Saturday, April 25, 2009

Perceptions of average intelligence by race

++Addition3++Blode gives a detailed breakdown of group self-perceptions, perceptions of 'outside' groups, and actual IQ averages in the comments. Rather than try to summarize, I recommend taking a look for yourself.

++Addition2++Steve Sailer wonders about Jewish perceptions of intelligence. Sample sizes for takes on Hispanics, Asians, and Jews are prohibitively small (N = 29), but with that disclaimer in mind, it's still an interesting question. Consequently, I've added a table for Jewish perceptions of the five groups considered and added a bar set for Jews in the graph presented in the body of the post.

++Addition++FeministX comments:
Average people seem to have almost perfectly identified intelligence as a reflection of income by race.

It seems people subconsciously know how much ethnic groups earn on average, except for the jews.
When Herrnstein and Murray published The Bell Curve in 1994, the racial 'hierarchy' for income followed the intelligence hierarchy. The massive influx of unskilled immigrants from Latin America over the last couple of decades (with first-generation Mexicans averaging an eighth grade level of educational attainment) has since pushed Hispanic earnings below that of blacks in the US. That working blacks earn more than working Hispanics do (and due to remittances, many Hispanics in the US appear even poorer than they are) likely plays into the higher intelligence ascribed to blacks compared to Hispanics.


Continuing with the interaction between self-described and actual intelligence, the GSS asks respondents to rate groups by intelligence on a scale from 1 (unintelligent) to 7 (intelligent). To gauge contemporary perceptions, the following table shows the mean value for all responses from the beginning of this decade on. The third column converts answers to IQ scores under the assumptions that the mean intelligence value for whites represents the equivalent of an IQ of 100 and that a standard deviation is 15 points:

Public perception of...MeanIQ

People push group averages in both directions toward the mean, but with the exception of rating the average intelligence of blacks higher than Hispanics (probably owing in part to the fact that there are far more black than Hispanic celebrities with household name recognition and also to the educational system's trumping up of black thinkers and inventors like Du Bois and George Washington Carver without any corresponding Hispanic intellectual giants), the American public sees through the blank slatist mythology. Groups are not identical, and it requires ignoring what you see with your own lying eyes to believe that they are.

Across all racial categories, people essentially see the group intelligence 'hierarchy' in the same way. The results, by race of responders*:

White perception of...MeanIQ

Black perception of...MeanIQ

Hispanic perception of...MeanIQ

Asian perception of...MeanIQ

Jewish perception of...MeanIQ

For ease of digestion, the following graph shows how intelligent each racial group perceives itself and the other racial groups to be. The white bars show how intelligent whites see each group (shown along the y-axis) to be, the black bars show black perceptions of each group, the brown bars show Hispanic perceptions, the yellow bars show Asian perceptions, and the baby blue bars show Jewish perceptions.

There is a caveat to the trend mentioned above. With the exception of self-deprecating whites, each group sees itself as more intelligent relative to other groups than the population as a whole sees it. That is, blacks see blacks as more intelligent than anybody else does. The same holds true for Hispanics, who see Hispanics as more intelligent than anybody else does, and Asians, who see Asians as more intelligent than all but Hispanics do (and Asian perceptions are depressed relative to the population across the board, so their high self-perceptions stand out).


* Because the question concerning Hispanic, Asian, and Jewish intelligence was only asked in 2000, Hispanic, Asian, and Jewish sample sizes for these three groups is small (34, 43, and 29, respectively). Consequently, they should only be seen as suggestive.

Notice that the Hispanic sample is more generous across the board than other racial groups are. This could be random optimism in the small Hispanic sample. Comparitive, rather than absolute, values assigned to a group by members of each racial group are what is most worthy of consideration.

Hispanic also includes "some other race", as it is a method of racial identification used almost exclusively (97% of the time) by Hispanics. The Asian category includes Indian, Chinese, Japanese, Filipino, Korean, and "other Asian".


ironrailsironweights said...

