Tuesday 11 November 2014

Global IQ and health

Until relatively recently, intelligence and health were seen as two unrelated matters. This is not surprising, because it is usual to think in compartments, and often very sensible. By compartmentalising thoughts one avoids them spilling over, and making a total mess of disorganised observations. A compartment makes things manageable, keeps things clean and tidy, and brings them within the scope of our limited minds. The downside of such cognitive cleanliness is well known: a link can be overlooked because it lies across a conceptual boundary. Like good generals, reality attacks best when it finds the boundary between two opposing armies: that particular point in the enemy line is weak because each army sees it as of peripheral interest, and any intrusion can probably be blamed on the neighbouring army group. We need liaison officers to bridge the gaps between disciplines if we are to go forward in good order.

For most of my academic life I taught about intelligence in terms of scholastic attainment, and the individual assessment of brain damage, but linked it to little else. Health behaviours were a totally different matter, dealt with by a completely different set of researchers. So, to my shame, I taught about patient compliance with doctor’s instructions and reported the findings of the compliance literature that it did not seem to be related to intelligence. I taught about Flesch readability formulas without considering their g loadings. This was wrong, but it took intelligence researchers to show the omission years later. Gottfredson’s work showed how measures of this sort were distributed about a bell curve, and could be plotted against other intellectual measures. The breakthrough in health linkage came with the Lothian Birth Cohort work, showing that intelligence tested at age 11 was a good predictor of health and lifespan measured 60 years later.

Does this link apply globally?

First, a few preliminary observations. If we can show a link within a particular population, measured by variables (lifespan, problem-solving ability) which can be applied to all populations,  then it is likely (but not proven) that it will be found in all populations. If so, we can look within each population to see if the lifespan/intelligence link holds true. It is worth testing.

Second, one way to do this is to plot each nation in intelligence/lifespan space and see if there is a relationship. Tatu Vanhanen has done this in his book “Global Inequality as a consequence of human diversity” Ulster Institute for Social Research, London, 2014. ISBN 978-0-9573913-7-6

Here is a snapshot of Fig 4.4 on page 104.




Vanhanen is interested in testing the explanations for wealth given by those who favour external forces (geography, climate, colonialism, governance) rather than the internal force of human differences, which he sees as a more probable cause. He reviews the very broad canvas of environmental explanations for national differences, and also looks at their many criticisms of the causal nature of differences in ability and character. Most economic theorists dispute the effects of intelligence on society, and would not agree that intelligence may be involved in health differences. They argue that the wealth of a society, and how it is organised and governed, are the key drivers of health. They usually leave unexamined the view that intelligence may be a big part of the reason why a society is wealthy and well-organised. Indeed, their approach is one of universal silence on this possibility: differences in ability are never mentioned.

Back to Fig 4.4. The life expectancy measure is given as years above 40 years of age, so as to show fine detail. Across 178 countries the correlations with IQ are as follows: Global Inequality .86; Life expectancy .82; Under 5 mortality .80; Tertiary education .79; Sanitation .73; Income .65; Democratization .56.

So, intelligence (at a national level) correlates with life expectancy (also at a national level) at .82 which is 67% of the variance. Correlation is not causation, but the correlation of intelligence with global inequality makes sense: there is a bell curve of ability and a bell curve of wealth. Correlation, as we have discussed before, is necessary to show a causal link, but not sufficient. The causal chain from intelligence to health is easy to argue, link by link, vaccine by vaccine, public water supply by public water supply.

The correlation between intelligence and lifespan has a simple explanation: clever societies manage threats more expeditiously. They become safer. However, that simple explanation may need to be qualified. Perhaps societies with brighter people live longer because they have environmental and also genetic advantages. This finding has been shown within the Lothian cohort. Early intelligence is a great predictor. Simple reaction time measured in later life is a good predictor, and attenuates the intelligence effect. Perhaps the common factor is “system integrity”, a factor which give the lucky recipients a good brain and a good body.  More of that later.


  1. Well the one thing that the normalisation of life expectancy values doesn't convey is that the range of values at the bottom is fairly restricted. Many poor/low-IQ countries have higher life expectancy than you would expect given their economic & socio-political conditions, thanks to the global secular rise in life expectancy since 1950. I suppose that must be in large part because medical technology, especially immunisation, developed in the higher-IQ countries, is fairly cheap and widespread. But there are other cheap life-extending measures which don't require too much governance ability to implement, like sanitation or birth practises.

    Vanhanen wonders why Cuba is such an outlier (page 134). Well-trodden territory :

    Cuba & developing countries, health, education, political ability, and a little bit on cognitive ability

  2. Thanks, Very informative and helpful.The range of life expectenties at low income is an eye opener. Good points on Cuba (and Kerala). Interesting how many countries try to fake international test results.

  3. The amount of aid (both medicinal and the provision of food, etc) flowing from the developed to the developing nations in Africa has been enormous in the post WWII period. That must have affected the correlations between IQ and life expectancy to some degree, but I don't know how one could correct for that.

  4. I thought life expectancies were mostly driven by infant mortality rates?

  5. Yes, primarily. There's less variation in life expectancy at 5, for example.

    1. I am ignorant of western government and NGO policies towards lowering infant mortality rates in developing countries. But I can't believe that their impact is zero.