[belorn]

I think a good comparison is to look at global warming studies. They too have a lot of unexplained data which are unaccounted for which could be due to natural phenomenon, but their conclusion is that global warming is man made and those unexplained parts are just unaccounted variables.

A scientific method to address the mix of unexplained data is to use prediction models on the assumption that the theory is correct and see if it works to explain the data. In the case of global warming we know that the prediction model does not work and the measured data is significant outside of what a natural phenomenon would look like. scientists has thus mostly discarded the theory of natural phenomenon and instead looked to combine the multiple studies and find man made explanations for unexplained data.

The same thing was done with gender pay. If we assume gender discrimination is causing the gender pay gap then a more gender equal nation should predictable have a lower gender pay gap. What we get instead is the opposite, the so called gender-equality paradox, where the more gender equal a nation is the higher the gender pay gap becomes. We really should follow the same scientific standard as global warming scientists and discard the theory of discrimination as the cause of gender pay gap, but we don't. Instead we see "half" of people trying to explain the failed predictions with the injection that the more gender equal nations is in reality less gender equal, and the other half arguing that it doesn't matter that the prediction model do not explain the data.

This is where I think most contention lies in the discussions around gender pay gap.