Average IQ in the USA: What the Data Shows

What is the average IQ in the USA?

The average IQ in the USA is 100. That isn’t a coincidence or a measurement: it is built into the tests. The most widely used IQ tests in the world, the Wechsler Adult Intelligence Scale (WAIS), the Wechsler Intelligence Scale for Children (WISC) and the Stanford-Binet, are normed on American samples. The middle score of each US age group is defined as 100.

So asking “what is the average IQ in America?” is a bit like asking “what is the average percentile in America?” The answer is fixed by the method. What you can usefully ask is how the norms are built, how they shift over time, and what international comparisons do and don’t show.

Why do US norms center on 100?

Because the US population is the reference group. When a publisher releases a new edition of a Wechsler test, it gives the test to a standardization sample, about 2,200 people for recent adult and child editions, chosen to match US census figures on age, sex, race and ethnicity, education level and region.

The steps:

1. Each person takes the full battery under standard conditions.
2. Results are grouped by age band.
3. Within each band, raw scores are converted so the middle score becomes 100 and one standard deviation becomes 15 points.

The result is a table that turns any American’s raw performance into a rank against other Americans their age. If you take the test in the US, you are being compared with that US sample. For how the bands and percentiles work, see the IQ scale and ranges chart.

Other countries often build their own norms for translated versions (there are UK, Canadian, Australian and many other adaptations), and each of those is centered on 100 for its own population. That’s one reason you can’t line up two national averages from two national test editions: each one is 100 by its own definition.

Does the average shift over time?

The reported average doesn’t, but the underlying performance does. The Flynn effect, the steady rise in raw test performance across the 20th century, was first documented in detail by James Flynn using American data in 1984. He showed that Americans scored better on older test norms than the people those norms were built on, at a rate of roughly 3 points per decade.

In practice this matters in a few ways:

• Old norms inflate scores. Someone tested on a version normed 20 years earlier will tend to score higher than on a freshly normed version.
• Renorming resets the bar. Publishers release new editions every 10 to 20 years, and each one sets the current population’s middle back to 100.
• High-stakes cases notice. In US legal settings, especially death penalty cases where an intellectual disability claim depends on a score near 70, courts have heard arguments about adjusting for outdated norms. The details vary by jurisdiction.

Whether the Flynn effect is still running in the US is less clear. Some countries, notably Norway, show a reversal for cohorts born since the mid-1970s. US evidence is mixed and depends on which tests and which age groups you look at.

How does the US compare internationally?

Any international comparison of IQ depends on datasets that are heavily contested, so the honest answer is: nobody knows the US “rank” with any precision.

The best-known national IQ figures come from Richard Lynn and Tatu Vanhanen (IQ and the Wealth of Nations, 2002, and later works) and revisions of their dataset by other authors. Those tables set British norms as the reference point of 100 and place the US close to that reference, around the middle of the developed countries. The criticisms of the dataset apply to every country in it:

• Uneven samples. Some countries are represented by large, careful samples; others by a few hundred people from one city or one school.
• Different tests and decades. Results from different tests given decades apart are compared directly, even though the Flynn effect alone can shift scores by 10 points or more over that span.
• Estimated values. Early versions filled gaps by estimating a country’s figure from its neighbors.
• Disputed uses. The tables have been used to support claims about national and ethnic differences that many researchers reject, and reanalyses by other researchers produced different numbers for some regions.

International school assessments such as PISA (for 15-year-olds) and TIMSS (math and science) are much better documented. They aren’t IQ tests, though, and converting their scores into IQ points adds another layer of assumptions. On those assessments, US results have tended to sit around the average of wealthy countries, stronger in reading than in math. That’s a statement about school achievement, not about intelligence.

For a wider look at country data, see the average IQ by age, country and job.

What about IQ by state, or by group?

There is no dataset of IQ test results for each US state. Lists that rank states by IQ are usually derived from school achievement scores (such as the National Assessment of Educational Progress) and converted into IQ-style numbers. Those state figures move with poverty rates, school funding, immigration patterns and demographics, so they say more about schools and economics than about any fixed trait.

Group differences in average test scores within the US do exist and are a long-running and heated research topic. The gaps have changed over time, which points to environmental influences, and their causes remain disputed. For any individual, group averages say almost nothing: the variation inside every group is far larger than the differences between group averages.

Common myths about American IQ

A few claims come up again and again. Here is how they hold up:

• “Americans are getting dumber.” There’s no clean evidence for a large US decline. Some countries show a recent reversal of the Flynn effect, and some US studies report flat or falling scores on particular tasks, but the picture is mixed and depends on the test and the age group.
• “The US average is below 100.” Only on someone else’s ruler. Against US norms it is 100 by definition; against British norms in contested international tables it lands near 100, and the exact gap depends on which samples and decades are compared.
• “The average American reads at a low grade level, so the average IQ must be low.” Literacy surveys and IQ tests measure different things, on different scales, for different purposes. One can’t be converted into the other.
• “Online test sites know the US average.” They know the average of the people who visited their site, which is a self-selected group, not a sample of the population.

What should you do with the number 100?

Use it as a reference point, nothing more. If you take a properly administered clinical test in the US, you’ll get a score that tells you where you sit relative to Americans your age, with a margin of error of a few points. What your IQ test score actually means covers how to read that range.

If you take an online quiz, the number has no such anchor. Most online tests aren’t normed on any representative sample, US or otherwise, so “you scored above the US average” is a marketing line, not a finding. The free Mentarium IQ practice test is honest about that: it’s practice with the kinds of items IQ tests use, not a clinical score.

The Mentarium take

“The average IQ in the USA” sounds like a fact about Americans. It’s really a fact about how tests are built: the US sample is the ruler, so the US average is the 100 mark on it. Most of the noise around this topic, viral country maps, state rankings, “America is getting dumber” headlines, comes from forgetting what the ruler is and who made it.

That’s a pattern worth spotting everywhere. When a number confirms something you already believe about a country or group, that is exactly when to check the sample, the test and the date. Confirmation bias feeds on clean-looking rankings. The habits in critical thinking are the cheap defense: ask how the number was made before you repeat it.