According to Lynn Stout (LS), the "homo economicus" model -- the rational, calculating selfishness that is presented as a universal trait of human behavior in modern economics -- is anything but universal. While we may exhibit this sort of behavior some of the time or in certain contexts, e.g. when trading on the stock market, most of us are actually unselfish a great deal of the time. We don't drop garbage onto freshly-cleaned floors, we line up patiently for ice cream on a hot day, we usually obey the law and expect others to, as well.
LS describes experiments with certain types of games, which show that this unselfishness is common across cultures. They also show that depending on the rules of the game, "unselfish prosocial behavior" can be encouraged or undermined. Financial incentives -- the favorite remedy of economists, and judges and lawyers who follow the "law and economics" school of thought -- can actually undermine unselfishness. She cites the famous example of the (Israeli) day-care center that decided to fine parents who were late to pick up their children: lateness was even more frequent than in the pre-fine regime, because parents' guilt was transformed into the feeling that they were paying a bit more for the convenience of showing up later. Certain aspects of American law take this unselfishness into account. For example, those who injure the person or property of others through negligence (an example of what's called a "tort," in legalese) aren't expected to pay the full amount of the victim's damage, while those who seem to have been deliberately callous might have to pay far more than that amount, in punitive damages.
The book makes many good points. The "H. economicus" model is certainly flawed. Usual law and economics scholarship does often come up with flawed or even perverse recommendations. Legislators and regulators ought to structure laws so that they don't undermine our tendency to be unselfish. And what I found to be maybe the simplest and most important point: we just don't notice enough about how nice people are most of the time. Bad behavior grabs our attention more, misleading us into believing it's more prevalent than it is. These points are presented in a very clear style that reads as smoothly as the work of a professional journalist, which is quite an achievement for a law professor.
While I agreed with these big points, and also with some of LS's more specific recommendations (such as that companies not be treated the same as individuals under the law, since they more often exhibit selfish behavior than humans do), I was less persuaded, and sometimes confused or even troubled, by some of LS's supporting argumentation. Moreover, with its combination of utilitarianism, scientism and elitism, the book comes across more as a sort of heterodox Law and Economics, rather than a "blistering attack" on it, as a blurb would have it. The rest of this long review focuses on these reservations.
1. LS's use of the word "conscience" seems to be inconsistent throughout the book. After emphasizing that the distinction between acts and feelings is crucial (@12f), LS defines conscience as behavior:
"Emotions like pride, guilt, and irrational worry feel subjectively selfish. But the _behavior_ they inspire is not. However egoistic her motive, the person who sacrifices time or wealth to help or avoid hurting someone has acted, objectively, unselfishly. ... As a result, this book will define an act as unselfishly prosocial when it requires the actor to sacrifice time, money, or some other valuable resource to help, or to avoid harming, others. ... A principal theme of this book is that unselfish prosocial behavior -- conscience -- is a very real, very common, very powerful and very important phenomenon." (@13-15; emphasis in original.)
Often in the rest of the book, though, the text vacillates between speaking of conscience as behavior and conscience as feeling. E.g., LS speaks of "the idea that most people have a conscience" (@16), and faults scholars who don't consider "how conscience might influence behavior" (@36). And later:
"To the regulator or policymaker, it doesn't matter if 'selfish' feelings prompt people to keep promises, follow rules and help others. What matters is that they _do_ keep promises, follow rules and help others -- even when they have little or no external incentive to do so. We don't need to fully understand the workings of conscience to study, and value, how it affects behavior." (@55; emphasis in original.)
And again: "'[C]onscience' may be a form of cognitive error, a prosocial glitch in our thinking that prompts us to behave as if we expect to meet people again and as if we believe we are being observed ...." (@135f.) If a person's conscience is *behavior,* it cannot be a cognitive error or "prompt" her to behave in a certain way.
I'm surprised that none of the Princeton U. Press editors or peer reviewers caught this ambiguity about the most important word in the book.
2. Declaring psychological utility to be a "crafty and common" strategy of economists (@33), LS points out that "[a]ny and all types of philanthropic, destructive, or downright bizarre behaviors ... becomes by definition self-interested. This circularity ... erodes the value of economics as a predictive tool. ... For economics to preserve its predictive power, we must assume people get utility primarily from improving their own material circumstances."
By excluding psychological pleasure or pain as motivators in economics, LS rejects more than 200 years of economic theory (dating back at least to the French utilitarian economist J.-B. Say). The early neoclassical economists, such as Léon Walras and Francis Edgewood, whom most would say *gave* economics its predictive power, were firmly in this psychological camp. They would agree that such behaviors can be self-interested; indeed, the Chicago School accounts of altruism, crime, etc. are based exactly on this sort of psychological pay-off.
But does LS's revisionism actually improve the predictive ability of economics? According to her theory, there are times when people do act in a "self-interested" way, e.g. when shopping for cheap gasoline, and times when people act with a "conscience." Her theory still needs to account for how we can discriminate between the two regimes of behavior *predictively.* Moreover, by defining "conscience" as behavior, she doesn't offer any *explanation* of why people behave that way in the specific circumstance. Conscience becomes merely a descriptive term.
Conventional utility theory is much more elegant, because it explains ALL our actions through the notion of utility. Sure, it too needs some criterion for predicting when we will be motivated by material benefits and when by psychological ones -- but this is no weaker than LS's own theory. By having a unified explanation, instead of introducing an operational notion of conscience that is unexplained, conventional utility theory avoids dragging unnecessary entities into the theory, as Occam's Razor recommends. (NB: I'm not a fan of conventional utility theory. I'm simply pointing out that it isn't weaker than LS's theory when it comes to prediction, and in fact is superior in some formal ways.)
