Many hard choices turn out to contain interior decisions that have to be adjudicated separately, and often in some kind of preordained sequence, as in the Abbottabad raid. To make the right choice, you have to figure out how to structure the decision properly, which is itself an important skill. With the pursuit of bin Laden, the CIA had to make a decision about who was in the compound, and then it had to make a decision about how to attack the compound. But each of those decisions was itself made up of two distinct phases, sometimes called divergence and consensus phases. In a divergence phase, the key objective is to get as many perspectives and variables on the table as possible through exploratory exercises designed to reveal new possibilities. Sometimes those possibilities take the form of information that might influence your ultimate choice of which path to take; sometimes those possibilities take the form of entirely new paths that you didn’t contemplate at the beginning of the process. In the consensus phase, the open-ended exploration of new possibilities reverses course, and the group begins to narrow down its options, seeking agreement on the correct path. Each phase requires a distinct set of cognitive tools and collaborative models to succeed. Of course, most of us don’t separate the two phases in our minds at all. We just look at the options, have a few informal meetings, and make a decision, either through some kind of show of hands or an individual assessment.
In the bin Laden pursuit, the CIA deliberately set up a divergence phase at both stages of its investigation into that mysterious compound. A few weeks after Panetta first heard word of the “fortress” on the edge of Abbottabad, his chief of staff ordered the bin Laden team to conjure up twenty-five different ways of identifying the occupants of the compound. They were explicitly told that no idea was too crazy. This was the exploratory phase, after all. The goal was to generate more possibilities, not narrow the field. The analysts turned out to be all too willing to propose unlikely schemes. “One idea was to throw in foul-smelling stink bombs to flush out the occupants of the compound,” Bergen writes. “Another was to play on the presumed religious fanaticism of the compound’s inhabitants and broadcast from loudspeakers outside the compound what purported to be the ‘Voice of Allah,’ saying, ‘You are commanded to come out into the street!’” In the end, they proposed thirty-seven ways of getting surreptitious access to the compound. Most of them turned out to be utterly useless in identifying the occupants, dead ends in the exploratory phase. But some of the schemes ended up opening new paths. One of those paths would eventually lead to the death of Osama bin Laden.
BOUNDED RATIONALITY
What is it about complex decisions that makes them so challenging? For most of the preceding two centuries, our understanding of decision-making largely revolved around the concept of “rational choice” from classical economics. When people confronted a decision point in their lives—whether it involved buying a car or moving to California or voting to leave the European Union—they evaluated the options available to them and considered the relative benefits and costs of each potential outcome (in economics-speak, the “marginal utility” of each option). And then they simply picked the winner: the path that would lead to the most useful destination, the one that satisfied their needs or produced the most happiness with minimal cost.
If you had to specify a point in our intellectual history where that classical foundation first began to crumble, you might well land on the speech Herbert Simon delivered in Stockholm in 1958 when accepting the Nobel Prize in Economic Sciences. Simon’s work had explored all the ways in which the “rational choice” framework concealed the much murkier reality of choices made in the real world. For rational choice to make sense, it required four significant leaps of faith:
The classical model calls for knowledge of all the alternatives that are open to choice. It calls for complete knowledge of, or ability to compute, the consequences that will follow on each of the alternatives. It calls for certainty in the decision-maker’s present and future evaluation of these consequences. It calls for the ability to compare consequences, no matter how diverse and heterogeneous, in terms of some consistent measure of utility.
Think of a decision like the one to bury Collect Pond in these classical terms. Were all the potential options visible to the decision-makers? Were the decision-makers fully aware of the consequences of each potential path? Of course not. You might be able to narrow a decision down to a fixed set of alternatives with reasonably predictable consequences if you were deciding whether to buy frozen pizza or filet mignon for dinner tonight. But in a situation as complex as the one facing the residents of Manhattan circa 1800, the rational choice is not so easily computed. Simon proposed supplementing the elegant (but reductive) formula of rational choice with the notion of what he called “bounded rationality”: decision-makers cannot simply wish away the uncertainty and open-endedness of the choices they confront. They have to develop strategies that specifically address those challenges.
In the sixty years that have passed since Simon’s address, researchers in many fields have expanded our understanding of bounded rationality. We now understand that farsighted decisions are challenging for many different reasons. They involve multiple interacting variables; they demand thinking that covers a full spectrum of different experiences and scales; they force us to predict the future with varying levels of certainty. They often feature conflicting objectives, or potential useful options that are not visible at first glance. And they are vulnerable to the distortions introduced by individual “System 1” thinking, and by the failings of groupthink. There are eight primary factors that contribute to the challenge of farsighted decision-making.
