Willful

by Richard Robb

  After a distinguished career of fifty years, Edmund Phelps sat down with a blank piece of paper to work out how an economy might best engage workers. The answer, according to Phelps, lies in a particular type of modern capitalism that he identifies in his book Mass Flourishing. That economy is “dynamic.” Entrepreneurs and workers test out new ideas, and innovation takes hold at the grassroots. People flourish not only in terms of consumption, but also because “a good economy promotes lives of vitality.”22

  Phelps shows that this dynamism comes not from breakthroughs in basic science, the accumulation of physical capital, human capital (education and training), or the emergence of business geniuses like Bill Gates. If science were enough, dynamic economies would have sprung up at various points in the ancient world. The Greeks, for instance, achieved little by way of business or commercial innovation in spite of their astonishing scientific achievements. Not only did they know that the earth is round, around 240 BCE Eratosthenes calculated its circumference within a few thousand miles. They understood the causes of lunar and solar eclipses and accurately predicted when they would occur. Anaximander, born two hundred years before Aristotle, speculated that life first formed in water. Ancient Romans surmised that disease is caused by invisible little animals—an idea that could have made the Middle Ages considerably more pleasant, if only they had run with it.23 But unfortunately, most of this science never made it to the shop room floor.24

  Having examined when commercial activity thrives and when it does not, Phelps concludes in Mass Flourishing, daringly for an economist, that values play a central role. Values determine the extent to which people, whether swashbuckling entrepreneurs or staffers in established firms, pour their energies into business. Certain values give rise to an entrepreneurial spirit that seizes particular ages, including Britain starting in 1815; the United States, Belgium, and France starting in the 1830s; and Germany and Prussia in the second half of the nineteenth century. This spirit is necessary for mass flourishing.

  While mass flourishing is a social phenomenon fostered by common values, each person’s engagement within a dynamic economy is necessarily for-itself. The values that prompt people to take up challenges are more elusive than traditional inputs into society’s production function, yet as Phelps argues, they can still be studied and gradually understood. Employees and entrepreneurs caught up in the “sport” of economic life supply agile thinking and take risks, feeding into the innovation crucial to the good economy. That economy in turn makes rewarding jobs broadly accessible—not only to those with the skill set to become robotics engineers in Silicon Valley, but also to those whose work depends on muscle, sociability, organizational skills, and practical cleverness.

  Not long ago, the line between sports and work could become quite blurred. In American Work-Sports, Frank Zarnowski documents the number-one team sport in America in the 1850s, after cricket faded and before baseball took hold: the firemen’s muster. Firemen would compete in tournaments with standardized rules, mostly as to how far and how high they could pump water. Firemen’s sports were covered widely in the press, star firemen were famous, and the musters attracted crowds averaging nearly three thousand. Many other “work-sports” were enormously popular, from laying railroad tracks, lumberjacking, and rock-drilling to office sports such as typesetting and dictation. In 1938, roughly a million spectators attended cornhusking competitions across America, not counting listeners of the live radio broadcasts. Typewriter manufacturers sponsored popular speed typists and large crowds would gather to watch the champions face off.

  The analogy to sport helps explain certain corporate strategies. We are so accustomed to executives promoting plans for growth that it would be eccentric to ask: “Why do you need to grow?” It’s a rare executive who announces that the firm’s business opportunities are dwindling, its competitors are too formidable, and the cost of fighting them is too high, so the plan is to shrink at the right pace. Perhaps growth is a default imperative for businesses because it allows workers to tackle real difficulties. A stagnant firm, no matter how profitable, deprives employees of a key reason to come to work. It’s no fun. Even mergers and acquisitions may be gambits to stave off boredom. Given that they rarely raise the share price of the acquiring company, generating fresh challenges may be part of the appeal.25

  Under the right conditions, workplace victories can be sweet even if they lose money. In the mid-1990s, my boss’s boss at the Japanese bank DKB, Mr. Hosaka, visited from the bank’s head office in Tokyo. He wanted to win a strategic transaction for an important client. I asked if it was okay to take a loss on the deal. He answered, “We have other ways to make money.” I took these instructions as a beautiful frontal attack on the simplistic model of work as an exchange of time and effort for financial compensation. Skeptics may claim that Mr. Hosaka planned to make even more money from the client by giving away a “loss leader.” But I know that’s not what he meant. He wanted this victory.

  SUMMING UP

  Purposeful versus For-Itself

  A Peace Treaty

  We are now ready to expand the earlier diagram and take one final tour of the main categories of action. Along the way, we’ll address the question: What are the benefits of recognizing the boundary between the purposeful and for-itself realms?

  In the updated diagram, alternatives to purposeful choice appear in parentheses under each category of for-itself action, while the list on the right describes rationalizations that we use to conceal the importance of for-itself activity from ourselves. When we hew to preestablished beliefs, we concoct reasons to mistrust arguments to the contrary; when we frame an unpredictable altruistic gesture as purposeful, we might explain that “what goes around comes around” or that it will win points with a divine scorekeeper; and, finally, for action over time, we retroactively supply a “reasonable” motive after behaving impulsively.

