Fighting for Equity in Education

The struggle is long but hope is longer

Kevin Donnelly - The "Cory Bernadi" of Education Policy

Monday January 20, 2014

Kevin Donnelly has been a tireless campaigner against the Gonski report, using a variety of spurious arguments. But a recent opinion piece in The Australian breaks new ground, because he has explicitly linked educational inequalities in Australia to genetic differences in cognitive ability.

Here, in Donnelly’s own words, is the argument.

The article starts by stating that “58% of the variation in (British) General Certificate of Secondary Education scores ……. is explained by students’ genetic make-up.” He continues to argue that “ innate cognitive ability” is highly influential when explaining why some students outperform others. He contrasts this view to the tenet of the “progressive and cultural-left education establishment” that “success or failure are determined by socioeconomic background associated with being working class, from a non-English speaking background or indigenous.”

He then turns to the highly controversial American academic Charles Murray (author of The Bell Curve ), quoting Murray to the effect that “the changes we can expect in academic achievement in the lower half of the ability distribution are marginal, no matter what educational reforms are introduced.” This was part of Murray’s critique of attempts in the US to improve educational standards in disadvantaged groups.

Donnelly concedes that “home and school environment play an important role,” but then adds that “a student’s innate cognitive ability is highly influential when explaining why some students outperform others.”

From here, he slips into a discussion of work which has been used to suggest that “cognitive ability has a considerably stronger influence than socioeconomic background on educational outcomes in many different contexts.” But note that the crucial word “ innate ” has disappeared.

Donnelly has dressed his views up as science, but the evidence he cites is highly controversial. Unpicking the poor science that underpins the idea that differences between social groups in cognitive ability as measured by IQ are significantly genetic in origin is therefore important, but quite complex.

The evidence for Donnelly’s first claim, that 58% of the variation in GCSE scores is genetic in origin, comes from a study of outcomes for identical twins (MZ- monozygotic twins) compared to those of non-identical twins (DZ – dizygotic twins).

Unfortunately, twin studies have two fundamental flaws. Firstly, they assume that parents treat members of DZ twin pairs as similarly as members of MZ twin pairs. Only on the basis of this assumption, can it be assumed that any tendency for MZ twins to be more similar than DZ twins can be attributed to the greater genetic similarity of MZ twins. The validity of this assumption has not been established, which means that no firm conclusions can be drawn from the study.

The second problem is that twin studies can often overestimate genetic effects, because the environmental differences within twin pairs, and particularly within MZ twin pairs, are inevitably very small. Once again, this means that firm conclusions cannot be drawn.

These two limitations of twin studies have been clear on theoretical grounds for many years, but they have been largely ignored in practice. However, the proof of the pudding is in the eating, and with increasingly sophisticated techniques for genetic analysis, it has increasingly been reported that there is much less defined genetic variation associated with variation in a complex trait such as cognitive ability, than would expected from the twin study heritability. This is now so common that the field has come to have its own name – missing heritability.

In contrast to the claim that genetic variations in ability account for 58% of the variation in GCSE scores, it has so far been found that identified genetic differences account for less than 1% of the variation in cognitive ability.

There are possible genetic explanations of this gap – one being that many genetic variations are involved, with each having such small effects that they have not yet been detected, despite the large scale of the current studies. Another equally plausible explanation is that, given the assumptions on which they are based, twin studies may not be a good guide to the extent of genetic contributions to variation in complex traits – in other words, the missing heritability was never really there in the first place.

None of this means that the idea that cognitive ability has a substantial genetic component has been refuted, but it does mean that assuming that cognitive ability is innate is not based on established scientific fact, but is more an act of faith.

The second strand of Donnelly’s argument comes from Charles Murray’s The Bell Curve. A core part of the argument in this book is that variations in IQ within the white population in the US are mainly genetic in origin – a conclusion largely based on the sorts of flawed analysis of twin studies outlined above.

