In the September 18 edition of The New Yorker magazine, writer John Cassidy leads a fascinating tour through parts of the new field of neuroeconomics, the study of the neurological underpinnings of economic decision-making. Sadly, a number of the economists Cassidy interviews are stuck in a conceptual quagmire about the relationship between reason, emotion, and paternalistic public policy. Cassidy, a good journalist, inherits their confusion. There is, in fact, nothing in the research Cassidy reviews that helps to justify, in his words, “a new political philosophy based on the idea of saving people from the vagaries of their limbic regions.”
Cassidy chooses in his article to emphasize the way fragile human deliberative capacities can be overwhelmed by emotion — what he calls a “reason-versus-passion” model. Cassidy refers to Plato’s picture of the soul as “a charioteer attempting to steer the twin horses of passion and spirit,” and Freud’s “contest between the ego and the id.” However, Cassidy’s approach may be most reminiscent of the 18th Century German philosopher Immanuel Kant, who believed that our freedom and dignity depends on acting solely from the motive of rational duty, and never from emotion. Kant’s is the old philosophy of saving people from the vagaries of their limbic regions. But is the limbic system — a network of structures in the brain implicated in the production of emotion and motivation — really something we need to be saved from?
Cassidy, like old Immanuel, appears worried that if human decision-making violates highly abstract standards of rational choice due to emotion, then we will fail to live well, both individually and collectively, and Cassidy suggests that something may need to be done about it — by the government.
Cassidy writes, “Today, most economists agree that, left alone, people will act in their own best interest, and that the market will coordinate their actions to produce outcomes beneficial to all.” However,
“Neuroeconomics potentially challenges both parts of this argument. If emotional responses often trump reason, there can be no presumption that people act in their own best interest. And if markets reflect the decisions that people make when their limbic structures are particularly active, there is little reason to suppose that market outcomes can’t be improved upon.”
It would seem that neuroeconomics, as Cassidy understands it, straightforwardly threatens the liberal presumption in favor of liberty, providing a basis for both paternalistic social policy and regulatory intervention in the economy with one fell brain scan. Cassidy is not putting an exotic spin on the new research, but is simply following the lead of some of the emerging field’s most important thinkers, such as economists George Loewenstein, Colin Camerer, and David Laibson — all quoted heavily in Cassidy’s article, and all proponents of psychology and neuroscience-based paternalistic policy.
The mundane observation that people sometimes let their emotions get the best of them is certainly more exciting when illustrated by colorful pictures of brain activity. But can brain scan evidence about how exactly the brain generates behavior in different scenarios really have such profound implications for politics and policy? Probably not.
Cassidy’s passage above helps illustrate the trouble with one brand of neuroeconomics. The orthodox formal theory of economic rationality has been understood traditionally to provide both a descriptive theory of human behavior and a prescriptive standard for evaluating it. Behavioral economics, including neuroeconomics, shows that the economist’s theory of rational choice is empirically false — a descriptive failure. Where behavioral economics goes wrong, when it goes wrong, is to retain the empirically falsified theory as nevertheless stating a binding ideal of rationality in light of which we can and should evaluate behavior. We don’t, in fact, behave like homo economicus, but we should. As Cassidy shows, it’s a short step to the thought that maybe the government should deploy policy to try to fix our failure to act rationally in the economists’ sense.
The problem is made clearer once you understand that the formal model of choice laid out in microeconomics texts involves some truly wild idealizations of human cognitive capacities. For example, the economists’ standard model stipulates that each person has a well-defined and consistent ordering of preferences over all possible states of affairs (of which there are many), knows exactly what every other person believes and prefers, processes new information instantaneously and without error, and more. The assumptions of the standard theory are not simply psychologically implausible, but probably physically impossible, requiring either periods of deliberation longer than the age of the Universe, or computation faster than the speed of light.
It cannot be surprising, then, that real human subjects with exceedingly limited information and soggy brains that compute slower than a spam-clogged Dell fail to live up to the standards of the formal theory of economic rationality. It would amount to a literal miracle if we did not so fail.
But then why are we concerned at all with a formal theory of rationality, which, after all, applies only to a physically impossible mathematical fiction? Why would we measure ourselves against that standard? Angels, the theologians tell us, are weightless, but those of us bound by gravity are not therefore obese. We can’t be rationally required or obligated to do something impossible. ‘Ought’ implies ‘can’, as the philosophers say. To find that real humans are “irrational” relative to an impossible standard is simply to find that that standard doesn’t apply to us — that the standard of rationality that does apply to us is different. We can only be required to do the best we can do; failing to do the impossible is no real failure at all.
