No idea. Between the two of us, it might be discrimination. Otherwise, I'd look at the job profiles and at the whole process. Just by looking at the average, I can't make an anti-discrimination law.
Wide averages show men life expectation is 10 years shorter than women's (TBG). What does that tell you? Shouldn't we send the men to retirement 10 years earlier? What's the actual situation?

Bottom line: sector or industry wide average is not enough, you've got to look into it, to the root causes. Or else, you're doing statistical experiments on the society's defenceless body. Frankenstein. Or Stalin.

"And in fairness should an individual's contract be down to how well that individual can negotiate it"

I admit it can be unfair, on the other hand we should know how to "sell" ourselves, I guess... I don't know.

"a systematic pay scale and grading and job evaluation methodology should be used.  So people who do similar jobs with the same levels of responsibility, should be paid the same for that"

Now you're speaking my language - just when I was thinking it extremely unlikely! :)
It looks like my previous proposition for a standardised CV. (unapplicable, that is :P )

"There's no question of trying to make more men be cleaners or more women be bin collectors but where it can be shown that the work is equivalent, they should be paid fairly"

I totally agree on principle (SEE!!! :) ). About this precise case: I'm wondering, maybe bin collecting is considered much more physical a job?... I'm not trying to excuse the system.

"a pay gap for certain ethnicities compared to white people"

Personally I know cases where immigrants hesitated to ask for standard market salaries. But it's hard to make a process of intention for that. Salary is negotiated, not imposed by the state. So one is offering, the other accepting, or the other way around. Now maybe some companies systematically propose inferior salaries to non-whites. Maybe they have reasons (I'd like to hear them), or just taking advantage. The others should not accept. I know it's easy to say - just the case with women. Employment market is extremely delicate, mixing arbitrare criteria with economic ones with human ones.

In any case, it doesn't work as easily as you expressed it in the statement I originally replied to:

"With recruitment the monitoring forms gives an idea of the numbers of people from different groups who apply and who get appointed in an organisation.  So it helps to highlight whether they are getting no applications from ethnic minorities or if they get a high number of applications from women but no women are getting appointed. Therefore the organisation needs to make changes to be more inclusive or to address any causes for there being a poor diversity of applicants or appointees."

I'll give you the benefit of the doubt, you probably simplified the process.

Free at last! Free at last! Thank God Almighty, we are free at last! (Martin Luther King)