It looks as if everyone gave low estimated for Jewish IQ. Why would that be?


Nick said...

I suspect Jewish IQ is underrated because Jewish dominance in various fields is a lot easier to obscure with egalitarian dogma, as they're relatively uncommon and not always easily identifiable. You'd have to unbelievably unaware not to notice the disproportionate academic failure and criminality of Blacks and Hispanic, while you have to be somewhat informed to spot the powerful and successful Jews in the US. (Not very informed, of course, but most people know very little.) Still, it does seem unnaturally low, even in spite of the more difficult observation.

Blode0322 said...

Not to nitpick, but it looks like Hispanics think Asians are smarter than Asians think they are. Also I think the post title is misleading.

Some observations: It it completely unsurprising that whites think white intelligence is less than everybody else thinks it is. Lots of whites cultivate the image of the illiterate bible-thumping rural white, partly as a result of MSM conditioning.

Quite surprising that most groups assessed black intelligence ahead of Hispanic. I suppose your reasoning is pretty sound. MSM conditioning may play a role here ... there are countless black computer nerds in all shows, plus lots of black female judges, etc., without many corresponding Hispanic figures.

My wife just asked which group has the most accurate view overall ... and I told her I can't do the math. (No spreadsheet on this charming little laptop.)

Audacious Epigone said...

Will respond with something more worthy later, but I just have a moment now. Re: the title, I got it mixed up with another post I'm working on! Sorry.

rob said...

The low hispanic estimate may be a result of less than fluent English. No matter why, it's pretty natural to assume that someone who can't communicate with you is dim.

Darwin's Sh*tlist said...

I'd be interested to see how this plays out within different IQ groups of the same ethnicity.

With blacks, for example, my impression is that the brighter ones have the most inflated view of themselves. I've always attributed this to affirmative action: many of them have had amazing opportunities opened to them for only being the most competitive within their own ethnic group. Of course, since the etiquette of affirmative action prevents anyone from pointing this out, the biggest AA beneficiaries would be most likely to have distorted views.

I wonder to what extent this holds true for the middle and left of the black bell curve, where AA is less of a factor.

Audacious Epigone said...


The relatively low Jewish IQ estimate doesn't really surprise me. I think Nick hits it. The Steveosphere is not intellectually representative of our larger culture, not by a long shot. What percentage of the population is even vaguely familiar with Cochran and Harpending's Ashkenazi intelligence theory? Ten percent is probably a generous estimate. What percentage of the population is aware that Bernie Madoff is Jewish? Secular Jews are not conspicuous to Joe Sixpack like blacks, Hispanics, or Asians are, and there are some relatively high profile Hasidic Jewish communities that seem pretty backwards.


It's not optimally worded, but the point I was trying to make is shown by the following:

Black IQ (all) - 94.0
Black IQ (blacks) - 98.5

Hispanic IQ (all) - 91.9
Hispanic IQ (Hisps) - 96.8

Asian IQ (all) - 100.8
Asian IQ (Asians) - 102.1

White IQ (all) - 100.0
White IQ (whites) - 99.6


Good point. That Hispanics are taciturn compared to blacks probably also has some affect.


Thanks for giving me more ideas to work with.

ironrailsironweights said...

Audacious -

I must respectfully disagree, the Jews = high I.Q. stereotype is pretty much a universal one, at least in America. What might account for the survey's relatively low estimates of Jewish I.Q. is the concern that acknowledging that Jews are smart might be perceived as vaguely anti-Semitic.

It's not my impression that the underachieving Hasidic communities are particularly well-known outside the New York area. Maybe in a few other cities, but it's not a nationwide thing.



Nick said...

I don't see that the stereotype is universal. I see you live in New York, Peter. Well, I've lived in the Midwest, Southwest and Mountain West my entire life, and Jews are sufficiently rare in these areas that I don't believe most people think of them at all, and would be confused that anyone would see Jews as at all different from other whites. Moreover, to the extent that they're aware of Jewish stereotypes they probably assume that the stereotypes are false. (False egalitarian dogma is a lot easier to believe when you don't have the truth staring you in the face round the clock.) On a personal level, I wasn't aware of the gap until I'd really started paying attention to politics and HBD. It just didn't come up. Ever.

