Does anyone know any detail on this and how self identification might skew the results? New legislation from April requires all employers of over 250 people to take a snap shot of pay by 'gender' (conflated as always with biological sex) which then must be published. As I understand it the data is reported as a percentage gap between male employees and female employees to give an overall picture.
Which is all good, but as employees self identify now how will this give us meaningful data?
I don't think self identification will scew the results to a noticible degree - purely on the basis of the small number of people for whom classifying by gender is difficult/contentious vs the vast majority of people for whom it is straight forward.
You would need to have a sizeable % of your workforce be people whose biological sex and gender are different for this to materially affect the reported statistics. It's a small sample size but I work for a largish company (10k employees). We are looking at this return at the moment and when we factor in these "edge cases" it doesn't even change enough to affect the post rounding number that will be shown in our statutory accounts.
Yes, what Fraser said. However it also make companies publish data for each quartile, and the higher quartiles contain less incumbents. It is possible that you could have a small number in the highest paid quartile which includes someone who self identifies. Possible but unlikely.