  Rational Choice

  Which makes a better grilled cheese sandwich—Gruyère or cheddar? Gruyère, as far as I’m concerned. Which is better—a Ferris wheel or a roller coaster? Probably a Ferris wheel, but it would depend on the specifics. Which is better—a grilled cheese sandwich or a ride on a Ferris wheel? Well, that’s tougher. Out of context, the only sensible answer is, “They’re different, you can’t compare them.” But in our lives we do have to choose. I might settle for suboptimal cheese or skip the sandwich altogether to take my children to an amusement park (sadly, they’re too old for that, but you get the idea). If I live a remotely rational life, if I’m optimizing to satisfy my preferences in any meaningful way, I must choose among options that are not commensurable in the narrow sense of two types of cheese or two types of rides. The more difficult which-is-better questions must be answered in order to make the best use of limited resources.

  Do I renovate the kitchen or go on a vacation? Do I take the scenic route or hurry home to watch TV? Comparisons, whether complicated or simple, lie at the heart of the rational choice model. Before we can act, we must assign values to each option based on our preferences and how much we already have. Optimization requires that the utility we derive from the last dollar spent on each good or service is the same as the utility of the last dollar spent on every other good or service. If it weren’t, we could do better by redirecting our dollars to the good or service that generates more satisfaction on the margin.

  Money, though closely associated with the rational choice model, is not essential to this calculation. The question “how much would you pay for one more apple?” could instead be posed as “what fraction of an hour would you be willing to work for one more apple?” Same question, but in the second formulation we divide every price by an hourly wage. Without changing the analysis, we could express every trade-off in terms of hours worked or consumable goods. We speak in terms of money as a matter of convenience, but the same logic would apply to a barter economy.

  I suspect that those with a visceral dislike of economics or the quantitative social sciences are reacting
to this abstraction and to the premise of commensurability more broadly. Likewise, antipathy toward an excessive concern with money may at heart be uneasiness over the calculations that the use of money implies. Perhaps those who snipe at economics—arguing that it’s defective because it failed to predict the most recent economic crisis—are really rebelling against the whole enterprise. They may object to the way rational choice squeezes all our options into a common metric or the way a singular focus on rational choice obscures our ability to see the for-itself side of life.

  Whether cast in monetary or other terms, rational choice relates closely to the idea that people can quantify the strength of each desire on a common scale. If you think that sounds mechanistic, then consider Jeremy Bentham’s burlesque version of this idea. In the late eighteenth century, Bentham devised an algorithm for estimating the pleasure or pain that would result from any decision. His “felicific calculus” combines fourteen basic pleasures, including expectation, wealth, skill, and amity, as well as twelve basic pains, and adjusts for factors like the intensity and duration of each variable. Any of the twenty-six basic inputs can be traded for another to maximize well-being subject to the prevailing constraints.1

  In Bentham’s utilitarianism, each person is born in the straitjacket of the felicific calculus and can never escape.2 Although tastes may vary, there is a single correct course of action in every situation—one need only do the math. (To be fair, Bentham was mainly interested in social reform. He didn’t encourage individuals to work through the felicific calculus when making life decisions. Nevertheless, Bentham’s utilitarianism leaves his followers with a sterile conception of human agency.)

  Rational choice does at least allow for unlimited variation among individual preferences and some scope to shape those preferences over a lifetime. But the difference between rational choice when it’s applied to decision-making and Bentham’s utilitarianism is one of degree rather than kind. In both cases, choice becomes a passive calculation.

  Rational choice, then, reduces what people want to one word: more. This “more” is one-dimensional because everything is commensurable, everything can be compared and evaluated. In this view, a failure to attain “more” results strictly from bad luck or bad decisions.

  Behavioral Bias

  Purposeful choice’s explanation for non-maximizing action is behavioral economics. This field investigates the systematic mistakes that we could correct by recognizing our biases and mental shortcuts.

  It shouldn’t be hard to persuade anyone who falls for the gambler’s fallacy that the ball is no more likely to land on red in the next spin of the roulette wheel because it was red the last five times or the opposite, that black is not due to catch up.

  A person exhibiting loss aversion might reject a bet with a 50 percent chance of winning two dollars and a 50 percent chance of losing one dollar. She pays over the odds to avoid small losses because losing is accompanied by psychological discomfort. (This is different from risk aversion, which deals with losses on a large scale.) This bias had better be contained if you’re in the business of taking financial risks.

  After an event transpires, you may feel that you saw it coming, that you knew it all along, even though you didn’t. Behavioral economists call this the hindsight effect. This bias is pernicious, since perceiving outcomes as inevitable makes it harder to learn from experience.

  It’s not always easy to distinguish among behavioral bias, rational choice, and for-itself behavior.