I remember Sailer referencing a column by irritating columnist Joel Stein that claimed that now only 22% of Americans believe that the entertainment industry is run by Jews, as opposed to 50% in 1964, despite this being more or less true. I've no doubt there used to be widespread beliefs concerning Jews, but that they're receding because egalitarian dogma. Again, Jews are not easily identifiable and most Americans do not deal with large number of Jews in their day to day life, so they just believe what they're told.

FeministX said...

Average people seem to have almost perfectly identified intelligence as a reflection of income by race-

It seems people subconsciously know how much ethnic groups earn on average, except for the jews.

Audacious Epigone said...


Nick beat me to the punch.

Still, Jews are considered more intelligent than any of the four major racial/ethnic groups in the US. But does the average person have any idea that roughly 1 in 40 (Ashkenazi) Jews are "geniuses", while only 1 in 250 white Gentiles are? I find that doubtful (at least outside the Northeast)--instead, Jews are 'only' perceived as intelligent minorities.


Great point. That working blacks earn more than working Hispanics do (and due to remittances, many Hispanics in the US probably appear even poorer than they are) likely plays into the higher intelligence ascribed to blacks than to Hispanics.

Anonymous said...

i believe you have a typo -- the asian estimates of asian & jewish intelligence seem to be inverted.

Steve Sailer said...

Hispanics appear to have done the best job of getting the rank order right, perhaps because they are less exposed to English-language media propaganda.

Steve Sailer said...

Is there a big enough sample size to see how Jews rate Jews?

Audacious Epigone said...


The sample sizes are even smaller than for Hispanics and Asians, but Jewish perceptions are pretty similar to white perceptions. Like whites, Jews do not inflate their own group.

Dave said...

A very very interesting post which addresses an issue which has always been given a lot of discussion in the past. Thanks.

al fin said...

On your graph, it would be much clearer at first glance to group perspectives of whites toward others together, perspectives of blacks toward others together, etc. rather than distributing them among 5 groupings.

The colour bars are self-explanatory (except for blue Jews, which does rhyme ;0).

Audacious Epigone said...


Perhaps, but I think it's easier to compare how various races view each group this way.

I couldn't think of any obvious marker for Jews. The light blue is tied to the Israeli flag.

J said...

An other fun fact: Asians are most likely to recognize diversity.

If you take a group and then subtract their bottom perception from their top perception you get a diversity score.

W - 10.6

B - 9

H - 10.2

A - 12.3

J - 9

Audacious Epigone said...


Suggesting Asians are the least beholden to the blank slatism. Keep in mind, though, the Asian sample is small.

Blode0322 said...

My wife just asked which group has the most accurate view overall ... and I told her I can't do the math. (No spreadsheet on this charming little laptop.) - me

Okay, I've done some more numbers on which group's views are most accurate. I compared the numbers supplied by A. Epigone to what (I hope) are the real numbers. Caveat: they are from Wikipedia: whites 101.4, black 86.9, Hispanics 91, Asians 106, Jews 112 (the bottom of a supplied range, but higher than Lynn's estimate).

Doing so I generated absolute errors which I could then average. Let me apologize for not knowing the real term for non-absolute error (i.e., error with the + or - sign left on). I generated these numbers too, which I call "vectored errors" in contrast to the absolute errors.

Info generated from absolute errors:

Of political orientations, conservatives have the most accurate views (4.3) and liberals the least (5.6).

Of the ethnic groups, blacks have the least accurate views of relative group IQs (average absolute error: 6.1).

Asians have the most accurate views (4.5).

The second-most accurate group views are from whites (4.8), but the smartest group of all, Jews, have the second least accurate views (5.8).

Blode0322 said...

