In this section, I have presented just a few examples of the many components of the multifarious concept of rationality that can be taught. There are learnable macro-strategies for avoiding contaminated mindware. The forming of implementation intentions, mental bundling, and goal formation represent very learnable strategies in the domain of instrumental rationality (achieving one’s goals). These strategies complement nicely the learnable mindware that facilitates the optimal calibration and interpretation of evidence (probabilistic and scientific reasoning skills). Although there are no precise quantitative studies of the issue, it would appear that the propensity for rational thinking is at least as malleable as intelligence.
It’s the Portions, Stupid! Changing the Environment to Help the Cognitive Miser
Perhaps ridding ourselves of our humanity is not in the works; we need tricks, not some grandiose moralizing help.
—Nassim Nicholas Taleb, Fooled by Randomness, 2001
In previous chapters, I argued how our tendencies to process information as cognitive misers are a threat to our autonomy. If the cognitive miser is easily framed, responds to the most vivid stimulus present, and accepts defaults as given, then the behavior of misers will be shaped by whoever in their world has the power to determine these things. Phrased in this manner, the state of affairs seems somewhat ominous. But maybe there is an upside here. Yes, a malicious controller of our environment might choose to exploit us. But perhaps a benevolent controller of our environment could help us—could save us from our irrational acts without our having to change basic aspects of our cognition. The upside is that for certain cognitive problems it might be easier to change the environment than to change people. Because in a democracy we in part control our own environment, as a society we could decide to restructure the world so that it helped people to be more rational.
For example, in a cross-national study of organ donation rates, Eric Johnson and daniel Goldstein found that 85.9 percent of individuals in Sweden had agreed to be organ donors. However, the rate in the United Kingdom was only 17.2 percent.14 What is the difference between the Swedes and the British that accounts for such a large gap in their attitudes about organ donation? Is it that Sweden is a more collectivist society and the United Kingdom a more individualistic one? Are Swedes more altruistic than people from the United Kingdom? Perhaps a clue to the difference might be obtained by looking at the organ donor rate in the United States. It is roughly 28 percent, more similar to that in the United Kingdom than to that in Sweden. Could the difference be one between Anglophone nations and non-Anglophone nations?
You no doubt have guessed the answer to this puzzle by now. The difference in organ donorship among these countries has nothing to do with internal psychological differences between their citizens. The differences among Sweden, the United Kingdom, and the United States have nothing to do with attitudes toward organ donation. The differences are due to a contrast in the public policy about becoming an organ donor in these different countries. In Sweden—like Belgium, France, Poland, and Hungary, where agreement to organ donorship is over 95 percent—the default value on organ donorship is presumed consent. In countries with this public policy, people are assumed to have allowed their organs to be harvested, but can opt out by taking an action (usually by getting a notation on their driver’s licenses). In contrast, in the United States and United Kingdom—like Germany, Denmark, and the Netherlands, where agreement to organ donorship is less than 30 percent—the default value is no donation, with explicit action required to opt for organ donation.
In short, the difference between Sweden and the United Kingdom is not in the people. The citizens of both countries are cognitive misers and probably to a roughly equal extent. The great difference is in the form of a particular public policy. As misers, the citizens of both countries are strongly affected by the default heuristic. The option offered as the default is “sticky” in that it is overly influential. Johnson and Goldstein determined that when people really think about this issue without a default being given to them, roughly 80 percent (much closer to the percentage in Sweden and other opt-out countries) prefer to be organ donors. Since 1995, over 45,000 people have died while on waiting lists for an organ in the United States. A very small change in the donor decision-making environment that hurts no one (since an optout procedure is allowed in all countries with presumed consent) could save the lives of thousands of people. The tendencies of the cognitive miser have cost thousands of people their lives. But these tragic consequences are preventable. The best prevention in this case, though, is a change in the environment rather than a change in people because the former is so much easier to implement.
Examples such as organ donation are what led legal theorist Cass Sunstein and economist Richard Thaler to advocate a policy of what they call libertarian paternalism.15 The paternalistic part of their philosophy is the acknowledgment that government should try to steer the choices of people toward actions that will be good for them. The libertarian part of their philosophy is the guarantee that any policy changes preserve complete freedom of choice. How is it possible to steer people’s choices without interfering with freedom of choice? The answer is: exploit the tendencies of the cognitive miser. Specifically, this often means controlling the aspects of the environment that control the behavior of the cognitive miser—default values and framings.
