1971-1972

1971—1972 was the most productive year of Daniel Kahneman and Amos Tversky’s collaboration: a period spent in Eugene, Oregon, at the Oregon Research Institute, where an intense schedule of experiments and writing helped transform scattered observations about judgment into a general science of intuitive error. The year joined ordinary life, sustained friendship, and ambitious research, and it became the laboratory in which ideas later associated with availability, substitution, Bayesian discipline, forecasting failure, and systematic bias were developed and connected.

Setting and Collaborative Work

Daniel Kahneman and Amos Tversky were guests of the Oregon Research Institute, an institution that housed researchers working across judgment, decision making, and intuitive prediction. Their main host was Paul Slovic, Amos’s former classmate at Ann Arbor and a lifelong friend who was becoming a leading psychologist of risk. Paul and his wife, Roz, introduced the visitors to Eugene, where the researchers jogged, barbecued, and took their children to basketball games while maintaining a demanding professional routine.

The contrast between domestic familiarity and intellectual intensity defines the period. Kahneman and Tversky ran dozens of experiments and wrote articles on judgment heuristics, while Kahneman worked at night on Attention and Effort. The setting was not merely a backdrop: the Oregon year gave the collaboration uninterrupted time, a supportive network, and the practical conditions for turning conversations about intuition into a sustained program of empirical research.

The Availability Heuristic

One central project asked how people estimate the frequency of a category, such as people who divorce after age sixty or dangerous plants. The researchers proposed that people often retrieve instances from memory and judge the category to be large when retrieval feels easy and fluent. They defined the availability heuristic as judging frequency by “the ease with which instances come to mind.” At the time, the two-system approach had not yet been developed, and they did not decide whether availability was a deliberate problem-solving strategy or an automatic operation; later understanding placed both systems in its operation.

The Eugene work also clarified why vivid or easily remembered examples can distort judgment. A person may infer that a class is common not because its members are objectively frequent, but because examples arrive quickly in memory. The same structure appeared in a personal correction: an impression that adultery was more common among politicians than physicians or lawyers seemed to invite explanations involving power and life away from home, until the more basic explanation emerged that politicians’ transgressions were more likely to be reported. Availability could therefore reflect journalists’ choices as much as the underlying frequency of behavior.

From Experiments to a Research Program

The Eugene period belonged to several years in which Kahneman and Tversky studied biases in assigning probabilities, forecasting the future, assessing hypotheses, and estimating frequencies. In the fifth year of their collaboration, they presented their principal findings in Science in an article titled “Judgment Under Uncertainty: Heuristics and Biases.” The article described simplifying shortcuts in intuitive thinking and interpreted roughly twenty biases as both manifestations of those heuristics and demonstrations of their role in judgment.

Their argument challenged the prevailing assumptions of social science in the 1970s. The field generally treated people as rational and thinking as normally sound, while assigning departures from rationality mainly to emotions such as fear, affection, and hatred. The heuristics-and-biases program instead documented systematic errors in ordinary people and traced them to the design of the machinery of cognition. Its importance lay in making error predictable: intuition was not simply occasionally corrupted by passion, but structured in ways that could produce recurring mistakes even in normal judgment.

Consequences for Prediction and Decision

The research agenda developed during and after this period extended beyond availability. Substitution described how, when a satisfactory answer to a difficult target question did not come quickly, System 1 answered an easier related question instead. Bayesian reasoning supplied a contrasting discipline: Bayes’s rule required prior beliefs, or base rates, to be combined with the diagnosticity of evidence. A description of Tom W, for example, raised the probability that he was a computer scientist from a 3 percent base rate to only 11 percent when the description was four times more likely for computer science students than for students in other fields; with an 80 percent base rate, the corresponding probability became 94.1 percent.

Later findings on prediction gave the program a sharp empirical consequence. In a twenty-year study, Philip Tetlock gathered more than 80,000 forecasts from 284 people who earned their living commenting on political and economic trends. The experts performed worse than an equal-probability assignment across three possible outcomes, and even in areas they knew best they were not significantly better than nonspecialists. Greater knowledge produced only a slight improvement and could also strengthen the illusion of skill, while famous forecasters tended to be more flamboyantly overconfident.

The same pattern appeared in corporate finance. A Duke University survey collected 11,600 forecasts from chief financial officers of large corporations estimating the following year’s Standard and Poor’s index returns. The correlation between their estimates and actual outcomes was slightly less than zero. Their 80 percent confidence intervals also failed dramatically: outcomes fell outside the predicted range 67 percent of the time rather than the expected 20 percent. Such overconfidence reflected the tendency to construct a coherent story from information that came to mind while failing to account for information that was absent.

Experimental Legacy and Significance

The work associated with this period ultimately supported a broad taxonomy of judgment under uncertainty. The summary of the research identified representativeness, availability of instances or scenarios, and adjustment from an anchor as three economical and often effective heuristics that nevertheless generate systematic and predictable errors. Bayesian statistics, calibration, and explicit comparison with simple baselines supplied ways to discipline intuition rather than merely criticize it.

The consequences reached far beyond psychology. “Judgment Under Uncertainty: Heuristics and Biases” attracted more attention than its authors expected and became one of the most highly cited works in social science, with more than three hundred scholarly articles referring to it by 2010. Its concepts were applied to medical diagnosis, legal judgment, intelligence analysis, philosophy, finance, statistics, and military strategy. 1971—1972 therefore stands as both a lived interval in Eugene and a decisive turning point: the year in which a collaborative research style, sustained by friendship and relentless experimentation, helped establish heuristics and biases as a framework for understanding how ordinary minds make systematic errors.