IQ and Income: What Longitudinal Studies Have Found

Does a higher IQ lead to higher earnings? It is a question that surfaces in popular media, policy debates, and personal curiosity alike. Longitudinal studies — which follow the same individuals over many years — have generated more reliable evidence on this question than any other research design. The short answer: there is a moderate, consistent correlation between IQ and income, but it explains far less of the variance in earnings than many assume, and a range of non-cognitive factors account for much of what IQ leaves unexplained.

1. Why longitudinal studies matter for this question

Cross-sectional surveys can compare IQ scores and income at a single point in time, but they cannot establish which came first or rule out confounding variables. Longitudinal studies solve several of these problems by:

Measuring cognitive ability in childhood or early adulthood, before income outcomes are observed
Tracking the same individuals across decades, allowing genuine career trajectories to be observed
Collecting data on education, occupation, family background, and personality — variables needed to disentangle IQ's independent contribution

The major studies in this area include the National Longitudinal Survey of Youth (NLSY79 and NLSY97) in the United States, the British Cohort Studies, the Scottish Mental Surveys, and the Terman Life-Cycle Study. Each has different strengths and limitations, but their combined picture is reasonably consistent.

2. What the data consistently shows

The correlation is real but moderate

Across major longitudinal datasets, the correlation between IQ scores and adult income typically falls in the range of r = 0.30 to 0.50. This means IQ accounts for roughly 9–25 % of the variance in individual income — real, but leaving 75–91 % of the variance unexplained by IQ alone.

A landmark analysis by Charles Murray using NLSY79 data found that, even after controlling for socioeconomic background, individuals in higher IQ bands earned substantially more on average. However, within each IQ band, income spread remained wide — a high IQ did not guarantee high earnings, and low IQ did not preclude them.

The relationship is strongest at the tails

The IQ–income link is not perfectly linear across the distribution. Research suggests:

Moving from average IQ (around 100) to the 85th percentile (IQ ~115) is associated with meaningfully higher median income
Further gains above IQ 130 show diminishing returns relative to the gains from the 85–115 range
Below the average, the relationship also steepens — individuals scoring well below average face structural labor market barriers that compound the cognitive factor

Education partially mediates the relationship

A significant portion of the IQ–income correlation operates through educational attainment. Higher IQ correlates with more years of schooling, higher degrees obtained, and access to credential-gated occupations — and these credentials then command income premiums. When studies statistically control for education, the direct IQ–income correlation shrinks, typically to r = 0.20–0.30. This does not mean IQ has no direct effect; it means some of its influence is channeled through school completion.

3. Key study findings at a glance

Study	Sample	IQ measure	Key finding
NLSY79 (U.S.)	~12,000, born 1957–64	AFQT (age 15–23)	IQ correlated with income at r ≈ 0.40; large within-IQ-band variance
British Cohort Study 1970	~17,000, born 1970	Age-5 cognitive tests	Childhood cognition predicted adult wages even after controlling for parental background
Scottish Mental Survey 1947 follow-up	1,000+, born 1936	Age-11 IQ	IQ at 11 modestly predicted income in midlife; occupational level was a stronger mediator
Terman Life-Cycle Study (U.S.)	~1,500 gifted children	Top 1% IQ	Wide income variance within the gifted group; personality and opportunity factors differentiated high earners
Project TALENT (U.S.)	~440,000, born 1946–47	Multiple cognitive tests	Cognitive composite predicted earnings above and beyond education alone

4. What IQ does not explain

The majority of income variance remains unexplained by IQ. Research points to several competing and complementary predictors.

Conscientiousness and non-cognitive skills

Studies using Big Five personality data alongside IQ consistently find that conscientiousness — reliability, self-discipline, follow-through — is an independent predictor of income. In some analyses, conscientiousness accounts for income variance roughly comparable to IQ, after cognitive ability is controlled for. Heckman and Kautz (2012) reviewed evidence suggesting non-cognitive skills rival IQ in predicting labor market outcomes.

Social capital and networks

Who someone knows, where they went to school, and what industry connections they inherit or develop play large roles in career trajectories, particularly for access to high-income opportunities that are never formally advertised.

Occupational sorting and industry

Average income varies enormously across industries and occupations — often more than IQ variation within those fields predicts. A highly cognitively able individual who enters a low-wage sector may earn less than a moderately able individual in a high-wage sector, for reasons unrelated to cognitive skill.

Luck and timing

Economic cycles, geographic labor markets, and being in the right place at a particular moment contribute variance that no individual characteristic can explain in advance.

Parental socioeconomic status

IQ and parental income are themselves moderately correlated (due to both genetic and environmental factors), making it difficult to fully separate IQ's independent contribution from the advantages that often accompany higher childhood IQ in practice.

