cooperative networks, and by embodying imagination into the products that augment our capacity to interact, we expand the capacity of these networks and ignite economic growth. In fact, the diversity and sophistication of a country’s products accurately predict future economic growth—contrary to what neoclassical trade theory would predict, seeing products as epiphenomenal, rather than central to economic development.

Is there a future for this unwieldy, sprawling post-neoclassical field? I believe that there is. Of course, I myself feel part of it, so I might have a vested interest. Nonetheless, I believe that the field is valuable and that several recent developments confirm that it will have a place in our economic thinking.

First, economics is undergoing a generational change. Decades ago, heterodox views of economics—and the scholars advancing them—were excluded from the academic elite and the world’s most prestigious institutions. The most famous example of this marginalization was the ousting of Sam Bowles from Harvard in a highly contested tenure case in the early 1970s. Bowles, Herbert Gintis, and others packed their bags and moved to Amherst, where they started a successful program in heterodox economics that has produced decades of quality research. Bowles and Gintis, important pioneers of behavioral economics, were deeply interested in human behavior and on the conditions under which people cooperate. Also, they were interested in how people acquired preferences through social learning, since they were unhappy assuming utility functions as given.

Nearly 50 years later, things have dramatically changed. Now, behavioral economists are hot in the academic market, and every economics department wants to employ at least one. Most of these new behavioral economists, like Sendhil Mullainathan at Harvard or Dean Karlan at Yale, are relatively young. These Generation X thinkers are serving as models for a new generation of economists, now in graduate schools, who are more willing to challenge the neoclassical tradition. These new generations are looking for niches to make a contribution, and areas once excluded from the economics mainstream provide the most fertile territory for the establishment of a new camp.

This generational shift has also been strong in policy-oriented organizations like the World Bank, the OECD, and even the IMF. Decades ago, these organizations were almost exclusively neoclassical in orientation, but now they are also populated by nontraditional thinkers. The shift in these organizations is important because it means that post-neoclassical economists have leverage within the world’s leading policymaking organizations.

“The diversity and sophistication of a country’s products accurately predict future economic growth.”

The deepening maturation of post-neoclassical thinking has also made the field increasingly relevant. Behavioral economics doesn’t just explore the quirkiness of human behavior; it also makes clear recommendations about how to “nudge” human behavior in (ideally) beneficial ways. The post-neoclassical toolbox goes far beyond this, however. Behavioral psychologists and economists have developed a formal understanding of how the framing of problems affects people’s decisions, even in situations that could be perceived as equivalent, at least from a neoclassical point of view. Too many case studies exist in which simple monetary incentives backfire—for example, the preschool in Israel that started charging parents who picked up their children late, only to see parents arriving even later.

The post-neoclassical approach has also become relevant in the context of innovation systems and regional economic diversification. For decades, as I’ve noted, neoclassical economics has struggled to account for innovation, beyond mathematically abstracting it as an important “secret sauce.” Evolutionary economists, from Richard Nelson and Sidney Winter to Ron Boschma, Marianna Mazzucato, and yours truly, have developed empirically validated theories of the process of economic diversification showing that a region’s productive structure is deeply affected by innovation policy. This literature, which views economic development as a form of collective learning rather than as the consequence of the accumulation of factors, has helped revive interest in some forms of industrial policy and encouraged the development of tools to assess the economic potential of countries and regions.

Finally, post-neoclassicalism also drew strength from the 2008 financial crisis, which encouraged criticism of the neoclassical tradition for not being more self-critical. Consider the abstract of this 2010 paper by Ricardo Caballero from MIT: “The current core of macroeconomics . . . has become so mesmerized with its own internal logic that it has begun to confuse the precision it has achieved about its own world with the precision that it has about the real one. This is dangerous for both methodological and policy reasons.” More recently, Paul Romer, now chief economist of the World Bank, made global ripples with a paper critiquing neoclassical macroeconomics.

Of course, neoclassical economists will not lose their place in history. After all, theirs has been a useful theory. But as economics continues to progress, the neoclassical tradition will need to become more comfortable sharing the spotlight with other theories that succeed where neoclassical theory fails. My bold prediction is that new historical figures will emerge in economics and that they will include people from the post-neoclassical field. These individuals might include those who bloomed at the economics fringe during the last generation—people like Kahneman, Bowles, Mark Granovetter, Tversky, Ostrom, and Gintis—but also those who still have their best work ahead of them.

Illustrations by Ryan Peltier


Send a question or comment using the form below. This message may be routed through support staff.