Introduction: What National IQ Averages Tell Us -- and What They Do Not
National IQ averages are among the most cited, most debated, and most frequently misinterpreted statistics in all of psychology. These figures -- estimates of the mean cognitive test performance of a country's population -- attract enormous public interest, fuel contentious policy debates, and appear regularly in headlines ranking nations from "smartest" to "least intelligent."
Yet the researchers who produce this data are often the first to caution against simplistic readings. Richard Lynn and Tatu Vanhanen, whose IQ and the Wealth of Nations (2002) popularized cross-national IQ comparisons, faced intense criticism for methodological shortcomings. More recent datasets by David Becker (the NIQ-Dataset) and James Flynn's country-level analyses have improved data quality but have not eliminated fundamental challenges.
"National IQ averages are rough indicators at best. They should never be taken as precise measurements of a nation's cognitive capacity." -- James Flynn, political scientist, University of Otago
This article presents the available data on national IQ averages, examines the Flynn Effect across regions, and -- critically -- explains the methodological limitations that make cross-national comparisons far less reliable than they appear.
How National IQ Averages Are Calculated
Before examining the data, it is essential to understand how these numbers are produced, because the methodology directly affects their reliability.
Data Sources and Methods
National IQ estimates come from several sources:
- Standardized IQ test administrations -- Tests like the Wechsler scales (WAIS, WISC) or Raven's Progressive Matrices administered to representative national samples
- International student assessments -- PISA, TIMSS, and PIRLS scores converted to IQ-equivalent scales
- Military conscription data -- Some countries (Israel, Finland, Denmark, Norway) test all conscripts, providing large representative samples
- Extrapolation from neighboring countries -- When direct data is unavailable, estimates are derived from geographically or culturally similar nations
Quality of National IQ Data
| Data Quality Tier | Description | Example Countries | Reliability |
|---|---|---|---|
| Tier 1: Excellent | Multiple large-scale standardized test administrations; national norms available | United States, United Kingdom, Germany, Japan, South Korea | High |
| Tier 2: Good | At least one large representative sample or strong PISA/TIMSS data | France, Australia, Poland, Brazil, Turkey | Moderate-High |
| Tier 3: Limited | Small or non-representative samples; heavy reliance on PISA conversion | Mexico, Iran, Indonesia, Egypt | Moderate |
| Tier 4: Estimated | No direct IQ data; estimated from neighboring countries or minimal samples | Many Sub-Saharan African, Central Asian nations | Low |
| Tier 5: Imputed | No cognitive data whatsoever; purely extrapolated | Some small island nations, conflict zones | Very Low |
"A substantial portion of the national IQ database consists of estimates rather than measurements. This distinction is often lost when the numbers are published in tables." -- Jelte Wicherts, psychologist, Tilburg University
National IQ Averages by Region: 2025 Data
The following tables present estimated national IQ averages based on the most recent compilations, including David Becker's NIQ-Dataset (v1.3.4, updated 2024) and cross-referenced with PISA 2022 results where available. All scores are normed to a British mean of 100 (SD = 15).
Important disclaimer: These figures are estimates with varying degrees of uncertainty. They should be interpreted as rough indicators, not precise measurements.