5. The "g factor" debate and income

Some researchers argue that general intelligence (g) — the common factor running through diverse cognitive tests — is the most powerful predictor, and that specific cognitive subtypes matter less once g is accounted for. The practical implication: if IQ tests reliably estimate g, their income-predictive validity is largely about this broad factor.

Others argue that specific cognitive skills — verbal reasoning, quantitative ability, spatial ability — independently predict income in specific occupational domains. A spatial-ability advantage may matter more for engineering salaries than verbal reasoning does; the reverse may hold in law or writing-intensive fields.

The evidence does not fully settle this debate. What it does confirm is that broad cognitive ability matters more for income prediction than any narrow subskill measured alone.

6. Important limitations of the research

No longitudinal study in this area is free from methodological challenges:

Attrition: studies lose participants over decades; those who drop out often differ systematically from those who remain, potentially biasing results
IQ measurement timing: tests taken in childhood or adolescence may not perfectly represent adult cognitive ability
Era effects: studies conducted in the 1970s–1990s may reflect labor market structures that differ from today's, when credential inflation, automation, and the knowledge economy have reshaped which skills command premiums
Causation vs. correlation: correlation does not establish that higher IQ causes higher income; it is consistent with both a direct causal path and a third-variable (e.g., genetic advantage affecting both) explanation
Publication bias: studies finding significant IQ–income correlations are more likely to be published than those finding null results

Frequently asked questions

What is the actual correlation between IQ and income?

Most large longitudinal studies report correlations between r = 0.30 and r = 0.50, depending on the sample, the IQ measure used, and whether the analysis controls for education or other variables. This translates to IQ accounting for roughly 9–25 % of the variance in income. The relationship is real but far from deterministic — large income variation exists at every IQ level.

Does a high IQ guarantee a high salary?

No. The correlation between IQ and income is moderate, not deterministic. Research consistently shows wide income variation within every IQ band. Factors such as occupational choice, conscientiousness, social capital, industry, and timing all influence earnings in ways that IQ scores do not predict. The Terman study of gifted individuals found enormous income spread within a group that all scored in the top 1 % on IQ — clearly demonstrating that very high IQ is neither necessary nor sufficient for high income.

Does IQ or education matter more for income?

They are related, and it is difficult to fully separate them. IQ influences educational attainment; education then influences income. When studies control for education, the direct IQ–income correlation is smaller but remains significant. When studies control for IQ, the education–income correlation also shrinks. Both contribute independently, and neither alone explains most of the income variance. Research by Cawley, Heckman, and Vytlacil suggests that in some datasets, education's conditional effect on wages is larger than IQ's conditional effect, once both are accounted for simultaneously.

Is the IQ–income link stronger in some countries than others?

Cross-national data is limited, but available evidence suggests the correlation varies with labor market institutions. Countries with strong credentialism or meritocratic hiring may show stronger IQ–income links, while countries with stronger wage compression (higher union density, narrower pay scales) may show weaker correlations. Most of the major longitudinal data comes from the United States and the United Kingdom, so conclusions about other countries should be drawn cautiously.

What do these findings mean for individuals?

Correlations describe group-level tendencies, not individual destinies. At the individual level, IQ score is one factor among many — and not necessarily the most important one. Choices about occupational field, sustained effort, skill development in high-demand areas, networking, and financial decision-making all shape earnings in ways that population-level correlations cannot capture. Understanding the research is useful context; treating it as a personal forecast is not warranted.

Are online IQ tests useful for predicting income?

Online cognitive tests, including Brambin's, are designed for self-exploration and are not validated for predicting income or other life outcomes. Even clinical IQ tests — validated, professionally administered — account for only a portion of income variance. An online score carries additional measurement uncertainty and should not be used to draw conclusions about financial prospects.

Summary

Longitudinal studies consistently find a moderate positive correlation between IQ and income, with most estimates in the r = 0.30–0.50 range. This is a meaningful statistical relationship, but it leaves the majority of income variance unexplained. Education partially mediates the link; personality traits, social capital, occupational choice, and structural economic factors explain much of what remains.

For individuals, these correlations are best understood as probabilistic tendencies at the group level — not reliable predictors for any single person. High IQ does not guarantee high income, and lower IQ does not preclude it. The research supports the value of cognitive ability in labor markets while simultaneously making clear that it is one of many relevant factors.

Brambin offers an eight-dimension cognitive profile for self-exploration. It is not a clinical assessment and is not intended for diagnosis, educational placement, or financial forecasting. Treat any online score — ours included — as a starting point for curiosity, not a verdict.

IQ and Income: What Longitudinal Studies Have Found