Info generated from vectored errors:

Jews are the most underestimated group (-9.9 ... average views shortchange Jews by almost 10 IQ points!) Asians view Jews the most harshly (-12.1); Hispanics the least harshly (-7.5).

Blacks are the most overestimated group (+7.3). Again, Asians have the least rosy views of this minority (overestimating their IQ by only 4.3); blacks have the most (+11.6).

Hispanics and whites are the most accurately-perceived groups (errors of 1.4, with whites shortchanged and Hispanics viewed as smarter than they really are)

Hispanics perceive aggregate IQs the most optimistically (+0.6 ... the only group to think we're collectively smarter than we actually are).

Asians perceive aggregate IQs the least optimistically (-2.8).

Blode0322 said...

Addendum to "vectored errors", above:

Moderates are the least accurate and optimistic in overall intelligence assessment (-2.0). Conservatives are the most accurate and optimistic, only underestimating us all by 1.2 IQ points.

Interesting points:

Every group, political or ethnic, overestimates black IQ. This is in the face of evidence that IQ tests overpredict black performance (i.e., the "real number" I am using above predicts lower dropout and crime rates than occur in reality). Mainstream opinion in every group is in at least some denial about black IQ deficits.

Hispanics are the only group to have correctly conceived Hispanics as smarter than blacks, though they underestimate the difference. Most groups do not think Hispanics are dimmer than they actually are, they just overestimate them less than they do blacks.

Asians are the only group to have incorrectly perceived Asians as smarter than Jews. Jews return the favor by having the harshest views of Asian IQ (-6.1), but they are even harder on themselves (-10.7).

The ranking of group IQ, and the ranking of the accuracy of that group's views of other groups' IQs (say that three times fast), is the same, with one glaring exception: the smartest group, Jews, have quite inaccurate views on IQ. This tends to confirm the default hypothesis among HBDcons that it takes brains to properly assess IQ, but political correctness can throw all the calculations out.

Final note: If the Wikipedia-supplied "real numbers" are bad, it will be a piece of cake for me to change those and generate a whole new set of analyses. Let me know if you have better IQ estimates for the groups.

Blode0322 said...

Can't stop the rock:

Liberals, moderates, and conservatives perceive black intelligence in similar ways. Politics causes us to really differ in our assessments of white and Hispanic IQ, with liberals and moderates differing by a full three points on Hispanic IQ (+2.8 versus -0.2, the latter being one of only two cases of NAM IQ being underestimated). Liberals rob whites of almost three IQ points, while moderates steal only one, and conservatives less.

Jews have an even lower view of whites IQ than liberals, but Jews are hard on everyone. Asians have the most accurate view of white IQ, though they are harsher in general than Jews.

On the question of self-assessment:

Actual white IQs beat generally-perceived white IQs, which beat self-perceived white IQs.

Self-perceived black IQs beat generally-perceived black IQs, which beat actual black IQs.

Self-perceived Hispanic IQs beat generally-perceived Hispanic IQs, which beat actual Hispanic IQs.

Actual Jewish IQs beat generally-perceived Jewish IQs, which beat self-perceived Jewish IQs.

Actual Asian IQs beat self-perceived Asian IQs, which beat generally-perceived Asian IQs. Thus they are the only group who self-perception lies between conventional wisdom and reality, i.e. Asians are the only group whose self-concept does not multiply a widely-held error.

Audacious Epigone said...


Wow, nice work. I pointed it out in the body of the post.

One cautionary point: I based the scale on the presumption that the average for whites from all groups is the equivalent of an IQ of 100, so perceptions of whites are artificially going to look pretty accurate.

Blode0322 said...

"Nice work"? I done petted the thread so hard I killed it. Good thing it wasn't a puppy.

Audacious Epigone said...

No, you just gave it a spectacular finale. Thanks!

Anonymous said...

Do those ones count as Blacks? Were they treated differently?

There are more than a billion reasons why you should take any attempts at statistifying humans with a kilo of Sodium Chloride.