East Asia and Pacific
| Country | Estimated IQ | Data Quality | PISA 2022 Math Score | Primary Data Source |
|---|---|---|---|---|
| Singapore | 105-108 | Tier 1 | 575 | National norms, PISA |
| Japan | 104-107 | Tier 1 | 536 | WISC-IV norms, PISA |
| South Korea | 104-106 | Tier 1 | 527 | K-WISC norms, PISA |
| Taiwan | 104-106 | Tier 2 | 547 | National studies, PISA |
| China (select regions) | 103-106 | Tier 2 | 591 (B-S-J-Z) | PISA, regional studies |
| Hong Kong | 105-108 | Tier 1 | 540 | PISA, local norms |
| Australia | 99-100 | Tier 1 | 487 | WAIS norms, PISA |
| New Zealand | 99-100 | Tier 1 | 479 | WAIS norms, PISA |
| Vietnam | 94-99 | Tier 3 | 469 | Limited studies, PISA |
Europe
| Country | Estimated IQ | Data Quality | PISA 2022 Math Score | Primary Data Source |
|---|---|---|---|---|
| Netherlands | 100-102 | Tier 1 | 493 | WAIS norms |
| Germany | 99-102 | Tier 1 | 475 | WAIS norms, PISA |
| Switzerland | 100-102 | Tier 1 | 508 | PISA, national studies |
| United Kingdom | 100 | Tier 1 (reference) | 489 | WAIS norms (baseline) |
| Finland | 99-101 | Tier 1 | 484 | Conscription data, PISA |
| Estonia | 99-101 | Tier 1 | 510 | PISA, national studies |
| Poland | 97-99 | Tier 1 | 489 | National norms, PISA |
| France | 98-100 | Tier 1 | 474 | WAIS norms, PISA |
| Italy | 97-99 | Tier 1 | 471 | National norms, PISA |
| Spain | 96-98 | Tier 1 | 473 | National norms, PISA |
| Greece | 93-95 | Tier 2 | 430 | National studies, PISA |
| Romania | 91-94 | Tier 2 | 428 | Limited norms, PISA |
North America
| Country | Estimated IQ | Data Quality | PISA 2022 Math Score | Primary Data Source |
|---|---|---|---|---|
| Canada | 99-101 | Tier 1 | 497 | WAIS norms, PISA |
| United States | 97-100 | Tier 1 | 465 | WAIS norms, PISA |
| Mexico | 87-90 | Tier 2 | 395 | Limited norms, PISA |
South America
| Country | Estimated IQ | Data Quality | PISA 2022 Math Score | Primary Data Source |
|---|---|---|---|---|
| Chile | 90-93 | Tier 2 | 412 | National studies, PISA |
| Uruguay | 90-93 | Tier 2 | 409 | PISA |
| Argentina | 87-93 | Tier 2 | 378 | Limited norms, PISA |
| Brazil | 85-90 | Tier 2 | 379 | National studies, PISA |
| Colombia | 84-89 | Tier 2 | 383 | PISA |
| Peru | 84-88 | Tier 2 | 391 | PISA |
Middle East and North Africa
| Country | Estimated IQ | Data Quality | PISA 2022 Math Score | Primary Data Source |
|---|---|---|---|---|
| Israel | 95-97 | Tier 1 | 458 | Conscription data, PISA |
| Turkey | 90-93 | Tier 2 | 453 | PISA, limited norms |
| United Arab Emirates | 86-89 | Tier 2 | 431 | PISA |
| Saudi Arabia | 82-86 | Tier 3 | 389 | PISA, limited data |
| Morocco | 78-84 | Tier 3 | 365 | PISA, limited data |
Sub-Saharan Africa
| Country | Estimated IQ | Data Quality | Notes |
|---|---|---|---|
| South Africa | 77-82 | Tier 3 | Highly variable by region and SES; no PISA participation |
| Kenya | 75-82 | Tier 3 | Limited samples, urban-rural divide |
| Nigeria | 71-78 | Tier 4 | Very limited data; large regional variation |
| Ghana | 70-78 | Tier 4 | Small sample studies |
| Tanzania | 72-78 | Tier 4 | Very limited data |
Critical note on Sub-Saharan African data: The quality of IQ data for many African nations is extremely poor. Most estimates derive from small, non-representative samples, often tested in non-native languages using culturally inappropriate instruments. Wicherts, Dolan, and van der Maas (2010) re-analyzed the Lynn/Vanhanen dataset and found that when systematic errors were corrected, Sub-Saharan African IQ estimates rose by approximately 5-10 points.
