In part 1 of this post I bemoaned the limited research on the economics of G&T education.
One consequence is that, when it comes to defending gifted education against recession-driven spending cuts, there are very few weapons in the arsenal. This is a major problem and urgent action is needed to rectify it.
In reviewing the research I could find – and searching for the ‘magic bullet’ – I was struck by the untapped potential of ‘smart fraction theory’. This is highly controversial and contested territory, but it would certainly repay further exploration.
We need to backtrack a little in order to explain how ‘smart fraction’ thinking has developed.
The impact of cognitive ability on economic growth
In 2008, Hanushek and Woessmann published a comprehensive and influential paper on The Role of Cognitive Skills in Economic Development.
Their core argument might be summarised as follows:
- Empirical studies show that extending the quantity of schooling does not necessarily bring about improvements in human capital that impact positively on economic growth
- Schooling is too narrow an indicator; there are other factors at play, both environmental and genetic. So it is better to assess improvements in cognitive skill, as demonstrated by performance on international comparative tests such as PISA and TIMSS
- Cognitive skills have a significant positive impact on personal earnings and, in macroeconomic terms, on economic growth, whether in developed or developing countries
- Analysis of PISA and TIMSS maths, science and reading scores suggests that, were countries performing at the OECD mean of 500 points to improve to 540 points – the level of top performers like Finland and South Korea – they would secure a 5% improvement in GDP over a 20 year period
- The model suggests that improvements in top end performance and in average performance have separate and complementary effects on economic growth.
This last point is revisited in a 2009 follow up study .
This finds that that, were countries to secure an improvement of 10% in the proportion of students scoring at 400+ points and 600+ points respectively, each would have a positive impact on economic growth – and the intervention at the top end of the ability range would have 4 times greater impact than the intervention at average ability levels.
Unfortunately though, the statistical analysis shows that a 10% improvement at 600+ points would be much harder to achieve than the commensurate improvement at 400+ points.
However, support for top level performers is likely to have relatively stronger impact in developing countries compared with developed countries.
The overall conclusion is that providing better basic education for all and also pushing significant numbers to very high achievement levels is the best policy for economic growth.
The Smart Fraction
The concept of the smart fraction was first introduced by the pseudonymous La Griffe du Lion, who may or may not be the American academic Robert Gordon.
In 2002 he published a Socratic dialogue called ‘The Smart Fraction Theory of IQ and the Wealth of Nations’.
It explores the relationship between national average IQ and per capita GDP, finding a strong positive correlation. The author suggests that:
- contributions to national productivity require a minimum level of cognitive ability
- this is around IQ108 , about the level needed for a first degree
- the smart fraction is the percentage of the population which record at least this IQ.
The economic and statistical argument behind this analysis appears slightly frail. Moreover, the fraction is pitched at a level which makes it barely relevant to apologists for G&T education.
Worse still, the smart fraction has since been appropriated by those who would wish to prove racial differences and even to justify eugenics. This territory is a cul de sac for G&T education and we should give it a very wide berth indeed.
The SMART fraction reinvented
The missing link between these two studies was provided in 2009 when Rindemann, Sailer and Thompson published ‘The Impact of Smart Fractions, Cognitive Ability of Politicians and Average Competence of Peoples on Social Development.
The authors are psychologists rather than economists and it will be interesting to see whether their work stands up to scrutiny from the latter.
Their key argument and findings follow:
- There are strong links between findings in the economic tradition (human capital) the educational tradition (literacy) and the psychological tradition (intelligence) suggesting they are measuring ‘the same underlying latent factor’ of cognitive ability
- The smart fraction should be pitched similarly to Hanushek and Woessmann’s higher level, defined here as the 95th percentile on TIMSS, PISA and PIRLS tests of comparative academic performance, which is said to be equivalent to IQ125
- It is also possible to identify a ‘non-smart fraction’ comprising those performing at the 5th percentile (IQ 70-75)
- Proxies are established for high national cognitive achievement – the average annual number of patents per million people, the number of Nobel Prizes relative to population size, the number of scientists and engineers employed in R&D per million people and the proportion of high technology exports as a percentage of all manufacturing exports
- There is a much stronger correlation between high national GDP and the smart fraction than high national GDP and average cognitive ability
- There is also a strong correlation with most of the indicators of national cognitive achievement (less so for Nobel Prizes in literature) and with the identified indicators of political sophistication.
- The positive impact on GDP can be isolated mainly to STEM-related achievement asopposed to achievement outside the STEM fields, suggesting the formers are the main drivers of national affluence
- As might be expected, the negative impact of large class sizes is more pronounced on the ‘non-smart fraction’, whereas the positive impact of early tracking (setting) is stronger for the smart fraction
- Further research is required that examines different ability levels and uses other indicators of high achievement
Their ultimate conclusion: ‘our results emphasise the importance of nurturing the highly gifted’.
What can we learn from this?
To date there is scant evidence online that the findings of Rindermann et al are attracting peer review.
I would give much to know whether Hanushek and Woessmann find this treatment convincing. The methodology and logic do need to be tested rigorously, especially the vital issue of causation.
There is a pressing need for a leading economist of education to endorse or qualify Rindermann’s conclusions.
That might spark further collaboration between economists and experts in G&T education, doing much to stiffen the backbone of gifted education research, as well as providing a much needed justification for continued investment during this period of financial retrenchment.