Complex decisions involve multiple variables. When we mull one of those classic lab experiment decisions—“Get $900 for sure OR 90% chance to get $1,000”—there are indeed subtle ways in which our brains steer us to irrational choices, but there are no hidden factors in the choice, no layers that need to be uncovered. Even the unpredictable element—the 90 percent chance—is clearly defined. But in a hard choice—what to do with Collect Pond, how to determine if bin Laden is living in Abbottabad—there can be hundreds of potential variables that might impact the decision and its ultimate consequences. Even intimate decisions can involve a significant number of factors: Darwin’s pros-vs.-cons list calculated the impact of marriage on his social life with “men in clubs,” his desire to have children, his financial stability, his need for romantic companionship, his intellectual ambitions, and more. And in many complex decisions, key variables are not evident at the outset; they have to be uncovered.
Complex decisions require full-spectrum analysis. Imagine the many scales of human experience as slices of the frequency spectrum of audible sound. When we adjust the EQ of a recording, we are zooming in on one of those slices: we want to turn the low end down a bit so the bass doesn’t rumble, or boost the midrange so we can hear the vocals. Music producers have surgically precise tools that allow them to target astonishingly narrow slices of that spectrum, tools that let you extract the background hum of a 120 Hz electric current from a mix, but nothing else. With sound, there are two polar extremes of listening: narrowband and full spectrum. You can carve everything else out of the mix and only hear that hum, or you can listen to the whole orchestra.
Decisions can be imagined in a similar way. The blizzard of decisions that you make over the course of an ordinary day are largely narrowband in nature, like choosing this brand of ketchup over that one or deciding which route to take on your morning commute. But the decisions that really matter in life, the hard choices, can’t be understood on a single scale. It’s not just that they contain multiple variables; it’s also that those variables draw on completely different frames of reference. They are multidisciplinary. Consider the public decisions of voting or rendering a jury verdict. To make those decisions well, you need to force your mind out of its narrowband priorities. You have to think about a problem from multiple perspectives. Voting for a candidate demands th
at you think about the temperament of the politicians in the race, their economic positions and their impact on your own pocketbook, the global forces likely to shape their tenure in elected office, their ability to work with their colleagues in government, and many other variables. A juror has to cognitively shift from the microscopic realm of forensic evidence to the arcane history of legal precedent to the intuitive psychology of reading the facial expressions of witnesses on the stand. Most of us have a powerful urge to retreat to narrowband assessments: She just looks guilty; I’m voting for the guy who will lower my taxes. But we decide better when we break out of the myopia of the single scale.
Complex decisions force us to predict the future. Most decisions, big or small, are fundamentally predictions about the future. I choose vanilla ice cream over chocolate because I can predict, with an accuracy long-buffered by experience, that I will enjoy the vanilla more than the chocolate. The consequences of the US government staging a raid on a private residence in Pakistan were not quite as easy to predict. A modern-day environmental planner might well include microorganisms in weighing the decision to bury Collect Pond, since cleaning up the drinking water involves ridding it of dangerous bacteria. But it seems unlikely that she would have included the microorganisms that caused the fill beneath Five Points to degrade, thus triggering the collapse of housing values in the neighborhood. These are the very definition of chaotic systems: they contain hundreds, if not thousands, of independent variables, all locked into feedback-heavy relationships, where small agents can trigger unimagined tidal waves.
Complex decisions involve varied levels of uncertainty. In many of the classic lab experiments of behavioral economics, psychologists may introduce a level of uncertainty to the decision being studied, but that uncertainty itself is clearly defined by the terms of the experiment: If you choose the 90 percent route versus the sure thing, you know exactly how much uncertainty you are willing to tolerate. But complex decisions in the real world necessarily involve different levels of uncertainty: If you are contemplating a move from New York to California, you can be certain that the winter temperatures will be, on average, warmer if you move, but the question of whether your children will thrive in the state’s public schools is necessarily more ambiguous. Yet in many cases, the outcomes with the highest uncertainty are the ones we care about the most.
Complex decisions often involve conflicting objectives. Narrowband decisions are easy because you don’t have the intermingling of signals from different parts of the spectrum. You don’t have to think about microorganisms altering property values, or how your professional ambition as a scientist might affect your desire for emotional intimacy with a spouse. The chains of causality are simpler. But full spectrum also poses challenges because people often have incompatible value systems at different points on the spectrum. It’s easy to go with your heart when you are only calculating the impact on your emotional state. It’s much harder when your heart conflicts with your politics, or your community roots, or your financial needs—or all three. And, of course, these conflicts become even more severe when the decision involves multiple stakeholders or an entire community.
Complex decisions harbor undiscovered options. As Simon observed, hard choices also confound us because the choices available to us are often not fully defined. They may appear, at first glance, to offer a binary set of options: choose A or choose B. But often the best decision—the decision that somehow finds the most artful balance between the competing bands of the spectrum—turns out to be an option that wasn’t visible at the outset.