The Flynn Effect: How National IQ Averages Change Over Time
The Flynn Effect -- named after James Flynn, who documented it systematically in the 1980s -- refers to the substantial and consistent rise in IQ scores observed across generations in virtually every country studied.
"The gains are massive -- amounting to as much as 20 points over a single generation. They demonstrate that IQ tests are not measuring some fixed, immutable capacity." -- James Flynn (1987)
Flynn Effect Magnitude by Country
| Country | Period | Total IQ Gain | Points Per Decade | Primary Test Used |
|---|---|---|---|---|
| Netherlands | 1952-1982 | +21 points | +7.0 | Raven's Progressive Matrices |
| Denmark | 1958-1998 | +15 points | +3.8 | Military conscription test (BPP) |
| Norway | 1954-2002 | +13.5 points | +2.8 | Military conscription test |
| United States | 1932-1978 | +13.8 points | +3.0 | Stanford-Binet, WAIS |
| Japan | 1951-1975 | +12 points | +5.0 | WISC adaptations |
| South Korea | 1989-2002 | +7.7 points | +5.9 | K-WISC, Raven's |
| Israel | 1972-2004 | +11 points | +3.4 | Military conscription (Otis) |
| Kenya | 1984-1998 | +11.7 points | +8.4 | Raven's CPM |
| Brazil | 1930s-2000s | ~15 points | ~2.5 | Various |
| Estonia | 1935-2006 | +14 points | +2.0 | Raven's |
The Negative Flynn Effect
In a significant recent development, several developed nations have reported declining IQ scores since the 1990s:
| Country | Period of Decline | Decline Rate | Data Source |
|---|---|---|---|
| Norway | 1995-2009 | -0.33 points/year | Bratsberg & Rogeberg (2018) |
| Denmark | 1998-2014 | -0.27 points/year | Dutton & Lynn (2013) |
| Finland | 1997-2009 | -0.25 points/year | Dutton et al. (2016) |
| France | 1999-2009 | -0.48 points/year | Dutton & Lynn (2015) |
Proposed explanations for the negative Flynn Effect include:
- Dysgenic fertility -- the hypothesis that lower-IQ individuals have more children (controversial and empirically weak)
- Immigration effects -- changes in population composition (contested)
- Education changes -- shifting pedagogical approaches
- Technology and media effects -- reduced deep reading, attention fragmentation
- Ceiling effects -- environmental improvements reaching diminishing returns in developed nations
"The Flynn Effect reversal in Scandinavia is one of the most important findings in intelligence research in the past two decades. It demands explanation." -- Bernt Bratsberg, economist, Ragnar Frisch Centre for Economic Research
Importantly, Bratsberg and Rogeberg (2018) showed that the decline in Norway occurs within families -- younger siblings score lower than older siblings from the same parents -- ruling out dysgenic fertility and immigration as primary causes.
Why Cross-National IQ Comparisons Are Problematic
While the data tables above attract enormous attention, researchers have identified serious methodological limitations that make cross-national IQ comparisons far less reliable than country-internal data.
Key Limitations
| Limitation | Description | Impact on Data |
|---|---|---|
| Sampling bias | Many national estimates derive from small, urban, or convenience samples | May overestimate or underestimate by 5-10+ points |
| Test translation issues | IQ tests developed in Western contexts may not translate validly | Items may change in difficulty or meaning |
| Cultural relevance | Test content may favor Western cognitive styles (abstract reasoning, timed performance) | Systematically disadvantages populations with different cognitive emphases |
| Testing conditions | Motivation, familiarity with testing, physical conditions vary enormously | Lowers scores in populations unfamiliar with formal testing |
| Norming differences | Different tests use different norms and standardization years | Scores not directly comparable across tests |
| Socioeconomic confounds | Poverty, malnutrition, disease burden, educational access all lower IQ scores | Conflates environmental deprivation with cognitive capacity |
| Publication bias | Studies with extreme or dramatic results more likely to be published | May distort the available dataset |
The Wicherts Critique
A pivotal 2010 study by Wicherts, Dolan, and van der Maas systematically evaluated the data quality of African IQ studies used in Lynn and Vanhanen's dataset. They found:
- Many studies used non-representative samples (e.g., only urban children, only students from specific schools)
- Some studies tested participants in non-native languages
- Several studies included participants with known health or nutritional deficiencies
- When these problematic studies were excluded, the African IQ mean estimate rose from approximately 67 to 78-82
"Many of the lowest national IQ estimates are artifacts of poor sampling, inappropriate tests, and adverse testing conditions -- not reflections of a population's cognitive potential." -- Jelte Wicherts, Tilburg University
What Explains National IQ Differences? Environmental Factors
Researchers have identified multiple environmental factors that explain a substantial portion of cross-national IQ variation.
