The headlines for Scottish education from the latest round of the Programme for International Student Assessment (PISA) are dismaying. In all three subject areas covered by PISA, the scores of Scottish 15-year-olds declined between 2018 and 2022. The drop was 18 points in mathematics, 11 points in reading, and 7 points in science. A change of 20 points is approximately equivalent to one year of mid-secondary schooling. So these falls correspond to nearly a year in mathematics, over six months in reading, and a term in science.
The explanation might seem to be the disruption caused by Covid. Attainment fell in most countries, and so Scotland’s problems are certainly not unique. But that is nowhere near to being the whole reason for decline. Doubts are raised about the Covid explanation by Andreas Schleicher, director of the PISA studies, in a detailed analysis of the global data (available on the PISA web pages). Not all countries did show a fall. For example, attainment rose in Japan and Korea in all three domains, and rose in two of the three in several others, for example Singapore, Italy and Israel. In fact, there was only a weak overall relationship between countries’ change in score and the extent of their school closures during the Covid period.
Of most relevance to the Scottish explanation, however, is Schleicher’s further point – that attainment was declining long before Covid, and that Covid merely gave it an extra downward push. That is certainly true of Scotland, as is well known. Scottish attainment fell from early in the century to the middle of the first decade, stabilised for a few years, and then, from 2012, started a steady decline which was unmitigated except for a brief rise in reading in 2018 (which was wiped out by the 2022 fall).
As a result, over the whole decade from 2012 to 2022, the Scottish decline was equivalent to about 16 months of schooling in mathematics, 8 months in reading, and 18 months in science. The baseline of 2012 is significant because it is the first PISA group to have any experience of the Curriculum for Excellence after it was officially inaugurated in 2010. Thus the decline started to become noticeable at the moment when the new curriculum started to impinge systematically on children’s learning. The 2022 group was the first to have all 10 years of their schooling with the new curriculum, and attainment has never been so low as it is now.
That the loss over the decade was greatest for science, next for mathematics, and least for reading is consistent with one of the main criticisms of the new curriculum, that it neglects knowledge of the kind that students can obtain only from expert teachers. To some extent reading can be taught by parents, but fewer parents are likely to remember enough mathematics to teach their children beyond the end-of-primary stage. Science knowledge is even rarer, and in any case learning science needs access to specialist equipment.
This interpretation of the impact of the curriculum on students’ knowledge is reinforced by two further aspects of the 2022 data, both relating to inequality of attainment. The first is the difference between students at the 10th and 90th percentiles of attainment in each domain. Table 1 shows the score at each of these points between 2012 and 2022. In reading there was barely any fall at the 90th percentile between 2018 and 2022, and in fact a rise of 9 points from 2012. At the 10th percentile, by contrast, there was a fall of 22 points from 2018 and 33 from 2012.
A similar contrast is evident for mathematics and science. In mathematics, the fall from 2012 to 2022 was 16 at the 90th percentile but 36 at the 10th. In science, the corresponding falls were 13 and 47. The contrast between the two percentiles is thus, again, greatest for science, next for mathematics, and lowest for reading. For all three domains, an analogous but weaker form of this contrast was found in the middle of the distribution of attainment (not shown in the table), the decline being greater at the 25th percentile than at the 75th.
Table 1: Scottish results in the PISA studies, at 10th and 90th percentiles in each domain, 2012-2022
Sources: PISA data bases (2012-2018), and pp. 358, 360 and 362 of Vol 1 of OECD report on PISA 2022.
It may be inferred from this that the decline most affected students who are weakest academically, and thus are most dependent on the formal teaching of schools. Weaker learners depend more on formal, structured teaching than do stronger learners. That has been one of the critiques of skills-based curricula since the 1970s, and would certainly appear to be applicable to the Curriculum for Excellence.
A further contrast along similar lines is in the changing social inequality of attainment, shown in Table 2. Social inequality is measured in the PISA studies by an index of ‘economic, social, and cultural capital’. The PISA reports compare mean attainment in each of the four quarters of this index. The table shows the trajectory in this from 2012 to 2022.
Table 2: Scottish results in the PISA studies, in lowest and highest quarters of the index of economic, social, and cultural capital, 2012-2022
Sources: PISA data bases (2012-2018), and online supplementary data for 2022 at https://stat.link/ax46rt.
In reading, inequality rose between 2012 and 2022 because the score in the lowest-status group fell twice as fast (a drop of 20 points) as in the highest-status group (a drop of 10). The change of inequality for the other two domains is somewhat stronger. In mathematics, the contrast is between a fall of 38 points in the lowest-status students and 25 for the highest-status. In science, it is 39 and 29. Again, the contrast between the middle two quarters of social status (not shown in the table) is a weaker version of these comparisons of the lowest and highest.
The speculative explanation is then that students with the least economic, cultural and social capital from home are most dependent on gaining access to these through the school curriculum. These are less available from a curriculum that places less emphasis on knowledge than on skills and on well-being – such as Curriculum for Excellence.
So a plausible explanation of the widening inequality is the curriculum. This widening is not merely between 2018 and 2022, but stretches right back to 2012. Curricula that do not concentrate on formal knowledge are particularly unhelpful to academically weak students and to students who come from homes where that knowledge is not readily available. The school curriculum is not so indispensable for able students, and for students from affluent homes where the parents themselves have abundant formal education.
Many writers have suggested this egalitarian potential of a curriculum based on knowledge. It can narrow social and academic inequalities in student understanding. Well-known examples in England are Daisy Christodoulou and David Didau. In Scotland, Bruce Robertson – rector of Berwickshire High School – has argued in his Teaching Delusion series of books that knowledge is empowering.
When the distinguished American sociologist E. D. Hirsch noted the widening educational inequalities in France after its move away from a knowledge-based curriculum in the three decades from the mid-1980s, he concluded this in his Why Knowledge Matters:
“If students gain the knowledge and vocabulary of the public sphere, they will score well. … If accidents of birth have excluded that knowledge from the home environment, and if the school does not supply it, then they will score badly.”
After the shock of these new PISA results, Scottish policy makers ought to take note.
Lindsay Paterson is emeritus professor of education policy at Edinburgh University.