The kids are back to school. And while Canada has long maintained a high level of education spending, it’s worth asking: Is it paying off?

Analysts often compare education performance among countries by using measures such as percentages of eligible students enrolled, per-pupil spending, and levels of education in the adult population. Canada shines on these metrics, spending 5.9 per cent of GDP on education (from kindergarten to post-graduate studies) in 2018, which was third-highest among G7 countries after the United Kingdom (6.1 per cent) and the United States (6 per cent). On the output side, Canada’s 60 per cent of 25 to 64 year olds with a post-secondary education outperforms all other OECD countries.

But such metrics tell us nothing about student achievement. As we all know from experience, high costs and quantities do not guarantee quality. As discussed in my new study published by the Fraser Institute, PISA—the OECD’s Programme for International School Assessment—offers the best, albeit limited, comparative achievement measures of student performance on the planet. For Canada, it’s a good news bad news situation.

Current PISA data are all pre-COVID. The planned 2021 assessment was postponed for a year, with results expected in 2023. But PISA has tested random samples of 15-year-old students in reading, math and science every three years since 2000. Canada did very well in each test cycle, outperforming all G7 countries except Japan, which achieved higher average scores in science and math than Canada in recent tests (although Canada has an overall superior record in reading, only scoring below Japan in 2012).

Given the Canadian public’s concern about math performance, the PISA scores for these two top-scoring G7 countries tell an interesting tale. In 2003, 2006 and 2009, the two countries were statistically tied in average math scores. Since then, Japan has consistently outscored Canada, although average scores in both countries have drifted downward recently.

Beyond the G7, Canada has also done well on the broader world stage. In the most recent 2018 PISA assessment, Canada’s average subject scores were close to or above the 90th percentile among all 78 participating countries, ranking 6th in reading, 8th in science and 12th in math.

Yet we must be careful when comparing ranks, as sampling and measurement errors can make it difficult to distinguish between close scores. Canada’s 2018 reading scores, for example, were so close to those of five other participants (Hong Kong, Estonia, Finland, Ireland and Korea) as to make them statistically indistinguishable. Yet only three participants—Macao, Singapore and four municipalities in China—had statistically higher average reading scores than Canada at or beyond the 95 per cent level of statistical confidence.

Averages are indispensable when comparing performance but are blind to the range of student scores, which is key when comparing student achievement. As such, PISA reports percentages of high and low performers in each subject. Again, Canada does well on this metric, with a significantly larger percentage (15 per cent) of high achievers in reading in 2018 than all other G7 and OECD countries other than the U.S., which was statistically tied at 13.5 per cent. Canada’s proportion of high achievers in math (18.3 per cent) was statistically tied with Germany (17.8 per cent) but below Japan’s significantly higher share (21 per cent). In science, Canada’s 11.3 per cent share of high achievers was statistically indistinguishable from those in Japan, Germany, the U.K. and the U.S. Only France and Italy had statistically significant smaller proportions of high performers.

That’s the good news for Canada. Here’s the bad news. In each subject, Canada’s average scores have been declining over time, as have the percentages of high performers. Comparable metrics in other G7 countries have either been stable or improving, which means Canada has been slowly but steadily losing ground, especially in science and math.

Again, current PISA data are all pre-COVID. Across Canada, government polices during the pandemic (including school closures) surely affected student performance. Given that Canadian students, particularly in Ontario, endured some of the longest school closures and most acute learning disruptions in the developed world, further score declines are likely.