Environmental Predictors of National IQ
| Environmental Factor | Correlation with National IQ | Mechanism | Key Evidence |
|---|---|---|---|
| GDP per capita | r = 0.50-0.70 | Resources for education, nutrition, healthcare | Consistently found across datasets |
| Years of schooling | r = 0.60-0.75 | Direct cognitive stimulation and skill development | Each year of schooling adds ~1-5 IQ points (Ritchie & Tucker-Drob, 2018) |
| Childhood nutrition | r = 0.40-0.60 | Brain development requires adequate protein, micronutrients | Iodine supplementation alone can raise IQ by 10-15 points in deficient populations |
| Infectious disease burden | r = -0.60-0.80 | Parasitic infections divert metabolic resources from brain development | Eppig, Fincher, & Thornhill (2010) found disease burden was the strongest predictor |
| Urbanization | r = 0.40-0.50 | Exposure to complex environments, formal education access | Urban residents consistently score higher than rural |
| Lead exposure | r = -0.30-0.50 | Neurotoxic effects on developing brain | Estimated 2-5 point IQ reduction per 10 ug/dL blood lead |
The Iodine Deficiency Example
Perhaps the most dramatic demonstration of environmental effects on national IQ comes from iodine deficiency. Iodine is essential for thyroid function, which regulates brain development. In populations with severe iodine deficiency:
- Average IQ reductions of 10-15 points have been documented
- Universal salt iodization programs have been associated with IQ gains of similar magnitude
- A natural experiment in Switzerland showed that iodine supplementation in the 1920s coincided with a dramatic rise in IQ scores and a decline in goiter-related intellectual disability
"Iodine deficiency is the world's greatest single cause of preventable brain damage." -- World Health Organization
This example illustrates a crucial point: national IQ averages reflect environmental conditions at least as much as they reflect cognitive potential.
How to Interpret National IQ Data Responsibly
Given the complexities outlined above, how should an informed reader approach national IQ data?
Guidelines for Responsible Interpretation
- Treat numbers as rough estimates, not precise measurements. A difference of 2-3 points between countries is likely within the margin of error.
- Always check data quality. Tier 1 data (multiple large-scale standardized administrations) is far more reliable than Tier 4-5 estimates.
- Consider environmental context. Low national IQ averages in developing countries primarily reflect poverty, disease, malnutrition, and educational deprivation -- not genetic inferiority.
- Never apply group averages to individuals. Within every country, IQ scores span a vast range. A national average of 85 does not mean every person in that country has an IQ of 85.
- Watch for the Flynn Effect. Many estimates are based on data from 10-30+ years ago. Given the Flynn Effect (especially rapid in developing countries), current averages may be substantially higher.
- Be skeptical of rankings. Ranking nations by IQ implies a precision that the data does not support and invites misleading conclusions.
If you are interested in assessing your own cognitive abilities rather than comparing nations, our full IQ test provides an individual evaluation that measures working memory, processing speed, and reasoning. For a faster overview, try our quick IQ test.
