Thursday, 14 May 2015

Marginal tribes, disparate outcomes

There are two genetic/cultural groups who constitute minorities in Europe. For various reasons they were kept at the margins of society, made liable to restrictions in terms of the trades they could follow, and often hounded and maltreated. In the 20th Century both were slaughtered by German National Socialists.

One of those tribes is well known and highly accomplished, the other less well studied, and with far fewer accomplishments. The contrast is striking. However, explaining that difference must wait for another day. At this point we need to get a better understanding of the abilities of the less well studied tribe, particularly since studies on their abilities are often not published in English, the de facto language of Google science.

I have covered the topic before,

but the London Conference paper brings things more up to date, and carries out a formal meta-analysis.



Dalibor Jurasek has found 38 usable datasets from 7 countries, resulting in a total N = 4468. He concludes that the best estimate of Roma intelligence is IQ 74. This is extremely low. There does not seem to be any verbal bias in the tests; there is weak or no Flynn effect; Spearman’s rule seems to hold.

Below you will find his Powerpoint presentation and, in a move which I hope many speakers will follow, his speaker notes, which give further particulars.

Now we can start working on why the Roma and the Ashkenazi differ in their accomplishments.


  1. Two things instantly seem to be relevant:

    Godwin’s Law | West Hunter

  2. The closest relatives of European gypsies are the dalit tribes of north-west India. So 1000 years ago roma tribes started their journey through Europe. Their original cognitive capacity was not to high, and because of the caste system of India and the endogamy in Europe it may fixed. Shocking to see the same slums in India and parallel in Europe at the end of the towns and villages inhabited by dalits and gypsies where they show exactly the same behavioural features and lifestyle.
    The marginalised lifestyle and the use of the surplus and waste resources of the host population somehow exhibited a selection pressure toward not for IQ, but rather for immunity against diseases and for r-strategy. So the indicated 74 IQ for the gypsies may be correct. The low vocabulary of the languages of the gypsy tribes (e.g. lovari) also confirm this hypothesis.
    In Hungary Jozsef Vekerdi linguistic professor published a roma-english –hungarian dictionary in 2000. He wrote: „The dictionary comprises ca. 2600 items, including Hungarian loans and obsolete words. An individual speaker uses about 900-1200 words in every day conversations, not counting the Hungarian loans.”

  3. Strongly suggests genetic continuity and minimal impact of environmental variables over about 32 generations.

  4. I wonder there is no study from Hungary in this meta-analysis, despite the fact that huge gypsy community living here (10%). I bet the Hungarian gypsies have a much higher IQ than the indicated 74. Fortunately we can calculate the approximate IQ for Hungarian gypsies taking account the high number of roma pupils in remedial schools. For this special school, IQ of 70 and below need to be measured for a given child. Annually as high as 20-22% of the roma children versus
    3% of the Magyar pupils enter to this type of education.
    Taking these data I measured hypothetical average IQ for Hungarian roma by a simple normal distribution calculator (21% below IQ 70, SD 15) and I got 82 IQ. The 1-2% number of roma with higher education also verify this (IQ for HE 110, SD15)
    The higher integration status of gypsies and better SES in Hungary may cause this higher IQ estimate.

    1. Thank you for these calculations. The discussion on the paper raised the possibility that we have another example of "boiling off": people who do not like the lifestyle leave, often for something better, and become assimilated to some degree. Hungary may have encouraged this process, as you suggest.

  5. Sorry, I could not see any speaker notes on the slides. Were they included with them?

  6. Should be at the very bottom of the slides, and not all the slides have notes

  7. Prolific speaker peoples tend to score lower in iq, 70-80 on average.


  8. I wonder how this, combined with the reported fact that 80% of children in Romanian orphanages are Roma, bears upon the findings of longitudinal studies of children adopted from those orphanages?

  9. It's very interesting that the integrated urban Roma have normal IQs. Is this a case of higher IQ types leaving the Roma culture and assimilating or does the normal Roma environment depress IQs that much? Or a combination of the two factors. Is the 98 a robust finding?

  10. Probably due to "boiling off" (see above). Those who have ability leave the group, pass for locals. Remainder become more "pure" and deepen their identity. I think it is clear that the environment is self made.

  11. Probably due to "boiling off" (see above). Those who have ability leave the group, pass for locals. Remainder become more "pure" and deepen their identity. I think it is clear that the environment is self made.

  12. I'm fan of artistic gymnastics and i have curiosity about integrated roma people among romanians. Many non-officially recognized gypsy romanians look mixed with gypsy (seems) but they are treated as ethnically romanians and not as gypsy or mixed-race. Romania have enormous gypsy population.
    Many famous romanian gymnasts look octoroon, quadron... gypsies and seems have normal intelligence, completely.