Complex decisions are prone to System 1 failings. For the individual contemplating a complex decision, the quirks of System 1 thinking can distort the way the choice is framed or the potential virtues of the choices on the table. Loss aversion, confirmation bias, the availability heuristic—all the shortcuts that make it easy to get through the simple problems of life can turn out to be liabilities when we face a true crossroads.
Complex decisions are vulnerable to failures of collective intelligence. Groups by definition bring a wider set of perspectives and knowledge to the table. Large, diverse groups can be vital to the divergent stage of a decision, introducing new possibilities, exposing unseen risks. But groups are vulnerable to many failings of their own, including collective biases or distortions that arise from the social dynamics of human interaction. The word “groupthink” is a pejorative for a reason. As we will see, many of the techniques that have been developed to augment complex decision-making have been specifically engineered to steer around the potential blind spots or biases of group behavior, and to uncover the wide range of knowledge that a well-curated group possesses.
These eight factors are the shoals on which countless long-term decisions have foundered. It is almost impossible to avoid them all in navigating a difficult choice. But over the decades that have passed since Simon first proposed his notion of bounded rationality, decision-makers in many fields have developed a set of practices that help us steer around some of them, or at least fortify our vessel so that the inevitable collisions do less damage as we make our way to safe harbor.
FINGERPRINTS AND THREADLIKE PRESSURES
In the simplest terms, deliberative decisions involve three steps, designed specifically to overcome the unique challenges of a hard choice: we build an accurate, full-spectrum map of all the variables, and the potential paths available to us; we make predictions about where all those different paths might lead us, given the variables at play; we reach a decision on a path by weighing the various outcomes against our overarching objectives. The first three chapters explore the techniques for making those group decisions, roughly following the sequence that most decision paths unfold along: mapping, predicting, and ultimately making the choice. The final two chapters take a more speculative look at decisions made on the two extremes: mass decisions about broader issues, like the decision we face in battling climate change; and personal decisions, like the one Darwin wrestled with in his notebook.
There’s a wonderful scene in the first half of George Eliot’s Middlemarch that captures the challenges of complex decision-making. (We’ll come back to Middlemarch and an even more famous decision from the book in the final chapter.) The scene follows the internal monologue of an ambitious young physician named Tertius Lydgate in 1830s England as he weighs a particularly vexing group decision: whether to replace the amiable local vicar, Camden Farebrother, with a new chaplain named Tyke, who is supported by Nicholas Bulstrode, the sanctimonious town banker and the main source of funding for Lydgate’s hospital. Lydgate has struck up a friendship with Farebrother, though he disapproves of the vicar’s gambling habit. As a town council meeting approaches, Lydgate churns through his options:
He did not like frustrating his own best purposes by getting on bad terms with Bulstrode; he did not like voting against Farebrother, and helping to deprive him of function and salary; and the question occurred whether the additional forty pounds might not leave the Vicar free from that ignoble care about winning at cards. Moreover, Lydgate did not like the consciousness that in voting for Tyke he should be voting on the side obviously convenient for himself. But would the end really be his own convenience? Other people would say so, and would allege that he was currying favor with Bulstrode for the sake of making himself important and getting on in the world. What then? He for his own part knew that if his personal prospects simply had been concerned, he would not have cared a rotten nut for the banker’s friendship or enmity. What he really cared for was a medium for his work, a vehicle for his ideas; and after all, was he not bound to prefer the object of getting a good hospital, where he could demonstrate the specific distinctions of fever and test therapeutic results, before anything else connected with this chaplaincy? For the first time Lydgate was feeling the hampering threadlike pressure of small social conditions, and their frustrating complexity.
What is striking here is, first, the nuance of the portrait of the decidin
g mind: all those “threadlike pressures” drawn in exacting detail. (Indeed, the excerpt above is only a fraction of Eliot’s treatment of Lydgate’s musings on this one choice, which take up the good part of a chapter.) But the pressures themselves originate with forces wider and more varied than the individual mind. In just this one paragraph, Lydgate is wrestling with his personal friendship with Farebrother; his moral objections to Farebrother’s weakness for cards; the social stigma of being seen as voting on the side of his patron; the economic cost of potentially betraying his patron in a public forum; the threat to his intellectual ambitions if Bulstrode should turn on him; and the opportunities for enhancing the health of the Middlemarch community, thanks to his growing scientific understanding of the “specific distinctions of fever.” The choice itself is binary: Farebrother or Tyke. But the array of factors that shape the choice are scattered across multiple scales, from the intimacy of personal connection to long-term trends in medical science. And the choice is further confounded by the fact that Lydgate himself has conflicting objectives: He wants to see his hospital funded, but he doesn’t want to be mocked by the community for “currying favor” with the banker.
Farsighted Page 3
