The suggestion that you can get better education outcomes for less money has to be an attractive one. That was the prospect held out by the Centre for Public Policy for Regions in its recent briefing on Scottish school spending and attainment.

The briefing concludes that Scotland has gained little from its increased spending on schools since devolution; that the other home nations, with lower spending, have seen significant rises in attainment; and that we could save mega bucks from school budgets with no detriment.

The report poses interesting questions, but also states that its conclusions are subject to considerable “uncertainty”, that there are “important areas of concern over data comparability” and that comparing exam results between jurisdictions will never be straightforward. I agree, yet CPPR persists in doing just that. Is it legitimate to base any policy conclusions on such “uncertain” data? Probably not, but that does not minimise a central question in the briefing about what secures the best outcomes at best price, and have we chosen those in Scotland?

I suppose the first big question to resolve is what are the outcomes being sought? The briefing views the world of education through exam results and costs incurred and, while legitimate as one dimension of debate about schools, it is only one dimension.

What our national debate showed, and the principles of A Curriculum for Excellence embody, is something broader and deeper. We explicitly want to develop successful learners, confident individuals, effective contributors and responsible citizens, and among the great current challenges to our system is how to measure success against those objectives. Exams will rightly be a part of that measurement, but how limiting if the only data we have and can use is exam results.

The report does offer some explanations for our costs and results - our sparse population adding to cost and our relative levels of deprivation contributing to low attainment. But CPPR says these can only explain part of the picture, and no real explanation for apparent variances in performance exists.

I concluded long ago it would almost always be possible to boost attainment in some chosen subject. You would narrow the curriculum, allocate more time to that subject, test more and coach pupils to meet the specific demands of your exams. You might even encourage more poorly- performing pupils to seek a future elsewhere!

I also concluded that those results would plateau and that parents, business and teachers would eventually rebel over the narrowing of the curriculum, the over-testing and coached performance taking precedence over the depth of learning. A balance always needs to be found between the constant pursuit of better exam results, breadth and depth of learning and building pupils’ wider experience. That balance will adjust over time but, whatever the temptations, we must never view exam results as the only measure of success and the first pursuit of education.

While the CPPR briefing shows how the other UK countries have progressed in English and maths, the evidence does not run only one way. For example, results in modern languages fell in England over the same period by as significant a margin, while Scottish results varied much less. Could this be one price of more focus on English and maths in England?

The briefing, rightly in my view, endorses the idea that sustained improvement will most likely come through better teacher quality and leadership, and that is exactly the source of part of the increased costs since devolution. McCrone was a first step to restoring the status of a beleaguered profession and makes teaching more competitive in the jobs market, bringing in more mature entrants. Add the standard for full registration, the teacher probation scheme, the qualification for headship, the chartered teacher, a new focus on CPD, A Curriculum for Excellence - all measures to strengthen the profession and quality of teaching - and you can see where some of the cost increases have come from.

There are suggestions in the report that performance-related pay and more parental choice could be prescriptions for better performance, yet there was no evidence advanced as to why better results or lower costs would result. Perhaps that is because there isn’t any strong evidence.

Having contended that there are hundreds of millions of savings to be made, with no loss to standards, the report doesn’t offer any suggestion of what is to go. The initiatives to raise performance in teaching, with limited exceptions, have found approval from parents, teachers, headteachers, industry and politicians across the political spectrum.

A lot of our costs reflect local and national policy choices. We have chosen to spend on education in the ways we have in order to craft the particular school system we want. Others will have chosen differently. There is little sign of any willingness in Scotland to reduce spend or reverse what has been created. The bill on its way through Parliament making the closure of rural schools more difficult and embedding key class sizes in statute serves to illustrate the point.

So, should we ignore the questions CPPR raises as overly simplistic? Tempting, but no: they warrant more attention and work, and echo my own persistent calls to constantly re-evaluate what we do, open ourselves to challenge and change whenever necessary. I also conclude that we need a better vocabulary and the measures to assess the outcomes we have defined and want - or exam results v costs will dominate and limit debate.

Peter Peacock was Education Minister in the former Scottish Executive.