I think what LS might be getting at is falsifiability, not predictability. Psychological utility theory isn't *falsifiable,* because it can justify any action ex post by the notion of "well, you did it, so that means you really preferred to do it." If that's really what she meant, I'd agree that this is indeed a "crafty" weakness of the theory, and one that can lead to evil applications of economic argument.
3. Evolutionary theory plays a strange role in this book. Several times LS speaks of observations of unselfish behavior "finding support" in evolutionary theory (e.g., @ 101, 146). Shouldn't it be the other way around -- that observations support a theory (or don't)?
Chapter 6 of the book is a long, speculative proposal for how conscience evolved. It has falsifiability issues of its own, being in the vein of that somewhat plausible speculation you find in Polybius, Hobbes and Rousseau about how different forms of human organization were created. Another flaw of the explanation -- and an ironic echo of the fallacy of psychological utility theory -- is that it assumes that because a trait exists today, it's therefore adaptive or useful in some way. LS deals neither with the notion of evolutionary neutrality, nor with Stephen Jay Gould's idea of "spandrels," characteristics that didn't originate through natural selection, though they may prove to be adaptive later.
Overall, the true function of evolutionary explanation in the book seems to be a form of scientism: borrowing the prestige of science for rhetorical purposes, even when it isn't justified in terms of the science being invoked. LS provides more in this vein by frequent references to fMRI brain scans and the like (10, 112, 138, 241). Modern economics is guilty of scientism in spades, with mathematics and physics; but I'm not sure that retaliation with biology is an improvement. And LS's gambit creates some other explanatory problems, as I'll mention below.
4. One explanatory problem is that LS ignores culture as a differentiator. Traits that evolve biologically and claimed as features of "human nature" should apply to all groups of people, and at all times on sufficiently short time scales (e.g., less than tens of thousands of years, for human evolution). So, in Chapter 7, LS points out that understanding conscience helps us to understand the US system of tort law (e.g., @174). What about tort law outside the US? What about US tort law at other times during the past 300 years? Should we credit human nature for the presence of certain features in current US law, if those features aren't shared by all legal systems that are near enough to ours in history?
5. Some important cultural features missing from the book are religion and deontological ethics. LS declares that we are all "intuitive utilitarians" (@115, 225). But does everyone always "weigh benefits to others against costs to ourselves in choosing to act prosocially" (@225)? Many people act because they *believe it is the right thing to do,* period. Far from being a "cognitive error " based on mistaken beliefs, some people might be nice to strangers because of a belief in karma, a belief that there is something divine in every human, or a reading of Kant, among other possible explanations.
6. On a more specific note: the book's analysis of contract law (Chapter 8) is far removed from actual business practice, being overly concerned with the roles of judges and lawsuits and the theories of professors. Based on my roughly 30 years' experience as a transactional lawyer, it sure doesn't seem to me that parties to contracts treat each other decently because of what judges say; most parties to contracts *don't have any idea* of what courts say, nor could they care less. The chapter also doesn't discuss the realities of unequal bargaining power, which affects people's choice of partner and constrains their subsequent behavior, and arbitration, which is on its way to making judges obsolete in contract disputes in the US.
This lack of realism also impacts the book's broader arguments. LS faults "behavioral economists" for relaxing the assumption of rationality in economics while retaining the assumption of selfishness. But she herself retains the fiction that economics consists in transactions between strangers (see esp. her copious references to experimental gaming). In fact, much of economic life occurs between parties who know each other. We are still waiting for a theory of economics that takes this adequately into account.
7. Finally, a point that troubled me most: LS encourages readers to understand conscience as *instrumental*, a tool of control. She presents a "model" of conscience based on three "variables": instructions from authority, beliefs about other's prosocial behavior, and the magnitude of benefits to others (@99). Immediately afterwards, in a section entitled "WHY THESE THREE SOCIAL VARIABLES?," LS notes "[T]hese three variables are especially easy for lawmakers and policymakers to manipulate; that is, they offer especially good levers for shifting human behavior on a large scale." (@100.)
This emphasis on regulators and policymakers (who in the US are typically unelected officials in the Executive branch), as well the book's frequent references to "lay" people and terminology, reminded me uneasily of the elite "Planners" in Sunstein and Thaler's "Nudge," an example of the behavioral economic analysis LS lambastes. There isn't any role for politics in this book, just as there wasn't in S&T's. Moreover laws are to be obeyed -- no role for civil disobedience or conscientious objection, either. Clearly, LS had many good and humane intentions when writing this book. But it is uncomfortable to see control as the central theme, and justice and democracy absent.
In her closing pages, LS contrasts the beliefs of two early 20th Century US Supreme Court Justices, Oliver Wendell Holmes and Louis Brandeis. Holmes believed that the law should assume that all men are bad, whereas Brandeis believed that "no small part of the law's function is to make men good." "The scientific evidence supports Brandeis over Holmes," she declares. I don't read her book as supporting this conclusion at all. For more than 250 pages LS emphasizes that the scientific evidence is that men are *already* good, at least most of the time. What law can do, she points out, is to create incentives for men to be selfish and bad -- so policymakers and regulators should be careful not to undermine their goodness. I think that's a very important message. I only wish that the book's own arguments put it into clearer relief.