In Praise of Computation

Environmental and Resource Economics https://doi.org/10.1007/s10640-025-00958-2 In Praise of Computation Cass R. Sunstein1 · Lucia A. Reisch2 Accepted: 6 January 2025 © The Author(s) 2025 Abstract A great deal of work in behavioral science emphasizes that statistical predictions often outperform clinical predictions. Formulas tend to do better than people do, and algorithms tend to outperform human beings, including experts. One reason is that algorithms do not show inconsistency or “noise”; another reason is that they are often free from cognitive biases. These points have broad implications for risk assessment in domains that include health, safety, and the environment. Still, there is evidence that many people distrust algorithms and would prefer a human decisionmaker. We offer a set of preliminary findings about how a tested population chooses between a human being and an algorithm. In a simple choice between the two across diverse settings, people are about equally divided in their preference. We also find that that a significant number of people are willing to shift in favor of algorithms when they learn something about them, but also that a significant number of people are unmoved by the relevant information. These findings have implications for current findings about “algorithm aversion” and “algorithm appreciation.” Keywords Law of small numbers · Biases · Algorithms · Algorithm aversion · Heuristics · Artificial intelligence I’ll talk about the most important event in my life, which is my collaboration with a close colleague named Amos Tversky.… We were the ones, I think, who introduced the notion of cognitive biases and cognitive illusions. And that work had some influence. But we didn’t do it to have influence. We just did it because it was so much fun that we could hardly stop ourselves.–Daniel Kahneman.Do people like algorithms and artificial intelligence? Will Lucia A. Reisch Cass R. Sunstein 1 Robert Walmsley University, Harvard University, Harvard Law School, Cambridge, MA, USA 2 El-Erian Professor for Behavioural Economics and Policy El-Erian Institute for Behavioural Economics and Policy Cambridge Judge Business School University of Cambridge, Trumpington Rd, Cambridge, UK 13 C. R. Sunstein, L. A. Reisch they trust them? Will they use them? Do they prefer human beings? When? These questions are increasingly central to many domains, including governance in general, corporate behavior, risk regulation, environmental protection, climate change, autonomous vehicles, antitrust, and competition policy. We aim here to offer some preliminary answers to these questions, with an emphasis less on the particular findings than on one empirical strategy to deal with them. But we start in an unlikely place, with two duos who changed the world. The Beatles’ first hit, back in 1963, was “Love Me Do,” and it contains much of what made the Beatles great. Sure, the group became much more sophisticated over time, and also much more complicated. “Tomorrow Never Knows” is on another level, and the same is true for “I Feel Fine,” “Rain,” “Paperback Writer,” “Get Back,” and of course “Hey Jude.” Still, “Love Me Do” has the core of much of what followed: the harmonies, the effervescence, the humor, the wink and the nod. John Lennon and Paul McCartney went in all sorts of directions, separately and together, and they were geniuses, separately and together. Still, back at the beginning, something special was there, and it was, well, Lennon and McCartney. Amos Tversky and Daniel Kahneman had their own “Love Me Do.” It is called “Belief in the Law of Small Numbers,” and it contains a lot of what made the duo great. Published in 1971, it was their first hit (and their first publication together). Sure, Tversky and Kahneman became much more sophisticated over time, and also much more complicated. “Advances in Prospect Theory: Cumulative Representation of Uncertainty” is on another level, and the same is true of “Loss Aversion in Riskless Choice: A Reference Dependent Model,” “The Psychology of Preferences,” “On the Study of Statistical Intuitions,” “Variants of Uncertainty,” and of course “Judgment Under Uncertainty: Heuristics and Biases.” Still, “Belief in the Law of Small Numbers” has the core of much of what followed: the harmonies, the effervescence, the humor, the wink and the nod. Amos Tversky and Daniel Kahneman went in all sorts of directions, separately and together, and they were geniuses, separately and together, but back at the beginning, something special was there, and it was, well, Tversky and Kahneman. “Belief in the Law of Small Numbers” emerged from a collaboration that began in 1969, when Kahneman invited Tversky to speak at a seminar that he was teaching. Invoking recent data compiled by Ward Edwards of the University of Michigan, Tversky attempted to show that people were good intuitive statisticians. As always, Tversky was dazzling. Nonetheless, Kahneman thought the idea was preposterous (“I knew that I was not a good intuitive statistician”), and he went at Tversky hard. Tversky almost never lost an argument, but he lost this one. He was excited about that. The two of them became inseparable friends and close collaborators, exploring judgments and intuitions, and lingering, often for hours, over sentences and paragraphs. “Belief in the Law of Small Numbers” was a pathbreaking paper. It found that “people have strong intuitions about random samples; that these intuitions are wrong in fundamental respects; that these intuitions are shared by naive subjects and by trained scientists; and that they are applied with unfortunate consequences in the course of scientific inquiry” (Tversky and Kahneman 1971, p. 105). At the time, that was a declaration of war (“these intuitions are shared by naive subjects and by trained scientists”). In 1971, Tversky and Kahneman did not claim, and might not have known, that their findings were relevant to public policy, including risk-related and environmental policy; but the relevance was hard to miss. For contemporary readers, the law of small numbers seems particularly pertinent to judgments about climate change. Data over short periods might affect judgments, including judgments of high-level policymakers; in fact, recent tempera- 13 In Praise of Computation ture has been found to have an impact on climate change beliefs (Sugerman et al. 2021; Hamilton and Stampone 2013). If patients and doctors overweigh evidence with respect to health over short periods (say, a month or two), belief in the law of small numbers might explain why. The basic problem, Tversky and Kahneman urged, was that “people view a sample randomly drawn from a population as highly representative, that is, similar to the population in all relevant respects” (Tversky and Kahneman 1971, p. 105). If a coin is flipped a few times, people think it is far likelier that it will come up heads half the time than would be predicted by the laws of chan (...truncated)