Today's news of £215m in new spending to help close that gap, all to be delivered this year, is undoubtedly welcome.
The situation is stark. Children from poorer backgrounds are less likely to do well at school than children from better-off backgrounds. Moreover, that gap in achievement is wide. In 2018/19, the proportion of school leavers achieving five or more awards at level five was 82.7 per cent for children from the least deprived areas. It was only 46.5 per cent for those from the most deprived areas.
Poverty is also widespread, meaning this educational gap is widely felt. According to Scottish Government figures, levels of child poverty have been stuck at levels well above targets for years, with 23 per cent of Scottish children living in absolute poverty in the last figures, for 2019/20.
This is a tragic waste of human potential, and one that perpetuates a vicious circle: without a good education children will struggle to progress in their own lives, with all the implications for their health and prosperity.
Education offers a way to break that cycle, yet efforts at improvement - even before the pandemic - were mixed. The most recent reports on the attainment gap, released in March but looking back at pre-pandemic Scotland, found some improvement, but also a gap that remained stubbornly wide.
Meanwhile, the OECD's long-awaited report on its review of Scotland's Curriculum for Excellence will finally emerge on Monday. The Government has already faced criticism that its late arrival, offering only three days' of debate before Parliament closes for summer, is intended to stifle scrutiny of its contents.
So while the promise of significant extra funds this year is welcome, it must be spent wisely if it's to have any impact. The recent election campaign offered little debate on the education problems that have dogged successive SNP governments. The fear must be that the same group of politicians, facing the same problems, may achieve the same grimly underwhelming results, only at greater expense. We hope to be proved wrong.