National IQ and Economic Development: Correlation, Not Causation
One of the most frequently cited relationships is the correlation between national IQ averages and economic development (GDP per capita). This correlation is real (r = 0.50-0.70 depending on the dataset), but its interpretation is fiercely debated.
Competing Causal Interpretations
| Interpretation | Proponent(s) | Argument |
|---|---|---|
| IQ drives economic development | Lynn & Vanhanen (2002) | Higher cognitive ability enables innovation, productivity, and institutional quality |
| Economic development drives IQ | Flynn (2007), Wicherts et al. | Wealth provides education, nutrition, and healthcare that raise IQ |
| Bidirectional causation | Rindermann (2018) | IQ and wealth mutually reinforce each other in a feedback loop |
| Third variables | Multiple researchers | Institutions, geography, colonialism, and trade routes drive both IQ and wealth |
"The idea that national wealth is caused by national IQ puts the cart before the horse. The correlation is real, but the causation likely runs primarily from wealth to IQ, not the other way around." -- James Flynn
The strongest evidence for the environmental direction comes from natural experiments:
- German reunification (1990): East and West Germans share the same genetic stock but diverged in IQ by approximately 7 points after 40 years of different socioeconomic systems. The gap narrowed rapidly after reunification.
- The Flynn Effect itself: Countries experiencing rapid economic development (South Korea, Ireland, Brazil) show correspondingly rapid IQ gains.
Conclusion: Data with Context, Not Data in Isolation
National IQ averages provide a useful but imperfect lens for understanding global cognitive patterns. The data clearly show that environmental factors -- education, nutrition, healthcare, economic development -- powerfully shape measured intelligence. Where these conditions improve, IQ scores rise. Where they deteriorate, scores decline.
The responsible use of this data requires acknowledging its limitations: sampling problems, cultural biases in testing, and the enormous within-country variation that no average can capture. National IQ data should inform efforts to improve conditions for cognitive development worldwide, not to rank or stigmatize populations.
"The most important thing about national IQ data is not the numbers themselves, but what they reveal about the conditions -- educational, nutritional, economic -- that either support or hinder cognitive development." -- Richard Nisbett, psychologist, University of Michigan
For those interested in their own cognitive profile -- a far more reliable and personally meaningful measure than any national average -- our full IQ test provides a standardized individual assessment. You can also practice with our practice test or take a timed IQ test to explore different cognitive domains.
References
- Lynn, R., & Vanhanen, T. (2002). IQ and the Wealth of Nations. Praeger.
- Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101(2), 171-191.
- Wicherts, J. M., Dolan, C. V., & van der Maas, H. L. J. (2010). A systematic literature review of the average IQ of sub-Saharan Africans. Intelligence, 38(1), 1-20.
- Bratsberg, B., & Rogeberg, O. (2018). Flynn effect and its reversal are both environmentally caused. Proceedings of the National Academy of Sciences, 115(26), 6674-6678.
- Eppig, C., Fincher, C. L., & Thornhill, R. (2010). Parasite prevalence and the worldwide distribution of cognitive ability. Proceedings of the Royal Society B, 277(1701), 3801-3808.
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358-1369.
- Rindermann, H. (2018). Cognitive Capitalism: Human Capital and the Wellbeing of Nations. Cambridge University Press.
- Becker, D. (2019). The NIQ-Dataset (v1.3.4). https://viewoniq.org/
- OECD. (2023). PISA 2022 Results (Volume I): The State of Learning and Equity in Education. OECD Publishing.
- Flynn, J. R. (2007). What Is Intelligence? Beyond the Flynn Effect. Cambridge University Press.
- Dutton, E., van der Linden, D., & Lynn, R. (2016). The negative Flynn Effect: A systematic literature review. Intelligence, 59, 163-169.
Frequently Asked Questions
What country has the highest average IQ?
