The Myth of Nigerian Excellence
Why do some people believe Nigerians are brilliant high-achievers?
Many people believe the following facts:
The Igbo are an exceptionally intelligent African group, performing on par with or above Whites1
Nigerian Americans are as intelligent as White Americans
Nigerian Americans have higher incomes than White Americans
Nigerian Americans are more educated than White Americans
Nigerian Britons are as intelligent as White Britons
Nigerian Britons have higher incomes than White Britons
Nigerian Britons are more educated than White Britons
Let us examine these facts, point-by-point.2
The National IQ of Nigeria is 68 per Richard Lynn and David Becker. Assuming I have the correct manual for conversion, the Mill Hill Vocabulary Scale test data for Nigerians suggest a national IQ of 73 relative to British norms in this sample and the Raven’s Standard Progressive Matrices Plus test data in this sample suggest a national IQ of about 71. Both of these are not corrected for the Flynn effect, so they’re presumably overestimates relative to the British samples, but they’re also not corrected for bias, so we’ll call it a wash.
The problem with these samples for figuring out the intelligence of the Igbo is that they’re derived from Lagos State and the Abuja Federal Capital Territory. If we want to figure out how the Igbo perform specifically, we need data from places like Abia, Imo, Ebonyi, Enugu, and Anambra—the cultural region known as “Igboland”. Since there is not Igbo-specific test data out there to my knowledge, it’s unclear how we should estimate the intelligence of the Igbo as a whole in a reliable manner.
One suggestion has been to use games performance, like performance in chess, videogames, and—for Nigeria in particular—scrabble. But scrabble is not a particularly attractive game for cognitive elites to play, nor is it very cognitively intensive. In fact, it overwhelmingly relies on memory, one of the least g-loaded abilities, and the one that most often trivializes tests when retesting or preparation for a particular form come into play. Consistent with this, scrabble performance was found to have a weak relationship with measured national ability, unlike CS:GO, chess, poker, and StarCraft II performance. That study also found—perhaps due to bias in assessments—that in nations with measured IQs below 85, games performance was unrelated to measured IQs.
I am not aware of any good reason to think the Igbo are particularly brilliant, nor do I know any reason to think they’re more or less bright than Nigerians in general.3
The factoid “the Igbo are uniquely brilliant” lacks empirical support.
Nigerians in America
Thanks to the American Community Survey4, we have large-scale, detailed, and recent (2022) data on individuals of Nigerian descent in the U.S. We also have several comparison groups to better understand how well Nigerians are doing.
For the purposes of my analyses, I will be examining educational attainment and total personal income5 adjusted for age, sex6, and census region and division. I will be primarily doing this among native Whites and Blacks, foreign-born Whites and Blacks, native and foreign-born Nigerians, native and foreign-born Indians, and South Africa-born Whites and Blacks. For educational attainment I believe the most sensible effect size is Hedge’s g. For income, the most sensible statistic is the median, since it is robust to outliers.
First up, let’s look at their incomes:
Next up, let’s look at their educational attainment relative to native Whites:
Since the latest available precomputed occupational status score was the Nam-Powers-Boyd score with a basis in 1990, results for occupational attainment would probably not be reliable. However, for those interested in the “OCCSCORE” differences (1950s basis), see this footnote.7
This data shows quite a lot.
With respect to education, we can see that immigrants are generally more educated than native Whites. This makes complete sense, as the immigration system selects for education. By the same token, we can see major regression to the mean in terms of education, with American-born Indians coming in below their parents, and American-born Nigerians coming in far beneath their parents.
With respect to income, we can see that, in order, Indians perform incredibly well, South African Whites outperform foreign Whites in general, Nigerians from Nigeria outperform native Whites, native Whites outperform African Blacks and foreign Blacks in general, as well as native-born Nigerians, who beat out South African Blacks, who beat out Native Blacks.
So, what can we say about claims about Nigerian American socioeconomic status?