Neuroscience is a category-buster, helping us understand that there is often nothing in the brain that answers to our pre-theoretical taxonomy of the mind. Peering into the brain, it becomes relatively clear that mental categories like ‘belief’, ‘desire’, ‘emotion’, ‘pleasure’, or ‘pain’ do not correspond one-to-one with well-defined, unified, underlying brain mechanisms. Perhaps most significantly — and this requires a lot of intellectual adjustment — there is nothing in brain that looks much like the kind of Reason with a capital ‘R’ that Plato or Kant would recognize and celebrate. The more one appreciates the complex, often unintuitive organization of the brain, the more it becomes apparent that the norms of strict rationality embodied in mathematics, logic, decision theory, game theory, probability theory, and statistics are not internal to the mind, but are remarkably recent, and remarkably precious, cultural achievements.
Capital ‘R’ Rationality is to the mind as ballet is to the body. Through rigorous, disciplined training over years, it is possible to co-opt the evolved nature of the mind to perform amazing feats of Reason, just as it is possible to co-opt the evolved nature of the body to perform breathtaking feats of balletic beauty. But the untrained mind, which depends on lots of quick and dirty cognitive shortcuts that fall short of strict capital ‘R’ standards, was evidently good enough to get us up to the Renaissance and Enlightenment, when the cultural disciplines of Reason helped us ramp up to cultural and economic modernity. And, even then, the growth of science and the extended market order has not required that all citizens become fully vested in the norms of Reason. We can’t all dance Swan Lake, but we also almost never fall down the stairs.
Cassidy’s article flirts with an empirically credible notion of rationality when he discusses the work of neuroscientist Paul Glimcher, who writes, with his co-authors Michael C. Dorris and Hannah M. Bayer, “There is, for example, no evidence that there is an emotional system, per se, and a rational system, per se, for decision making at the neurobiological level.” And that’s right. Glimcher’s pioneering approach assumes that computational resources are scarce, and that the brain must allocate them according to the expected payoff to the organism. In some contexts of choice, the expensive computational processes of the deliberative pre-frontal cortex come online. In others, the brain defaults to more frugal processes involving quick “gut” judgment.
Glimcher’s approach doesn’t attempt to integrate economics and neuroscience by simply comparing (and judging) actual human behavior against the rarefied standards of economic theory, tempting the conclusion that individual behavior and market outcomes can be “improved upon.” Instead, it applies economic theory to the way the brain itself allocates its scarce resources, which helps explains why real behavior — and embodied, ecologically embedded rationality — cannot correspond to a (therefore inapplicable) standard of rationality that assumes an unbounded budget of cognitive resources.
Unfortunately, Cassidy brings up Glimcher’s work only to allow the economist George Loewenstein to airily dismiss it. Cassidy incorrectly writes that Glimcher’s work “might undermine a lot of neuroeconomics,” when in fact Glimcher’s work integrates economic theory and neuroscience at the most promising level. It is neuroeconomics. (Glimcher’s groundbreaking book Decisions, Uncertainty, and the Brain is subtitled The Science of Neuroeconomics.) The point is not to understand how real behavior is anomalous relative to economic theory, but to use economic theory to help us understand real behavior by illuminating the economizing functions of the brain. But missing this point allows Cassidy to preserve his story’s strained “reason-versus-passion” narrative frame, and all the tantalizing policy implications that fall out of it.
If good neuroscience does not in fact recognize a clear distinction between reason and emotion at the physical level, then it cannot be the case that neuroeconomics really implies that, as Cassidy writes, “[i]f emotional responses often trump reason, there can be no presumption that people act in their own best interest.” At best, neuroeconomics shows that peoples’ representation of their best interest shifts from one decision context to the next as the brain shifts its resources from one brain region to another. Neither neuroscience nor economics speaks to which representation of our interest is the right one. That would be a substantive value judgment that goes well beyond science.
However, Cassidy’s account of neuroeconomics makes it seem as though science does reveals that decisions made when in “hot” emotional states — when the limbic system is fired up — threaten to lead us rationally astray, while only decisions made in “cool” deliberative states represent our true “best interests.” But there can be no principled basis for privileging our “cool” desires over our “hot” ones without an independent conception our best interests, and that conception is emphatically not the formal theory of economic rationality. Even in terms of economic theory, there is no principled reason to systematically weigh long-term preferences more heavily than short-term preferences, as economist Glen Whitman has argued in a paper, “Against the New Paternalism: Internalities and the Economics of Self-Control.”
It is no doubt safe to say that we shouldn’t act impulsively too often. But we shouldn’t override impulse too often, either. There are perfectly good grounds for loosening up with a few drinks, or blowing a paycheck in Vegas. Somewhere in the space between androids and orangutans there are human beings. But that isn’t to say that there is one correct way to balance impulse and deliberation. There are many kinds of good lives, and different forms of life strike a different balance between “hot” and “cool” decision-making. As Aristotle notes, we should all eat neither too much nor too little. But too much for you might be just right for Milo, the champion wrestler.
