Key Takeaways
High IQ did not mean less crime in this self-report study
The central result surprised. Oleson compared 465 high-IQ respondents with 756 controls. The high-IQ group included members of a high-IQ society, university participants, and incarcerated people. Its estimated mean IQ was 148.7, compared with 115.4 for controls, a group that was itself well above average. Participants reported whether they had committed 72 offenses, ranging from minor rule-breaking to serious crimes.
The pattern was not simple. High-IQ respondents reported higher lifetime prevalence for 50 of the 72 offenses. Across seven of nine broad offense categories, their prevalence rose with IQ. Yet the correlation between IQ and a weighted crime score was only -0.02, essentially no relationship. These findings challenge a simple claim that crime steadily declines as IQ rises. They do not establish that intelligence causes crime: the data are self-reported, selected, and not representative of the population.
The study’s strongest contribution is not a reversal of the usual IQ-crime story, but a warning against treating that story as universal. A sample with unusually high intelligence, including people already incarcerated, cannot represent all high-IQ adults. The control group’s average IQ of 115 also makes this a comparison between bright and exceptionally bright participants, not between genius and the general public. The result is best read as a research question sharpened by unexpected evidence, not a final estimate of how much crime highly intelligent people commit.
Read crime statistics as a map of detection, not total harm
Recorded crime is incomplete. Official statistics count offenses that are reported, recognized, and recorded by authorities. Victim surveys miss crimes without identifiable victims, such as illegal drug use, and may miss harms victims do not recognize, such as pollution or financial manipulation. Self-report studies can reveal conduct that never appears in police records. Oleson calls the unobserved portion the dark figure of crime.
A striking earlier study found that 99% of 1,698 adults reported at least one of 49 listed offenses. That result does not mean nearly everyone commits equally serious crimes. It shows how strongly rates depend on what researchers define and ask about. Oleson argues that powerful people’s harmful conduct may be overlooked or treated less seriously than visible street crime, while acknowledging that broad offense lists can blur crucial differences.
The distinction between harm and recorded crime is essential for sound policy. A database is not a transparent window onto wrongdoing; it reflects reporting habits, legal definitions, enforcement priorities, and institutional access. This insight also demands care. Counting a minor workplace misuse alongside violence can expose hidden conduct, but it can also make unlike acts look equivalent. A credible account of crime therefore needs multiple measures: who was harmed, how serious the harm was, and why the conduct entered official records at all.
Match explanations to offenses, because IQ-crime patterns change by category
Different crimes require different opportunities. The study separated 72 offenses into nine categories rather than treating crime as one behavior. In its IQ-band trend analysis, prevalence rose with IQ for white-collar crime, property crime, professional misconduct, drug crime, sex crime, violent crime, and vehicular offenses. The trends were particularly strong for white-collar offenses, property offenses, and professional misconduct. For example, the study estimated a 2.93 percentage-point rise in white-collar prevalence for each 10-point IQ increase in its sample.
These are patterns, not causes. Other categories and individual offenses moved in the opposite direction. The author’s detailed analysis found negative trends for justice-system and miscellaneous offenses, among others. Intelligence may shape which opportunities people can access, how they plan, or how they respond to risk. The survey cannot isolate those mechanisms, and its trend lines should not be mistaken for proof of causation.
Disaggregation improves the question. A single measure of crime can hide opposing relationships, much as an average can conceal important differences between groups. Here, education and professional access could help explain why some offenses rise with IQ without implying that IQ itself produces criminal intent. The same logic applies to policy: interventions aimed at impulsive street offenses may not address fraud or workplace misconduct. The study’s categories remain imperfect, but its key analytical move is valuable: ask which conduct, under which opportunities, and with what consequences.
Measure offenses per arrest before calling a group more criminal
The denominator changed the picture. High-IQ respondents reported more arrests and convictions overall than controls, partly because their group included incarcerated participants. But they also reported far more offenses. Across all offenses, the index group reported one arrest per 172.8 offenses, compared with one per 129.1 among controls. It reported one conviction per 304.5 offenses, compared with one per 193.1 among controls.
This fits two competing explanations. The differential detection hypothesis proposes that high-IQ offenders are less likely to be detected. The differential reaction hypothesis proposes that authorities respond differently to them. The study’s ratios are consistent with either, but cannot distinguish between them. Self-reporting, unequal samples, and inconsistent interpretations of arrest all complicate the comparison. More offenses per arrest is suggestive evidence, not proof that intelligence helps people evade justice.
