If we see that men tend to earn more than women, we can conclude that either
(a) equal pay for equal work/like work legislation is not in place or is not enforced, or
(b) such legislation is in place and is in force, and the discrepancy is due to the fact that men and women are not doing equal or like work.
If we’re exploring possibility (b) and we find that the discrepancy is not accounted for by age differences, educational differences, or other factors, we have to admit that it is accounted for by factors for which we haven’t controlled; e.g. if women of the same age as men and employed in the same capacity are paid less than men, it may be that they tend to have lower seniority or experience because their careers are more likely to have been interrupted.
The bottom line may simply be that women tend to earn less than men because they tend to make choices with respect to their careers which society does not value (at least in terms of paying them) as much as it values the choices which men tend to make.
Is this a problem? It may or may not be. Both the men and the women may be making the choices which maximise their own satisfaction, but women may tend to derive less satisfaction from money and more from other factors than men do. If in fact that is the case, then we should expect women to receive less pay than men.
Of course, all of this is an ideal model. Our very first assumption, that equal pay legislation is in place and is enforced, is probably not entirely correct in any society. And the assumption that men and women are freely making the choices which are reflected in their earnings is also a fairly theoretical ideal. Both men and women are constrained by culture, upbringing and other circumstances in their ability to make the choices which will maximise their own welfare.
But we cannot assume that, simply because women tend to earn less, they are more constrained or less free to make choices which maximise their own satisfaction or welfare. That would be to equate satisfaction/welfare with earnings, which would be a very silly equation. It’s entirely possible that in fact the men are more constrained than the women, and are forced into less satisfying choices – e.g. we offer high pay for unpleasant, difficult or strenuous jobs, and men are more disposed to pursue those jobs because of the obligation they feel to achieve material success/to provide materially not just for themselves but for others.
It is the case that men tend to enjoy poorer health, suffer more from stress and die younger then women, so it can’t be assumed that simply because they have higher earnings, they enjoy a better life, or that we would be making women’s lives better if we arranged matters so that they tended to make choices which earned them incomes comparable with men’s incomes. We might well be making them worse off.
The key to our policy here, surely, should be to maximise the freedom of both men and women to make the choices that are best conducive to their own welfare. If we succeed in this policy, we should expect people to make a diversity of choices with a diversity of outcomes, and if we find that men and women tend to make different choices with different outcomes, it doesn’t necessarily follow that either men or women are beiong discriminated against.