The butcher, the baker and the teaching assistant all work in what most people would concede are low-paid jobs but that some would label as low-skilled. It’s the same for the lab technician, the nursing assistant and the care worker. I could go on, with a long list of jobs paid way below the median pay of around £30,000. These are the sorts of jobs that keep our society and economy running smoothly and that will come into stronger focus as a result of Brexit and the new immigration policy.

The home office’s new immigration system includes a route into the UK for “skilled workers with a job offer from an approved employer sponsor” where the role will be paid over £25,600 and require skills at level 3 (A level equivalent) or above. All of the roles I mentioned above - and many more - fail to meet that bar on pay (by a long way, due to the median being relatively high because of the enormous inequalities in our society), even though many of them do require a level 3 qualification. So low-paid might be a fair description, and low-valued, but certainly not low-skilled.

Background: The highest-paying A-level subjects

Edu-research: 3 questions teachers should always ask

News: College cyberattacks on the rise

This conflation of salaries and skill levels with abilities and judgements of what is good or bad, of value or not is an important and potentially problematic trend. It’s wider though than just immigration.

Measuring the value of education

Take the Department for Education report released this weekend for another example of how salaries and potential earnings over a lifetime are coming to the fore in political and media circles. The report, based on IFS research using the Longitudinal Education Outcomes (LEO) datasets, shows that, on average, graduates will earn 20 per cent more than non-graduates over their lifetimes, with most reports focusing on how men do better than women and how some subjects are more lucrative than others.

The report is full of important analysis and worth a look. One of the overall messages is that while 80 per cent of graduates earn more as a result of their degree, 20 per cent don’t. I’m sure the Treasury will be looking at that, along with the finding that about half of those who attend university represent a net cost to the exchequer over their lifetime.

The report comes at a time when the government is starting to get to grips with the LEO data that links educational achievement at all levels with salaries some years later. The data is a seductively simple, perhaps simplistic, way of measuring the “value” of education and training and looks likely to inform policies on fees or capping for degree courses as well as in other areas.

So, whether it is immigration or education policy, it seems that salaries and educational achievement are going to be bandied around in policy circles and in the media for some time to come. Unsurprisingly, discussing things like this generates a lot of ignorance, prejudice and emotion - and none of those usually makes for good judgements nor the best policymaking. Three big issues arise from all of this.

What message does it give?

The first is to question what sort of message it conveys to young people at school and college and to people in work. Given that only 60 per cent of 19-year-olds achieve a level 3 qualification, it worries me what all of this says to the other 40 per cent. If even level 3 is deemed to be unskilled, then we will continue to have millions of people feeling undervalued and their contributions often in socially and economically vital jobs simply dismissed. That cannot be right.

Using median pay as an outcome measure

The second issue is about using salaries, and particularly median pay, as an outcome measure. The point is that even if you believe that education is about getting a good job as an outcome (and there is nothing inherently wrong with that as one of the outcomes) the data is difficult to interpret at anything other than a macro scale. Using the data at course level without considering gender, sectoral and regional pay differentials feels risky. How, for instance, can we compare the lifetime returns for a woman completing a nursing degree with a man completing a law degree? Or someone completing a law degree in the North East compared with a law degree in London?

What is education for?

The third issue is about how we value education and what we think it is for. My own view is that I want to live in a society that provides transformative educational opportunities to everybody, regardless of circumstance, and that the ultimate outcome for education should be informed, confident and active citizens. It’s unlikely that those citizens will not work for most of their lives, so education that helps them be productive, earn good salaries and pay taxes is vital. Vital but not sufficient, because I want people to volunteer, to vote in elections, to be tolerant, to understand our place in the world and help shape our society for a better future.

So, for me, education is about work and about society. I worry that because we can now measure salary outcomes, we will forget the wider outcomes we seek from education. Those wider outcomes are just too difficult to measure, and I’ve never believed the adage that if you cannot measure something it is not worth measuring.

David Hughes is chief executive of the Association of Colleges.