Rates answer different questions depending on their denominator. Counting arrested people may make one group appear more criminal, while comparing arrests with reported offenses may make it appear more successful at avoiding detection. Neither statistic alone settles the issue. The study usefully makes that tension visible, but its self-reported offense totals cannot verify how many acts occurred or how comparable the groups were. Stronger evidence would link detailed self-reports to independently checked records while preserving participant confidentiality.
Strengthen social bonds where exceptional ability becomes isolation
Relationships can restrain offending. Hirschi’s social bond theory explains conformity through four connections: emotional ties to other people, investment in conventional goals, participation in ordinary activities, and acceptance of shared rules. In 44 follow-up interviews, high-IQ participants described experiences that could weaken these connections. Some recalled being stigmatized as children; others felt alienated, dismissed conventional careers, had unstructured time, or distrusted laws they viewed as hypocritical.
The findings are suggestive. Several participants connected loneliness and failed relationships with destructive choices. One man described sexual offending as a desperate attempt to feel close to someone, an explanation that does not excuse the harm. The interviews show how participants understood their behavior; they cannot prove that isolation caused crime, and there was no interview comparison group. Still, they suggest that prevention should address belonging and meaningful participation, not just intelligence or punishment.
This account makes social connection more than a warm abstraction. Attachment can provide feedback, commitment gives future consequences weight, involvement limits idle opportunity, and belief makes rules feel legitimate. The framework also avoids the trap of treating intelligence as a complete explanation for conduct. Yet interviews are retrospective stories, shaped by memory and self-presentation. Alienation should not be used as a proxy for dangerousness, either. The useful question is whether institutions can offer gifted people genuine challenge and belonging without confusing nonconformity with pathology.
Notice how criminal involvement can grow through belonging and rationalization
Faulkner’s story shows a gradual path. In interviews, a man who reported an IQ of 162 described meeting Vince at 17, then being drawn into a criminal circle through attention, protection, money, and a sense of status. After his first shooting, Vince praised him and reframed the victim as deserving it. Faulkner said he later justified violence in stages, from self-defense to protecting reputation. He claimed to have killed 15 people, but the researcher could not verify that central claim.
The lesson is about escalation. The account depicts how a person can cross boundaries by redefining each act as necessary and by receiving approval from a group. The story is compelling precisely because it is not independently established. It illustrates a possible process, not a typical trajectory or reliable count of murders.
Faulkner’s account resembles a social process more than a portrait of an innate criminal type: trust creates access, access normalizes conduct, and group approval recasts violence as duty. That sequence is psychologically plausible, but plausibility is not verification. The researcher’s caution matters because dramatic narratives can gain authority simply by being vivid. Readers should take the account as a case for asking how belonging and moral justifications shape action, not as evidence that high intelligence commonly leads to organized violence.
Separate reasoning power from the self-control that governs behavior
IQ is not the whole mind. The book distinguishes general intellectual ability from executive function, the processes that organize thinking and regulate behavior, including inhibition and decision-making. Its analogy is an engine and transmission: IQ supplies cognitive power, while executive function directs how that power is used. A person can reason quickly yet struggle to control impulses, or show strong regulation without exceptional test scores.
The personality results add nuance. Some high-IQ respondents scored below age-matched norms for impulsiveness and above them for venturesomeness, a willingness to take considered risks. They also scored below norms for empathy. The study did not give the same personality questionnaires to controls, so these patterns cannot explain why the high-IQ group reported more offenses. Intelligence may help someone plan; it does not guarantee restraint, concern for others, or lawful choices.
Separating capacity from regulation prevents a common mistake: assuming that a person who can understand consequences will reliably act on that understanding. Planning, inhibition, empathy, and motivation are related but distinct capacities. That distinction helps explain why cognitive sophistication can serve prosocial or harmful ends without making intelligence itself morally directional. The study’s personality comparisons are limited because they use external norms rather than a matched control group. Its conceptual distinction is stronger than any claim about the particular scores.
Treat IQ as a sentencing clue, not a verdict on culpability
The legal question has no easy answer. An IQ near 130 is roughly as far above the average as an IQ near 70 is below it. Oleson uses the logic of Atkins v. Virginia, which barred execution of intellectually disabled defendants, to ask whether extreme intelligence should also affect culpability. One argument says greater foresight and reasoning can make deliberate wrongdoing more blameworthy. Another says profound social isolation or unusual moral reasoning could complicate judgments about responsibility.
