But actual disadvantage and principals’ perceptions of disadvantage don’t always align: 65 percent of principals in the United States say that more than 30 percent of their students are from disadvantaged homes, far more than in any other country. However, the actual percentage of disadvantaged students reported by PISA is just 13 percent, marginally higher than in Japan and Korea; but in those two countries, only 6 percent and 9 percent of principals, respectively, report a comparable share of disadvantaged students in their schools.

In other words, the actual incidence of child poverty is roughly the same among these three countries, but more than six times as many US principals reported that more than 30 percent of their students are disadvantaged. Conversely, in Croatia, Serbia and Singapore, more than 20 percent of students are disadvantaged, while 7 percent or fewer principals report significant populations of disadvantaged students.

Obviously, a child considered poor in the United States may be regarded as relatively wealthy in another country, but the fact that the perceived problem of socio-economic disadvantage among students is so much greater in the United States – and in France too – than the actual backgrounds of students also suggests that what school principals in some countries consider to be social disadvantage would not be considered such in others.

And there is a third important dimension, namely the actual impact of disadvantage on learning outcomes, which is shown by the size of the circles in the chart.4 That impact reflects whether an education system provides equitable learning opportunities. In countries like Finland, Iceland and Norway, one would expect this impact to be small because these countries have very little socio-economic disadvantage in their student populations. Achieving equity in school is easy when society distributes wealth and family education equitably.

But the more impressive examples are countries like PISA top-performer Singapore, where disadvantage is significant, but its impact on learning outcomes is only moderate. These countries seem very good at nurturing the extraordinary talents of ordinary students and at ensuring that every student benefits from excellent teaching.

In contrast, France has a comparatively small share of disadvantaged students, but school principals perceive this share to be large, and student learning outcomes are closely related to social background – more closely, in fact, than in any other country except Chile and the Slovak Republic. More generally, the results show that principals’ perception of disadvantage correlates with inequalities in education opportunities more strongly than real disadvantage does.