It is further claimed that the quite large difference in mean IQ between whites and African-Americans in the US is probably, at least in part, due to genetic differences in ability. This argument is also theoretically flawed, because there is no reason to assume that the causes of between-group variation are the same as those within-groups. Racial discrimination, lower standards of living and lower opportunities for education experienced by African-Americans could explain the lower mean IQ scores for African-Americans, without invoking genetic differences. While Murray sometimes hedges his bets on genes and environment, how he interprets the data is clear, given his warning, quoted above, that only marginal effects can be expected for those in the lower half of the ability distribution.

But, again, the proof of the pudding is in the eating. The US 2013 Mathematics and Reading Assessments show that over the period from 1990 to 2013, the percentage of students in the below basic category has dropped from 50% to 17% in Grade 4 maths, from 48% to 26% in Grade 8 maths, from 38-32% in Grade 4 reading and from 31-22% in Grade 8 reading. What is more, bigger gains over this period were achieved by African-Americans than whites. This means marked improvements in outcomes at the bottom of the distribution, in an ethnic minority which Murray, and apparently Donnelly, assume has on average genetically lower cognitive ability. This evidence alone leaves Murray’s (and Donnelly’s) genetically racist and socially elitist predictions in tatters.

Donnelly then engages in a sleight of hand. He outlines the evidence that cognitive ability is an important factor in determining educational and indeed life outcomes, and then, as noted above, slips seamlessly into the assumption that innate cognitive ability is crucial.

It is important to note that in most of the research that he cites, ability is defined as earlier performance in school. He never seriously stops to ask whether the same environmental and social factors that are associated with final educational outcomes are also important in determining early performance. If this is the case, then the analysis would lead to an underestimation of the importance of socioeconomic factors.

In contrast, it seems highly plausible that disadvantaged backgrounds exert their effects early, and that students from these background start off behind, and tend to stay behind. From this perspective, what schools need is the funding, staffing, and programs, to make a difference. But in Donnelly’s perspective, it is more about genetically determined cognitive ability.

The rest of the article consists of straw man arguments, in which Donnelly attributes particularly naive views to supporters of Gonski, and then demolishes them. So, he suggests that the “education establishment” believes that ability has nothing to do educational outcomes, and that “all that is needed to overcome disadvantage is to invest additional billions.”

But no-one has argued for that. He cites several studies which have shown that increased spending has not always led to improved outcomes, and simply ignores the many studies that show that funding increases for evidence-based programs can produce effects. Far from believing that money alone makes the difference, the Gonski report and most of its supporters argue that it is important to define what works, and direct the funding to these programs.

Donnelly is never explicit about how much of the current educational inequality in Australia could be due to innate differences in cognitive ability, although he does lead off with the figure of 58%. But it is clear that he believes that some of it, and perhaps a lot of it is. But at the same time he argues that this view “is not to suggest that inheritance is destiny, or that nothing can be done to lift standards and strengthen the performance of many under-achieving students.” But if this is true, then there might be a place for Gonski funding after all.

It is therefore imperative that Donnelly spell out how much he believes genetic differences in innate cognitive ability set limits to achieving the Gonski aims of eliminating the effects of socio-demographic factors on educational outcomes. Putting it very specifically for the Australian situation, Mr Donnelly needs to answer two specific questions:
• How much of the lower outcomes currently achieved by Indigenous children does he believe is genetic in origin?
• How much of the lower outcomes currently achieved by children from low SES backgrounds is genetic in origin?

Donnelly’s claim that genetic differences are important in relation to educational inequality in Australia is due to genetic difference is not established scientifically. Whatever the final outcome of the research on genetic contributions to individual differences in cognitive ability, and currently the identified genetic associations are very small, there is no evidence at all that differences in educational performance between social classes and ethnic groups are genetic in origin.

So, at the moment, Donnelly’s arguments look like just another in the long history of attempts to use bad science to justify the continuation of disadvantage and social inequality.

Professor Ian Morgan
Visiting Fellow at the Australian National University, Canberra, and Sun Yat-sen University, Guangzhou, China.

Commenting is closed for this article.

Previous Next


Contact us

News feed from Save our Schools

SOS Twitter

Chris Bonnor's Education Media Watch



Education news

Education blogs

Education information

Public education

Policy Briefs