Neither symmetry nor genius decides a case. The author notes that a legal insanity defense would require much more than a high test score, including evidence about what the person understood at the time. The comparison is a prompt for examining how cognitive differences affect judgment, not a claim that gifted defendants are legally impaired. IQ alone cannot establish innocence, illness, greater guilt, or lesser guilt.
A numerical mirror is not a moral mirror. Equal distance from an average score does not establish equal impairment, equal social experience, or equal responsibility. Still, the question exposes a real tension in punishment: law often treats people as equally capable once they cross a threshold, even when their capacities differ. Individualized evidence is more defensible than turning high IQ into either a privilege or an aggravating label. The study opens a jurisprudential problem; it does not supply a courtroom rule.
Analysis
Oleson’s central achievement is to make the criminal genius an empirical question rather than a familiar fictional character. The book moves from centuries of cultural images, from rebel angels to Moriarty, into an ambitious self-report study of adults rarely represented in criminology. Its most provocative result is not that intelligence predicts more crime. The crime-score correlation is nearly zero. Instead, the findings undermine a single, linear account of criminality and show why offense type, opportunity, visibility, and enforcement must be analyzed together.
The evidence is valuable but sharply bounded. The index group combines high-IQ society members, elite university participants, and incarcerated respondents, while the control group is not a population sample and averages 115.4 IQ. The groups differ in age, sex, education, and other characteristics. IQ is sometimes inferred, the surveys are separated by seven years, and no external records validate the reported conduct. The offense list also combines trivial rule violations with serious harm, and its seriousness weights come from an older study. These conditions make the results exploratory, not estimates of population crime rates or causal effects.
The book’s strongest conceptual move is to connect individual cognition with institutions. Intelligence may shape access to professional roles, the ability to plan, and the odds of detection, while class and institutional discretion shape which harms become visible and punishable. These mechanisms can coexist. A single score cannot tell whether an offense reflects opportunity, impulse, social isolation, or deliberate exploitation.
The social bond interviews supply a plausible account of alienation and rationalization, but they cannot establish causality. Faulkner’s dramatic story is best treated as a hypothesis-generating narrative because its most serious claims remain unverifiable. Overall, the book’s enduring value is methodological: study hidden conduct across the IQ spectrum, separate cognitive ability from self-regulation, and measure harm as well as arrest. It expands the field of inquiry without pretending to close it.
Glossary
dark figure
unrecorded crimeThe portion of crime that does not appear in official statistics because it goes unreported, undetected, unsolved, or unrecorded. It includes conduct victims do not recognize and offenses without obvious individual victims. Oleson uses the idea to argue that recorded crime reflects social recognition and enforcement as well as actual behavior.
differential detection hypothesis
unequal odds of detectionThe hypothesis that people with higher intelligence may commit offenses at comparable rates yet be less likely to be detected, arrested, prosecuted, or convicted. It offers one explanation for why official criminal populations often show lower average IQ. Research reviewed in the book is mixed, and the study’s offense-to-arrest ratios cannot prove the hypothesis.
differential reaction hypothesis
unequal justice-system responseThe hypothesis that authorities respond differently to offenders based on characteristics such as intelligence, demeanor, or ability to communicate. It proposes that differences in arrests or convictions may reflect official discretion, not only differences in offending or detection. The book presents this as a possible explanation, but its data cannot separate it from differential detection.
social bond theory
social ties restrain offendingTravis Hirschi’s theory that conformity is supported by ties to people and institutions. It describes attachment to others, investment in conventional goals, participation in ordinary activities, and acceptance of shared rules as connections that can discourage crime. Oleson applies the theory to interview accounts of isolation, weakened commitment, unstructured time, and distrust of law.
executive function
control of thinking and behaviorA set of mental processes involved in organizing cognition and guiding action, including decision-making, memory, inhibition, and behavioral regulation. The book distinguishes executive function from IQ: high intellectual ability does not necessarily mean strong self-control. This distinction matters because criminal behavior may involve regulatory difficulties that an IQ test does not capture.
crime score
weighted offense measureOleson’s method for comparing participants who reported different kinds and numbers of offenses. Reported offenses were weighted with seriousness values, largely adapted from Sellin and Wolfgang’s 1964 study, and then combined into an individual score. The measure allows comparison across offense types, but its weights are dated and cannot fully represent context, intent, or victim impact.
EPQ-R
personality assessment questionnaireThe revised Eysenck Personality Questionnaire is a self-report instrument used in the study to assess psychoticism, extraversion, neuroticism, and a lie scale, among other measures. Oleson also used its related scales to examine addiction and criminality. The control group did not complete the questionnaire, limiting direct personality comparisons between groups.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.