  Biases That Are Mostly Rational Choice

  Sometimes a phenomenon classified as a behavioral bias is really rational choice in disguise. For instance, given that the chances of winning are incredibly remote, the sale of lottery tickets gets blamed on the optimism bias or simply a bias toward overestimating small probabilities (twin to the bias toward treating small probabilities as if they were zero). But we don’t have to resort to cognitive biases to explain the lottery. Perhaps people buy lottery tickets to animate fantasies of wealth that will wipe their cares away. As Aristotle reminds us, “It is pleasant for [a person] to think he will get what he wants; but no one wants things that seem impossible for himself to attain.”3 Buying the ticket converts the dream of getting rich quick from impossible to possible, and some people must find that appealing.

  Biases That Are Mostly For-Itself

  The difficulty of recognizing for-itself actions makes them particularly prone to misclassification as behavioral biases. As we saw in Part 2, confidence in one’s own opinions is central to holding on to an identity. We tend to believe in ourselves. While this may look to an observer like a failure to optimize, it stands apart from purposeful concerns. If we miss this distinction, however, we might mistake a behavior that is both natural and necessary for overconfidence from the optimistic bias, which has been called “the most significant of the cognitive biases.”4 Or we might diagnose it as confirmation bias, a tendency to notice and remember information that confirms what we already believe.

  The omission bias refers to a tendency to favor inaction in decisions with moral consequences. Many people who would be unlikely to push the fat man in the trolley problem would be even less likely to catch him if he slipped on a banana peel and was about to fall onto the tracks and save five people. Either way, as we saw in Chapter 7, this choice is for-itself. It only looks like a bias if we assume that any concern for others has to be care altruism inside a utility function.

  Similarly, if we insist on seeing all behavior as purposeful, we might attribute the difficulty of staying with a plan over time to hyperbolic discounting. Today, we prefer one consumption stream; tomorrow another. We’d even pay to get out of the plan we’d selected. But as discussed in Chapter 8 and the Online Technical Appendix (www.willful-appendix.com), this is not a behavioral bias; rather, choosing through time belongs to the for-itself realm.

  Positive Psychology

  The final branch of purposeful behavior in our diagram is positive psychology, or happiness research. This field assumes that studying what appears to satisfy others can teach us about our own natures. Implicit to this project is the assumption that self-reported happiness corresponds to something real. But when asked, “Are you happy?” it’s natural to wonder, “Compared to what?” I can report my well-being relative to either how I used to feel or how I imagine people around me feel. Thus the major empirical findings of happiness research are unsurprising: reported happiness is more a function of the rate of income growth (habituation) and relative income (rivalry) than the income level itself.5

  This field has recently emphasized the importance of “flow” for enhancing self-reported well-being. As Martin Seligman says, it’s about “being one with the music.”6 To get more flow, Seligman, who plays bridge in the upper echelon, recommends picking challenges that correspond to your greatest talents (if he’s to be believed, then I should give up trying to learn Japanese).

  Seligman’s “flow” has elements in common with the for-itself realm: overcoming obstacles and emphasizing gameplay over utility maximization. Seligman goes further, though, by creating a quantitative measure of well-being. And while he doesn’t take that measure too literally or stray too far into Bentham territory, he does incorporate flow, or “engagement,” as an input into well-being. This creates a contradiction: when you’re at one with the music, you can’t also be evaluating, comparing, or optimizing. As we’ve seen, it’s a circle that Seligman will never fully square. Flow/engagement can’t be quantified and maximized as part of a larger formula.

  For-Itself

  The for-itself realm encompasses surprising or spontaneous actions, such as a wild leap out of character, a struggle to overcome a challenge that has been freely chosen, and a fierce adherence to an opinion after the time is long past to give it up. Exercising one’s will on the world is the key to this realm.

  The urge to place all behavior in the purposeful realm can be hard to resist. If people like adventure, sensation, challenges, and overcoming obstacles,
why not just plug them into the traditional rational choice framework? Couldn’t a rational agent assign a low utility to overcoming challenges if the probability of success were too close to either 0 percent or 100 percent? If people like to act in character, treat themselves to an occasional altruistic gesture and so on, why not put a price on it? Agents would then be able to solve the resulting maximization problem, and rational choice seems like it should do just fine.7

  But, as we have seen, this can’t be done. Some options are wholly incommensurable and inhabit the for-itself realm. Three major categories of action—acting within our identity, engaging in certain altruistic gestures (mercy and love altruism), and organizing ourselves through time—are unlike the choice between cheese versus cheese or even cheese versus Ferris wheel; they just don’t fit into an optimization problem.

  I have developed various arguments in this book to show that purposeful choice cannot explain all action. Just assessing potential new beliefs triggers “the irritation of doubt,” so we stick with what we think we know. Adoption of identity-stretching beliefs leads to new preferences and, in the absence of a homunculus to direct the formation of potential identities, there’s no clear way to optimize. We also saw that certain altruistic gestures—but not all—are spontaneous and unpredictable. Others are bigger than the purposeful choice apparatus. Finally, we showed mathematically that it’s impossible to have preferences about different arrangements of consumption through time unless we make strange, implausible assumptions. At heart, all these arguments share a common intuition: acts of will can’t arise out of passive calculations to satisfy preferences or they would no longer be acts of will.