Reliance on small samples and the value of taxing reckless behaviors

Abstract

New technology can be used to enhance safety by imposing costs, or taxes, on certain reckless behaviors. The current paper presents two pre-registered experiments that clarify the impact of taxation of this type on decisions from experience between three alternatives. Experiment 1 focuses on an environment in which safe choices maximize expected returns and examines the impact of taxing the more attractive of two risky options. The results reveal a U-shaped effect of taxation: some taxation improves safety, but too much taxation impairs safety. Experiment 2 shows a clear negative effect of high taxation even when the taxation eliminates the expected benefit from risk-taking. Comparison of alternative models suggests that taxing reckless behaviors backfires when it significantly increases the proportion of experiences in which a more dangerous behavior leads to better outcomes than the taxed behavior. Qualitative hypotheses derived from naïve sampling models assuming small samples were only partially supported by the data.

Contradictory deviations from maximization: Environment-specific biases, or reflections of basic properties of human learning?

“…Punishing one problematic behavior can move decision-makers to even less desirable behavior. The top panel in Figure 5 summarizes a recent replication of this pattern (Yakobi et al, 2020), using a within-subjects design with 82 Mturk participants. It simulates a situation in which a policymaker tries to reduce the probability of accidents abstracted as losses of 20 points.…”

Attention! Do We Really Need Attention Checks?

“…Note that increased attention to rare losses (resulting in lower checking rates) also implies higher maximization. By contrast, based on previous studies, inattentive participants are hypothesized to pay less attention to rare outcomes (i.e., to underweight rare events; Hertwig et al 2004;Yakobi et al 2020), resulting in higher checking rates. Still, it is also possible that although inattentive participants are less alert to the implications of feedback in general, they will be more sensitive to the most important rare events (Olschewski et al 2023), resulting in lower checking rates in the aggregate.…”

Myopia drives reckless behavior in response to over-taxation

“…Experimental details Yakobi et al (2020) conducted two experiments using a variation of the 𝑛-armed bandit problem (Sutton & Barto, 1998). In this task, participants repeatedly chose between three monetary lotteries for a total of 100 periods.…”

“…Reinforcement learning = delta-rule reinforcement-learning model. Naïve sampler = smallsamples model used byYakobi et al (2020) in Experiment 1. Two-stage naïve sampler = small-samples model that first eliminates one of the two riskier options and then compares the winner with the safe option, as used byYakobi et al (2020) in Experiment 1.…”

“…Naïve sampler = smallsamples model used byYakobi et al (2020) in Experiment 1. Two-stage naïve sampler = small-samples model that first eliminates one of the two riskier options and then compares the winner with the safe option, as used byYakobi et al (2020) in Experiment 1. Extended two-stage naive sampler = Two-stage Inertia, Sampling and Weighting model (I-SAW2) used byYakobi et al (2020) in Experiment 2.…”