While data from various IQ tests are useful for global scale analyses (e.g. GDPcc correlations), they are far less reliable for particular countries. That’s why I’m a big fan of the OECD’s PISA assessments, which are highly standardized, have large samples from similar age groups, take place concurrently once every three years, test those aspects of intelligence most intuitively relevant to economic success (i.e. application of numeracy and literacy skills in novel situations), and enjoy strong face validity (i.e. very few “strange” results).
However, there is also the temporal dimension. All these IQ maps that you see today are almost entirely based on testing children/teenagers from the current year to 1948, the earliest year on David Becker’s database as of now (although with attempts to correct them through reference to contemporary UK standardization samples). PISA and TIMSS are rather tidier, with large, representative samples of teenagers getting tested at set ages and at set years. Still, even this isn’t perfect, because countries vary in their educational and auxological histories, which will have varying knock-on effects on the intelligence of different cohorts.
The correlations between cohorts will still be very good (after all, IQ is strongly hereditary, and the quality of the environment will itself tend to be strongly correlated to average IQ). But there will be some interesting outliers, both positive and negative. For instance, the gap between the youngest and oldest cohorts can be expected to be greater in countries such as South Korea, which transitioned from the Third World to the First in the space of half a century. It can likewise be expected to be smaller in countries like the United Kingdom, which sprang off from a high base – it started the last century as the workshop of the world, and was less damaged by WW2 than most other European countries – but plummeted in relative terms ever since. Understanding cohort dynamics will also make it possible to do, say, more fine-grained analyses between national IQ and socio-economic success.
So it’s a bit surprising that hardly any attention has been devoted to another OECD program, PIAAC (Programme for the International Assessment of Adult Competencies), which tests a range of cohorts instead of just teenagers.
The first round of the assessment in 2012 covered the following 22 countries: Australia, Austria, Canada, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Ireland, Italy, Japan, Korea, the Netherlands, Norway, Poland, the Slovak Republic, Spain, Sweden the United States; Chile, Greece, Indonesia, Israel, Lithuania, New Zealand, Singapore, Slovenia and Turkey joining in 2014.
The following documents/dataset refers refers to the first round of PIAAC:
- OECD Skills Outlook 2013: First Results from the Survey of Adult Skills (pdf)
- Table A3.3 (N) [xls, tables]
- Figure 3.8a (L) [xls, charts]
Unfortunately, I have been unable to find data for the later countries in one place, though I haven’t looked very hard. It appears that the next round of PIAAC will take place in 2020.
In the rest of this post, I will highlight some of the more interesting data from there. For comprehensibility, all numbers have been converted to the IQ scale, with the England/N. Ireland average set to its traditional “Greenwich Mean” of 100 and S.D. = 15.
PIAAC 2012: Literacy, Numeracy, Average IQ
|England/N. Ireland (UK)||100.0||100.0||100.0|
No surprises here, except perhaps Korea’s figures being a bit lower than expected. We’ll come to that.
Difference in Performance: Youngest (16-24) vs. Oldest (55-65)
|England/N. Ireland (UK)||0.2||0.0||0.1|
As hypothesized, Korea has the largest cohort Flynn effect; in age-adjusted terms, while its young perform as well as the best (Japan, Netherlands, Finland), its elderly are near the back of the queue.
Incidentally, Turkey – not covered in the first round of this assessment, but I found figures for it in its national report from the second round – has a difference of 13.8 IQ points between its oldest and youngest cohorts. This makes patent sense in the context of it going from a Third World country in the 1950s, to an upper middle-income one today.
In contrast, the UK and the US – already relatively well developed countries in the 1950s, when their boomers appeared – barely eked out any increase in the ensuing fifty years.
The big exception here is Russia. As I speculated in my mega-article on Russian IQ for Sputnik and Pogrom, this may have been linked to the alcohol epidemic that began in Russia from around the mid-1960s, when life expectancy plateaued and consequently stagnated for the next half-century. It is not a big stretch to imagine there were similar dynamics in the country’s psychometric profile, with any Flynn effects from continuing development being annulled by the flood of vodka of the late Soviet era.
On the bright side, Russia’s alcohol epidemic has more or less ended, and – as I predicted back in 2012 – IQ amongst the youngest cohorts has been going up for the past decade (e.g. from 95 in PISA 2000-2009 to 99 by PISA 2015; Sugonyev’s yet unpublished military data).
Sociological observation of questionable validity: The three laggards here, the US, Russia, and the UK, are all especially (in)famous for developing a sizable lumpenproletariat class during this period (respectively, white/trailer trash, gopniks, and chavs).
Difference in Performance: Men vs. Women
|England/N. Ireland (UK)||0.8||4.0||2.4|
The Germanic lands – or perhaps countries characterized by the authoritarian family model – are characterized by significantly brighter males, while the Latin and Slavic lands lean in the other direction.
In Russia, it seems women are brighter, not only on average, but even in terms of numerical skills. Possibly this is also a function of Russian men having borne the brunt of the 1965-2015 alcohol epidemic.
Difference in Performance: Natives vs. Immigrants
|England/N. Ireland (UK)||10.3||11.8||11.0|
No surprises here, I think. Seems to correlate with the PISA data (see “Not Sending Their Best”: World Map of IQ Drop Due to Immigration).
Difference in Performance: Skilled vs. Elementary Occupations
|England/N. Ireland (UK)||7.8||9.5||8.6|
To maximize output and social welfare, you want to cluster your brighter people in the skilled, complex jobs of the “O-Ring economy“, while simple, “foolproof” jobs can be done by pretty much anybody reasonably effectively. It doesn’t matter if a waiter is 145 IQ or 100 IQ, but it certainly does if he’s a CEO.
As Murray and Herrnstein showed in The Bell Curve, the effectiveness of this cognitive sorting mechanism has increased by leaps and bounds in the US since the 1950s.
The PIAAC data suggests Anglos and Scandinavians are the best at this, which might be one more factor that explains their unusual economic and scientific dynamism, even relative to their IQs.
The Japanese do worse at this, possibly being held back by cultural factors (deference to age and other elements of social status).
Russia is at the very bottom of the list, suggesting highly inefficient cognitive selection. The pessimistic but plausible explanation that comes to mind is endemic corruption and nepotism.
1. In particular, we need a more serious, in-depth analysis of the PIAAC data. Conversion of disparate tables floating about in PDF and Excel format on the Internet into one database.
2. In general, greater focus on tracking IQ across all relevant dimensions (nations; subregions; time; cohorts) to enable deeper, finer-grained economic/demographic analysis.