Should gender be distinguished from sex? Yes, it is scientifically needed. Yes, it is particularly useful in regards to social sciences including psychology, sociology, economics, and political science.
As a researcher and economist, I often conduct surveys where I ask people about their gender. I have no way to verify the chromosomes of these people who are taking my survey. Therefore, at the time of interpretation of the survey data, I am compelled to interpret findings on the basis of the survey taker’s gender identity rather than their biological sex. Claiming the latter would be a misrepresentation of my data. Omitting the question completely would result in predictable omitted variable bias.
I may allow a choice between male and female. Alternatively, I may allow an “other” or “nonbinary” response option on the survey. Because the presence of the third option generally improves statistical models and reduces error and noise in the gender response estimates, and because gender is a table-stakes or standard control, omitting the third option is negligent. It is a matter of due diligence to provide this option and check for the effect. Many times there is no significant nonbinary effect. Great. Sometimes there is. Great. In any case, the construct is valid and should be studied as a matter of scientific due diligence.
Does sex generally correlate to gender identity? Of course, but there is still a value add to the explanatory power of my factors and models by allowing the additional expression, precisely because there are systematic differences in the responses to policy preference and so on exhibited by this third kind of person. See the video below which includes some interesting discussion on this and related topics:
I appreciate the views of the guest, but his many points generally line up with the validity of the nonbinary construct, even while he laments the use of it. He highlights that people of a certain kind, related to but not identical with nonbinary identity, can be expected “with 100 percent certainty” to act in a particular way. He’s embellishing of course, but this amounts to stating that the factor is scientifically important. The conclusion is that gender is important and distinct from sex, even while both are important and often correlate.