What research says about IQ and real-life outcomes

Cognitive test scores show consistent but moderate associations with educational attainment and some work outcomes at group level. Causality runs in more than one direction, since education itself raises measured scores. Associations with health and longevity have been reported but are confounded by socioeconomic factors. For any individual, prediction is weak: score bands contain people with very different lives.

Ritchie and Tucker-Drob's meta-analysis found that additional years of education are associated with higher measured intelligence, which complicates any simple story that intelligence causes educational success. Both directions operate, alongside shared causes such as family circumstances, health and schooling quality. Cross-sectional correlations can't separate these on their own.

Schmidt and Hunter's 1998 review reported a corrected correlation of about 0.5 between general mental ability and job performance, a figure that has shaped personnel selection practice ever since. Richardson and Norgate argued that such estimates depend on corrections for range restriction and criterion unreliability, on supervisor ratings as the measure of performance, and on samples that may not generalize, and that the uncorrected associations are substantially weaker. Reasonable readers disagree about the size. What is not defensible either way is treating the relationship as deterministic for an individual: even a correlation of 0.5 leaves roughly three-quarters of the variation in performance unexplained.

One of the most cited health results comes from the Scottish Mental Survey cohorts: Calvin and colleagues followed nearly 66,000 people for 68 years and found higher childhood intelligence test scores associated with lower death rates from causes including cardiovascular disease and respiratory illness. The association survived adjustment for socioeconomic status in their analyses, but the authors were careful about mechanism: education, health behaviors such as smoking, and access to care are all plausible pathways, and an observational cohort cannot settle which.