Based on the most recent data compilations, ***East Asian nations*** -- particularly Singapore, Japan, South Korea, Hong Kong, and Taiwan -- consistently show the highest estimated national IQ averages, typically ranging from 104 to 108. However, these estimates carry uncertainty of 2-3 points, and much of the data reflects PISA-equivalent conversions rather than direct IQ test administrations. China's data is particularly limited, as PISA results come only from select, high-performing regions (Beijing, Shanghai, Jiangsu, Zhejiang). It is also important to remember that these averages reflect current environmental conditions (excellent education systems, nutrition, healthcare) as much as any other factor.
How much can a country's IQ change in a generation?
Substantially. The Flynn Effect documents gains of ***3-7 points per decade*** in many countries during the 20th century. The Netherlands gained 21 points between 1952 and 1982. South Korea gained nearly 8 points in just 13 years (1989-2002). Kenya showed gains of 8.4 points per decade between 1984 and 1998. These changes are far too rapid to be genetic and reflect improvements in education, nutrition, and healthcare. Conversely, some developed nations (Norway, Denmark, France) have shown IQ ***declines*** of 2-5 points since the mid-1990s. Test your own cognitive abilities with our [full IQ test](/en/full-iq-test).
Are national IQ comparisons scientifically valid?
They are ***partially valid but deeply limited***. Data quality varies enormously across countries (from Tier 1 multiple-study estimates to Tier 5 pure extrapolations). Cultural biases in test design, sampling problems, and socioeconomic confounds all reduce reliability. The Wicherts et al. (2010) reanalysis showed that correcting for poor data quality raised Sub-Saharan African estimates by 5-10 points. Most intelligence researchers use national IQ data cautiously and acknowledge its limitations. The data is most useful for tracking ***changes within a country*** over time (where measurement conditions are more comparable) rather than for cross-national rankings.
What is the Flynn Effect and why is it reversing in some countries?
The Flynn Effect is the observation that IQ scores have risen by approximately ***3 points per decade*** across dozens of countries throughout the 20th century. It was documented by James Flynn in the 1980s. The effect is strongest on fluid reasoning (culture-fair) tests and is attributed to environmental improvements: better nutrition, more education, increased cognitive complexity of modern life. The reversal -- documented in Norway, Denmark, Finland, and France since the 1990s -- is called the "negative Flynn Effect." Bratsberg and Rogeberg (2018) showed the decline occurs within families (younger siblings score lower than older siblings), ruling out genetic explanations and pointing to environmental causes such as changing educational practices or media habits.
How does education affect national IQ averages?
Education is one of the ***strongest environmental predictors*** of national IQ (r = 0.60-0.75). A meta-analysis by Ritchie and Tucker-Drob (2018) found that each additional year of schooling raises IQ by approximately ***1 to 5 points***, with the strongest effects in early childhood. Countries that have expanded universal education have consistently seen rising IQ scores. The mechanism is straightforward: schooling trains precisely the abstract reasoning, working memory, and problem-solving skills that IQ tests measure. This is why investment in education is considered the most reliable pathway to raising national cognitive performance. Explore your own cognitive strengths through our [practice test](/en/practice-iq-test).
Why are Sub-Saharan African IQ estimates considered unreliable?
Multiple methodological problems undermine Sub-Saharan African IQ estimates: (1) ***sample sizes are often very small*** (sometimes under 100 participants); (2) samples are frequently ***non-representative*** (only urban students, only one ethnic group); (3) tests are often administered in ***non-native languages***; (4) participants may have ***nutritional deficiencies or parasite infections*** that lower scores; (5) ***testing conditions*** (noisy environments, unfamiliarity with test format) differ dramatically from standardized conditions. Wicherts et al. (2010) demonstrated that when the most problematic studies were removed, average estimates rose significantly. Researchers increasingly argue that these estimates tell us more about ***deprivation and testing conditions*** than about cognitive potential.
Related Articles
Curious about your IQ?
You can take a free online IQ test and get instant results.
Take IQ Test