With respect to education, Nigerian immigrants are, indeed, highly-educated. This is expected from other datasets too. For example, in the IAB Brain Drain dataset, as of 2010, some 0.6% of Nigerians had emigrated to the West, including 12% of highly-educated Nigerians. However, despite their considerable advantages in educational attainment, Nigerians have much lower incomes than other groups they’re more educated than, like South African Whites, native Indians, and foreign Whites. This makes it very clear that there are exceptions in the link between more meritocratic measures of socioeconomic status like income and more potentially non-cognitive ones like educational attainment. Because socioeconomic status measures can become disconnected like this, they create what looks like an attainment-achievement discrepancy for many groups, including Nigerians.
Regression to the mean for Nigerians is rather extreme. The ones born in the U.S. clock in well below Whites, in both education and incomes. This makes sense, however, as it seems their parents were exceptionally educated, so we should expect this sort of regression. It also makes plenty of sense that this regression is greater than the regression for Indians, since it involves regressing to different means.
Nigerians earn more than African Blacks and foreign Blacks in general, but they’re also not the highest-performing African group. That would be South Africans. But that’s misleading, because the South African group is peopled by Whites who command higher incomes than any of the Black African groups, while the Black South Africans perform worse than Nigerians. Nevertheless, Nigerians are not the highest-performing Black African group. Zimbabwean Blacks out-earned them ($48,793), and Cameroonians were plausibly tied with them at $45,352. The rest performed worse than them, which we might predict because Nigeria is the largest Sub-Saharan African nation by far. Due to their size, this suggests Nigerians are pretty much at the top of the pile for America’s Sub-Saharan African Black immigrants.
We have samples of Whites born in a few Sub-Saharan African countries. White South Africans, White Zimbabweans (Rhodesians), and White Nigerians had sample sizes that declined in that order, but they nevertheless all out-earned Black Nigerians, with respective median incomes of $59,578, $98,185, and $75,434. If we compare foreign-born Nigerians to even more groups of foreign-born Whites, we would quickly see that there are very many who out-earn them. There’s good reason to think about sample selectiveness when it comes to immigrants.
I judge the claim that Nigerians are more educated and higher-earning than White Americans is half-true.
Nigerians born in Nigeria out-earned and were more educated than native-born Whites, but native-born Nigerians earned less than and were less educated than native-born Whites, indicating major regression to the mean. Moreover, the foreign-born Nigerian advantage was also seen among foreign-born Whites, albeit to a lesser extent. Adjusting for this immigrant advantage across Whites and Blacks, Nigerians earn less than Whites and have less education.
Using ethnic self-identification alone would reduce this apparent regression. This was also documented by Emeka (2019) in a study on ethnic attrition among Nigerian Americans. Emeka noted that the individuals who ceased to identify as Nigerian were disproportionately the downwardly mobile, “leaving us with a positively select Nigerian second generation and, perhaps, unduly optimistic assessments of Nigerian-American socioeconomic advancement.”8
What about intelligence?
We only have one measurement of intelligence among specifically Nigerian Americans, and it’s a proxy measurement based on computing the representation of National Merit semifinalists identified as Nigerian through their surnames.9 If one believes that Nigerians are very motivated towards education, then this method overestimates. With known patterns of ethnic attrition and incomplete surname identification, it almost certainly provides overestimates of ability, but there’s no obvious correction factor for this method. When Nigerian Americans’ representation is contrasted with Whites, the gap is -0.03 SDs, or an IQ of 100.5, consistent with no difference. The median gap estimate for Africans in general is 0.39 standard deviations in favor of Whites, or an IQ of 94.2. Because this method should overestimate Nigerian and African IQs, it’s a bit odd.
