The relationship between measured cognitive ability and occupational success is one of the most studied questions in applied psychology. Across five decades of research — spanning hundreds of studies, millions of subjects, and dozens of countries — a remarkably consistent picture has emerged: general intelligence is the single strongest predictor of job performance available to employers, more predictive than experience, education, or personality assessments taken alone.

The Schmidt and Hunter Meta-Analysis

The landmark contribution to this literature is Frank Schmidt and John Hunter's 1998 meta-analysis published in Psychological Bulletin. Synthesising 85 years of research across 515 studies, Schmidt and Hunter found that general mental ability (GMA) — essentially g as measured by cognitive tests — had a corrected validity of 0.51 for predicting job performance across all job types. This is a remarkably high figure in the social sciences, where most predictor-outcome relationships in the real world fall below r = 0.30.

The relationship is not uniform across all roles. For low-complexity jobs — manual labour, routine clerical work — the validity of GMA is lower (around 0.23). For medium-complexity roles — skilled trades, sales, police work — it rises to 0.40. For high-complexity professional, managerial, and scientific roles, validity reaches 0.58. The more cognitively demanding the role, the more powerfully intelligence predicts performance. This makes intuitive sense: complex roles require rapid learning, novel problem-solving, adaptation to changing circumstances, and managing multiple streams of information simultaneously — all things that g directly supports.

Terman's Study and the Long View

For long-horizon evidence, Lewis Terman's famous "Termites" study — a longitudinal follow-up of 1,528 gifted Californian children (IQ 135+) begun in 1921 and continued for decades — remains instructive. Contrary to the popular myth of the brilliant misfit, Terman's sample showed dramatically elevated rates of professional achievement, occupational success, health outcomes, and personal stability compared to the general population. They published thousands of books and scientific papers, became doctors, lawyers, engineers, and professors at extraordinary rates. Subsequent analysis has found that even within this gifted group, those with higher IQs outperformed those at the lower end — the dose-response relationship continues even far above average.

The National Longitudinal Survey of Youth (NLSY97) provides more recent and more representative evidence. Studies using this dataset consistently find that IQ measured in adolescence predicts income, occupational status, and educational attainment in adulthood — relationships that persist after controlling for family background, education, and personality. Remarkably, these longitudinal effects strengthen rather than weaken with age, as the cumulative advantages of superior learning speed and information-processing compound over time.

The Diminishing Returns Hypothesis

One of the more debated questions in the field concerns whether the relationship between IQ and success is linear or whether it levels off above some threshold. The "threshold hypothesis" suggests that beyond roughly IQ 120, additional increments of measured intelligence add little to real-world achievement. Some early studies supported this view, but larger and more carefully controlled research has consistently rejected it. A landmark 2014 study by Kell, Lubinski, and Benbow, following mathematically gifted adolescents (top 1%) for 25 years, found strong linear relationships between SAT scores (a proxy for g) and patent applications, peer-reviewed publications, and tenure at research universities — suggesting that even at high levels, more is better.

However, the contribution of other traits grows stronger as IQ becomes less of a binding constraint. Conscientiousness, emotional regulation, motivation, and social skills — which are only weakly correlated with IQ — become increasingly important differentiators among people who all sit above a cognitive threshold sufficient for a role. Above some minimum, who works hardest and most effectively with others matters enormously.

Key Takeaway

The evidence that cognitive ability predicts occupational performance is not seriously disputed in the scientific literature. IQ is the strongest single predictor of job success available, especially for complex roles, and its effects compound across a career. What this does not mean is that intelligence is the only thing that matters: conscientiousness, interpersonal skill, and motivation are all significant contributors to success that correlate only weakly with IQ. The practically useful conclusion is that intelligence sets the upper bounds of learning speed and complexity management, while other traits determine how consistently and effectively that potential is realised.