Consider an example from a domain in which libertarian paternalism has actually been implemented. Financially, Americans are massively under-prepared for their retirement. They have not saved enough. Many people are not participating in the 401(k)s and other retirement savings options that they have available to them. Thaler and colleague Shlomo Benartzi have popularized a series of pension-plan enrollment reforms that could literally rescue the retirement years of millions of workers—years that would have been ruined due to dysrationalic decisions earlier in life. Their reforms are making their way into legislation, and many corporations are beginning to adopt them.16
Thaler and Benartzi’s reforms involve several steps, each involving a way of circumventing a well-known thinking error that is implicated when people make 401(k) decisions. The first step comes at the point when employees of most large companies must first choose to enroll. If they do nothing (do not fill out the relevant form) they are not enrolled. Here is where things first go wrong. Many employees do not enroll. In the Thaler/Benartzi program, employees are automatically signed up for the 401(F) and must choose (by filling out a form) to opt out of the system. Thus, their program exploits the default bias of the cognitive miser.
The second place where employees trip up when making 401(k) decisions is in the allocation of their (and their employer’s) contributions. The Thaler/Benartzi program makes additional use of the default bias by automatically allocating the employee’s contribution equally among a small set of mutual funds to ensure that the initial allocation is diversified. Another of Thaler and Benartzi’s suggested reforms involves getting employees to increase their 401(k) contributions by asking them to commit in advance to having a proportion of their future raises allocated to additional 401(k) contributions. This strategy ensures that the employee will never experience the additional contribution as a loss, because the employee never sees a decrease in the paycheck. Of course, the contribution is the same in either case, but such a procedure encourages the employee to frame it in a way that, according to prospect theory, makes it less aversive.
Thaler and Benartzi have developed a savings program called Save More Tomorrow™ (SMarT), which puts into practice many of the reforms discussed here. It has been used by major corporations such as Hewlett Packard and Philips Electronics. The important point for our discussion here is that it represents an example of inoculation against irrational behavior by changing the environment rather than people. The SMarT program demonstrates that some of the difficulties that arise because of miser tendencies can be dealt with by changes in the environment.
Even in the case of missing mindware, w
e can sometimes make the environment less onerous for those with critical mindware gaps. For example, cognitive psychologist Gerd Gigerenzer has demonstrated that many people have trouble dealing with single-event probabilities (for example, there is a 40 percent probability that the economy will go into a recession). In a survey, Gigerenzer and colleagues found that over 25 percent of survey participants in New York misunderstood the phrase “there will be a 30 percent chance of rain tomorrow.” This minority did not understand the statement to mean that on 30 percent of the days like tomorrow it will rain. They had alternative interpretations such as that it will rain tomorrow 30 percent of the time, or that it will rain over 30 percent of the region tomorrow.
Such misunderstandings of probability terminology are widespread in the domain of medicine. Physician Richard Friedman described the reaction of one patient on being told that the chances were 60 percent that an antidepressant prescribed for her would work. The patient said, “That means that 60 percent of the time I will feel better on this, right?”—displaying one of the classic misunderstandings revealed in the study by Gigerenzer and colleagues. Of course, people should be taught the use of this mindware of probability terminology. But it would be easy to supplement the communication of single-event probabilities with their correct interpretation (“there is a 30 percent chance of rain tomorrow, which means that that if there are 100 days like tomorrow it will rain on 30 of them”). This simple environmental change would prevent people without the relevant mindware from misinterpretation, and it would help them to acquire the mindware. Gigerenzer and other investigators have shown that the processing of probabilistic information, not only by laboratory subjects but also by practicing physicians, is facilitated by clarifying the point that probabilistic information refers to instances of classes.17
All of these examples show how simple environmental changes can prevent rational thinking problems. An even larger category of problems where people need help from their environments is problems of self-control. People over-eat, they over-spend, they procrastinate, they smoke, and they drink too much. Solutions to these problems with self-control are of two forms—corresponding to changes in the individual and changes in the environment. People try to bolster their “willpower”—that is, their internal powers of self-control. Alternatively, they try to rearrange their environments so that less exercise of willpower (autonomous system override) will be necessary. A common strategy here is to use pre-commitment devices. People enroll in automatic savings plans so that they will not over-spend. They prepackage meals so they will not over-eat. They commit themselves to deadlines so they will not procrastinate. Pre-commitments represent our deliberate attempts to restructure our environments so that they will be more conducive to our self-control attempts.
There is some evidence that these pre-commitment devices are successful—producing outcomes that people view as more rational when they are in a reflective state. There is massive evidence that pre-commitment to saving money is efficacious. In other domains there is also suggestive evidence.18 Dan ariely and Klaus Wertenbroch found that students using self-imposed deadlines in academic settings performed better than students not using self-imposed deadlines. Interestingly, however, the self-imposed deadlines were not as good at boosting performance as externally imposed deadlines.