That’s why “asymmetric paternalism” — Camerer and Loewenstein’s name for the “new political philosophy based on the idea of saving people from the vagaries of their limbic regions” mentioned at the start — isn’t really a special, unobjectionable kind of paternalism at all. It’s just paternalism, plain and simple.
Judgments about whether it is worth “saving people from their limbic regions” are no less morally loaded than judgments about whether it is worth saving people from their own sins. If the formal theory of economic rationality is impossible, and therefore cannot provide a binding standard of rationality for real people, it follows that deviations from that standard are not necessarily mistakes. Further, following Glimcher, there is no clear, neuroscientifically-grounded notion of rationality, either. Adherence to the cultural norms of Reason — much like adherence to the norms of ballet — requires disciplined training in the coordination of capacities not “built” to be coordinated in that way. Whether a decision counts as a “mistake” or not depends on contestable value judgments about the norms of reason and the aims of a good life. There may be excellent arguments for such judgments, but those will be arguments about what kind of people we should aspire to be, not arguments about economic rationality or neuroscience.
If you substitute “moral” and “morality” for “rational” and “rationality” in the passage from Cassidy’s essay below, you’ll see the problem more clearly.
“Asymmetric paternalism helps those whose rationality is bounded from making a costly mistake and harms more rational folks very little,” Camerer, Loewenstein, and three colleagues wrote in a 2003 issue of the University of Pennsylvania Law Review. “Such policies should appeal to everyone across the political spectrum.”
Paternalistic policies generally do appeal to everyone across the political spectrum when everyone across the spectrum agrees on the moral questions at issue. But mere agreement cannot overturn the liberal presumption in favor of freedom.
Paternalism is the use of coercion to force people to do or refrain from something against their will for their own good. Liberals of all stripes generally reject paternalism for reasons most lucidly laid out in J.S. Mill’s masterpiece On Liberty. First, we assume the individual is the best judge of her own good. Second, whether or not the individual is the best judge of her own good, we rightly doubt that another individual (or assembly thereof) has the legitimate moral authority to substitute their judgment for the individual’s by force — especially in light of widespread disagreement about the nature of a good life. Third, truth is hard to come by, and none of us can be fully certain we’ve pinned it down. Allowing people to act on diverse opinions about morality (or rationality) broadens the search for truth about good lives by setting up a decentralized system of social laboratories where experiments in living succeed or fail in plain view. So, unless an action harms somebody else, people should be at liberty to satisfy their preferences, whether saintly or sinful, coolly rational or impulsively emotional.
The conceit of the new paternalism is that the state isn’t going to be in the business of telling us which beliefs and desires we are allowed to act on, but will simply nudge people into doing what we wanted to do anyway, but couldn’t manage by ourselves. The idea is that there are things we want to do, but, due to some foible of mind, we are unable to do it without a little outside help. However, the allegedly helpful measures proposed by behavioral economist either aren’t paternalistic at all, or are indistinguishable from bad old fashioned paternalism.
Some of the new so-called “soft” paternalistic measures, such as employers helping workers to increase their rate of savings by requiring them to opt out of, rather than opt into, a retirement plan aren’t paternalistic in any sense; that’s a part of a fully voluntary labor contract. And policies like increasing the taxes on cigarettes or fatty foods in order to discourage potentially harmful consumption choices, are straightforwardly paternalistic in the old sense, requiring a one-size-fits-all value judgment about how much and for what reason we should consume certain goods.
Those kinds of judgments aren’t the proper work of government. In any case, if you really think people make systematic “mistakes” in judgment and choice, there is no reason to believe that democratic voters —who have less at stake when casting their ballots than when choosing what to have for lunch — will be especially good at populating the government with Spock-like rational legislators interested in tweaking cognition through expertly targeted policy rather than with well-coiffed primates interested in hoarding status and power.
What category-busting neuroscience really brings home is the contingency of the norms, disciplines, and institutions of Reason. As Nobel Prize-winning economist Douglass North said in his book Understanding the Process of Economic Change, “much of rational choice is not so much individual cogitation as the embeddedness of the thought process in the larger social and institutional context.” We can’t afford to take that context for granted.
Frictionless, perfectly competitive blackboard markets may fail if blackboard agents aren’t rational by the standards of the blackboard model. But when we find that we are not blackboard rational, we shouldn’t fret about the failure of blackboard markets because there never were any. We should worry instead about the real, rare underpinnings of our amazing prosperity and well-being. In the real world — the territory the blackboard fails to accurately map — cultural artifacts like calculators, clocks, mathematical symbol systems, books, Google, street signs, universities, the scientific method, and the market price system work together to enable forms of reason, and forms of life, that could not otherwise exist.
New findings in brain science will certainly help us to improve the technologies of Reason. But our freedom is perhaps the most important of those technologies. So before we get carried away with exciting brain-based arguments for paternalism and regulation, we should remember that it’s not rational, in any sense of the word, to burn the ladder you’re climbing.