That estimate generates further confusion because we also have estimates of the IQ of non-English Language Learner (ELL; i.e., likely second-generation) Somali Americans. These have been estimated at 91.32. At the same time, the native Somali income is $21,627 and the foreign-born Somali income is $25,824. A few potential explanations spring to mind. Firstly, the Somalis are likely not as selective as Nigerians on other dimensions, like educational attainment. This seems like it should obviously be true given that Somali Americans are largely refugee descendants. Perhaps the reason for this discrepancy is that the IQ estimate comes from Washington, which ought to be a somewhat more selective destination than the more typical one in Minneapolis. More likely, however, is that Somalis integrate more poorly, as higher proportions of non-immigrant Somali youths are English language learners than are the non-immigrant members of most other groups. Therefore, the non-ELL estimate for Somali Americans is probably an overestimate of their population average.
For measuring Nigerian American intelligence we’re at a loss. However, we do have estimates of the intelligence of African and Caribbean immigrants in the U.S. (and Canada), and for those, the results are clear-cut.
From the 1999 National Longitudinal Survey of Freshmen, native Whites had an estimated SAT/ACT-based IQ of 106.8, compared to 108.5 for foreign-born Whites. Blacks from Sub-Saharan Africa had an estimated IQ of 94.2 versus 92.8 for Blacks from the Caribbean and 90.5 for native-born Blacks. In other words, the native-born Black-White gap was typical (16.3 points), and the gaps between native-born Whites and Sub-Saharan and Caribbean Blacks were 12.6 and 14 points.
Richwine (2009) computed immigrant-native gaps for Europeans, northeast Asians, southeast Asians, Indians, Sub-Saharan Africans, Mexicans, Central Americans/Caribbeans, and South Americans. Relative to an American mean of 100, these groups had (culture-fair!) digit span-based IQs of 101.2, 107.8, 106.3, 113.8, 91, 84.4, 84.7, and 88.3. The same dataset also had second-generation scores for all of Asia, Mexico, and Central America/the Caribbean. These were consistent with IQs of 106.9, 84.1, and 91.9.
In the complete General Social Survey, Sub-Saharan Africans and Caribbeans scored 1.20 and 1.10 g below native-born Whites, which compares to a roughly 0.7 g gap between native-born Whites and Blacks. This might reflect bias, since the test is the wordsum. However, the gap for Sub-Saharan Africans was larger in the second generation and for Caribbeans, it was the same, although sample sizes were small in both cases.
Across several surveys, the gap between native Whites and Black immigrants is considerable. For first-generation Black immigrants, it averaged 0.99 SDs, and for the second generation, 0.84 SDs, but for the third, 1.00 SDs, albeit with identification problems afoot.
In neighboring Canada, African Blacks, Caribbean Blacks, and Canadian Blacks had estimated IQs of 91, 87, and 88.8 based on the average of 7ᵗʰ- and 8ᵗʰ-grade reading and mathematics test results. 9ᵗʰ-grade results showed IQs of 90.1 across the board, and 10ᵗʰ-grade results showed IQ estimates of 84.9, 80.2, and 89.1. In later data, Black Canadians, Caribbeans, and Africans had means of 95.2, 90, and 93.3. For the same tests, Somali language speakers scored 89.1 in 7ᵗʰ and 8ᵗʰ grade, and 87.1 in 9ᵗʰ and 10ᵗʰ grade. In later data, Somali language speakers scored 90.6 in 7ᵗʰ and 8ᵗʰ grade and 89.5 in 10ᵗʰ grade. Canadian LSAT data from 2019 showed a Black mean of 88.2 and a 2023 Black mean of 86.7. But more importantly, it showed a 10% larger standard deviation for Whites, so if the Canadian sample was split between lower-scoring Blacks from the U.S. and Caribbean, and higher-scoring African Blacks, they should have a larger variance if each of those groups has the same variance as Whites. Of course, we don’t know what each group’s variances are, so this isn’t conclusive.
The factoid “Nigerian immigrants to the U.S. are as bright as White Americans” is, at best, minimally supported. The evidence from other assessments of African Blacks suggests it is unlikely.
Nigerians in Britain
The situation with respect to data in the U.S. is relatively good. Basically everything in Britain is worse, but we do have some data.