One of the reasons why the domain of weight control remains so intractable is that people have found myriad ways not to pre-commit themselves to one of only two things that will bring their weight down—consuming fewer calories (the other being exercise, of course). The diet industry encourages this tendency by continually implying that there is a way around the prescription to eat less. There are numerous variants: eat only protein and avoid carbohydrates; eat the right carbs, not the wrong carbs; avoid high glycemic-index foods; eat the Top Ten Sonoma Diet Power Foods; eat only sushi; eat all the spaghetti you want as long as you don’t—fill in the blank; the list is endless. All of these prescriptions are ways of avoiding the real point: It’s the portions, stupid! As Consumer Reports (June 2007) advises: “The basic formula for losing weight has not changed: consume fewer calories than you burn” (p. 12).
We are getting no paternalistic help (libertarian or not) in the domain of eating. Our environment is literally making us ill. This is what the work of Paul Rozin and colleagues strongly suggested when they attempted to study the so-called French paradox.19 The mortality rate from heart disease in France is much lower than that in the United States despite the fact that the French have higher blood cholesterol levels and they have more fats (both saturated and unsaturated) in their diets. One reason for the higher heart disease mortality rates in the United States may be that Americans are more obese. Indeed, despite French people eating a higher-fat diet than Americans, the obesity rate in France is only 7.4 percent compared with 22.3 percent in the United States. Rozin and colleagues posited that one reason that Americans were heavier despite eating less fat was because they were routinely exposed to larger portion sizes.
Rozin and colleagues found evidence from a variety of sources indicating that this was in fact the case. They studied portion sizes in chain restaurants that exist in both countries and found that, for example, portion sizes were 28 percent larger in McDonald’s restaurants in the United States than in France. Portion sizes at Pizza Huts in the United States were 42 percent larger. Across eleven comparisons, the United States portion size was 25 percent larger than that in France. Rozin and colleagues examined equivalent recipes from The Joy of Cooking and Je sais cuisiner for seven meat dishes. The mean recipe size was 53 percent larger in The Joy of Cooking. Individual-portion foods were examined, and it was found that, for example, a lasagna dinner was 19 percent larger in the United States. A Nestle Crunch bar in the United States was 41 percent larger, and a yoghurt was 82 percent larger. Over a varied selection of such items, the individual portion in the United States was 37 percent larger. Clearly, in the United States it would be possible to provide much more environmental help than we are getting in the domain of weight control.
Rozin and colleagues have studied the so-called unit bias: that people will tend to eat one portion of something, regardless of the size of that portion, or will tend to eat a unit of something regardless of the size of that unit. In several different studies, the researchers left snacks (M&M’s, Tootsie Rolls, pretzels) in public places. When the size of the snacks was doubled or quadrupled, people did not cut their consumption proportionately. Instead, people consumed much more when the unit sizes were larger. A simple environmental fix—smaller portion sizes—could do a great deal in helping us with the obesity epidemic in the United States.
Society’s Selection Mechanisms
As the brief review in this chapter suggests, many suboptimal outcomes that result from thinking that is less than rational can be prevented. Interestingly, even if intelligence is malleable (as I think it is), the methods needed to increase it will almost certainly involve more long-term training than those involved in teaching most well-known skills of rational thinking.20 It is no wonder that our culture is so full of dysrationalic behavior, when we fail to maximize the use of known mental tools in one domain (rational thinking) while we go in desperate search of ways to facilitate another (intelligence) that, although not unimportant, is no more important.
Given the social consequences of rational versus irrational thinking, the practical relevance of this domain of skills cannot be questioned. Why then, do the selection mechanisms used by society tap only algorithmic-level cognitive capacities and ignore rationality? It makes little sense to test for the narrow concept of intelligence and then confer rewards as if someone had been vetted on the larger, broader concept.
In fact, the issue of the differential privileging of some cognitive skills over others deserves more explicit public discussion. For example, some philosophers have found demonstrations of irrationality in the cognitive science literature implausible because, they say, the subjects—mostly college students—“will g
o on to become leading scientists, jurists, and civil servants” (Stich, 1990, p. 17). I do think that these philosophers have drawn our attention to something startling, but I derive a completely different moral from it. Most jurists and civil servants, in my experience, do seem to have adequate algorithmic-level cognitive capacities. However, despite this, their actions are often decidedly suboptimal. Their performance often fails to measure up, not because they lack working memory capacity or memory retrieval speed, but because their dispositions toward rationality are sometimes low. They may not lack intelligence, but they do lack some rational thinking skills.
The poor performance of the college students in the experiments in the literature on reasoning and decision making is not in the least paradoxical. The college students who fail laboratory tests of decision making and probabilistic reasoning are indeed the future jurists who, despite decent cognitive capacities, will reason badly. These students have never been specifically screened for rationality before entering the laboratory. And they will not be so assessed at any other time. If they are at elite state universities or elite private schools, they will continue up the academic, corporate, political, and economic ladders by passing SATs, GREs, placement tests, and performance simulations that assess primarily the algorithmic mind. Rationality assessment will never take place.
What Intelligence Tests Miss Page 25