Science Explains Why Some People Keep Making Bad Decisions Even After Being Punished

SYDNEY — Someone continues texting while driving despite multiple tickets, or keeps spending beyond their means despite mounting debt. While most people learn to avoid behaviors that hurt them, some individuals persist in harmful actions even when faced with clear negative consequences. New research reveals there are distinct types of people when it comes to learning from punishment, and the differences run much deeper than simple stubbornness.

A study spanning 24 countries has identified three distinct behavioral “phenotypes” that explain why some people adapt to avoid harm while others seem immune to life’s hard lessons. The research challenges common assumptions about human decision-making and reveals that punishment-resistant behavior stems from specific cognitive deficits rather than moral failings or lack of willpower.

The research, published in Communications Psychology, followed 267 participants through an online game designed to measure how people learn from punishment. What emerged was a clear pattern: roughly 26% of participants were “Sensitive,” meaning they quickly learned to avoid harmful actions after experiencing negative consequences. About 47% fell into the “Unaware” category, showing that they failed to learn from experience alone but corrected their behavior once given explicit information about consequences. Most concerning, 27% were classified as “Compulsive,” meaning they persisted in harmful behaviors even after both experiencing punishment and receiving clear information about how to avoid it.

Three Types of People Based on How They Respond to Punishment

Researchers created an online game called Planets & Pirates where participants could click on two different planets to earn points. Initially, both planets offered equal rewards. But during the punishment phase, clicking on one planet began triggering “pirate attacks” that caused significant point losses, while the other planet remained safe.

Participants were randomly assigned to groups where pirate attacks occurred either frequently (40% of the time) with mild point losses, or rarely (10% of the time) with severe losses. After three rounds of punishment learning, researchers provided explicit corrective information, telling participants exactly which planet attracted pirates and which was safe. They then observed whether this information changed behavior in a final game round.

Sensitive participants figured out the pattern quickly through experience alone. These individuals showed clear behavioral adaptation, shifting their choices away from the punished option and toward the safe alternative during the punishment blocks. They accumulated points throughout the game while others lost them.

Unaware participants seemed oblivious to the connection between their actions and negative outcomes during the initial punishment phase. They continued clicking on both planets equally, losing points in the process. However, once researchers explicitly explained which planet attracted pirates, these individuals immediately corrected their behavior and avoided further losses.

Most intriguingly, the Compulsive group showed a puzzling disconnect between knowledge and action. Even after receiving the same explicit information as the Unaware group — and demonstrating they understood it by passing comprehension tests — they continued making choices that led to punishment.

Why Some Still Make Bad Decisions

Two distinct cognitive failures drive punishment insensitivity. The first, affecting the Unaware group, involves what scientists call “causal inference deficits.” Essentially, these individuals fail to correctly connect their specific actions with negative outcomes. They experience the punishment but can’t figure out what’s causing it.

The second mechanism, affecting the Compulsive group, involves problems with “cognitive-behavioral integration.” These individuals can correctly identify cause-and-effect relationships after receiving information, but struggle to translate that knowledge into behavioral change.

Understanding why experiential learning often fails has broad implications for policy and intervention design. The researchers noted that causal inference deficits may help explain why experiential feedback often fails to drive behavior change, an insight that sheds light on the limits of punishment-based policies like fines and sanctions.

Punishment Response Patterns Remain Stable Over Six Months

Perhaps most striking, these behavioral patterns proved remarkably stable. When researchers retested 128 participants six months later, they found the same three phenotypes emerged. A majority of participants maintained their original phenotype classification during retesting, demonstrating the stability of these decision-making styles over time.

Behavioral phenotyping at baseline successfully predicted individuals’ learning trajectories and choice patterns six months later—outperforming traditional self-reported measures of cognitive flexibility in forecasting long-term decision-making outcomes. These patterns represent stable, trait-like characteristics rather than temporary behavioral fluctuations.

What This Means for Changing Human Behavior

These discoveries illuminate persistent puzzles in human behavior, from why some people repeatedly engage in risky behaviors despite negative consequences, to why public health campaigns and warning labels often fail to change behavior in certain populations.

For the Unaware group, interventions that provide clear, explicit information about consequences could prove highly effective. These individuals possess intact behavioral control systems: they simply need help connecting the dots between actions and outcomes.

However, the Compulsive group presents a more challenging scenario. Traditional approaches that rely on providing information or increasing punishment severity are likely to prove ineffective. These individuals require interventions that target the disconnect between knowledge and behavior.

Rather than applying one-size-fits-all solutions, interventions could be tailored to address the specific cognitive mechanisms underlying each person’s punishment insensitivity.

What’s clear is that human decision-making is far more complex than simple cost-benefit calculations. For roughly one in four people, the usual assumption that negative consequences naturally lead to behavioral change simply doesn’t hold true; and addressing this reality requires fundamentally different approaches than society currently employs.

Disclaimer: This report summarizes findings from a study published in Communications Psychology. The study was conducted in a controlled online environment and may not fully capture the complexity of real-world decision-making. While the identified behavioral types were stable over time, individual behavior may still vary in different contexts. The research is not intended to diagnose or label individuals but to inform future policy and intervention strategies.