En vogue HR policy is to use skin color, ethnicity, sexuality and gender (hereon “social identity”) to inform hiring decisions. This is called hiring for “diversity”.

The argument made is two-fold:

  1. that this is redistributive social justice, that (for example) black people have less social privilege and therefore should experience positive discrimination.

  2. that diversity of social identity necessarily leads to higher performing teams by introducing people of different backgrounds and experiences.

I refute these arguments by:

  1. showing that social privilege, while sometimes correlated with, is not aligned with social identity. Instead it is aligned with more complex issues such as economic status, education, crime and policing, non-nuclear families and cultural norms.

  2. showing that diversity of background and experience is not predicated on diversity of social identity.

  3. arguing that therefore those discriminating along lines of social identity are not solving the underlying problem, but are instead carving new lines of societal division.

Could I argue that the “identitarians” are making a category mistake in that they view the category of relevant characteristics to be those of social identity, when in fact it is another less superficial category?

If not, can I point to a logical fallacy in the identitarian position (eg correlation not causation?)?

  • I'd agree that thinking on these issues is usually muddled and subject to just the objections you make. But the issues are complex and i see no way to sort them out. Positive discrimination is minefield of subtle problems and unforseen consequences.
    – user20253
    Nov 17, 2017 at 13:26
  • Thank you. This question is about the logic of the position rather than the subject mater. Specifically: have I identified a category mistake? I would like to check my understanding of the term.
    – 52d6c6af
    Nov 17, 2017 at 13:59
  • 1
    Yes, I would say it's at least partly a category-error. The categories have been sloppily defined.
    – user20253
    Nov 17, 2017 at 14:14

1 Answer 1


The problem with this line of reasoning is that your refutations are not refuting the initial arguments.

For the first argument, unless you're arguing that these groups do not have less social privilege, you're not affecting the view that that discrepancy can be ameliorated by positive discrimination.

At best, your argument means that there are other groups e.g. poor, white males who suffer from a similar lack of social privilege, possibly for similar underlying reasons, that aren't being helped.

This may well be true (I'd be very surprised if it isn't) but it doesn't mean that positive discrimination would not help the specific groups. It may open up a charge of unfairness but that's not a refutation.

And your refutations do not at all address the second argument i.e. that diversity helps an organisation's performance. If it can be shown that that statement is empirically true, then it becomes a secondary consideration as to why it's true.

Now, I'm reasonably familiar with this field and, to my knowledge, the only diversity metric that has been shown to have a material performance effect is male/female gender ratio. That's not to say the others don't but the academic backing is light.

Having said all that, I happen to agree that there are deeper societal issues. Or to put it another way, it's not explicit discrimination at a company level that is the cause of the discrepancy so cannot be the driving solution.

However, economic status is, I believe, a factor as is visible opportunity. As such, my view is these measures can help, and should be employed, but are by no means sufficient to resolve the underlying issues.

Edit: Category Mistake

To the question of whether the arguments are category mistakes.

For the first argument, I would say no. These identity group do exhibit a lack of social privilege and that positive discrimination, at a hiring level, can alleviate some of the symptoms of that lack.

For the second argument, I will have to hedge. It is not a category mistake for male/female gender diversity as that is well established to have a bearing on performance. I am not aware of research that concluded a positive impact of other identity diversity on team performance. So there may be a mistake there.

Finally, in reference to a number of comments, it is very possible that the idea of identity selection, at an organisational level, does not significantly resolve the underlying causes of social privilege difference. Nor is it necessarily the most effective or just mechanism. As such, it may well be a category mistake to presume such.

  • Thank you for highlighting my failure to refute the argument as stated. Thinking more carefully, I realise the argument I am refuting is that "hiring based on diversity of social identity is a net social good" given that it actually "misses the point" ie. that although (e.g.) black skin is correlated with social disadvantage, it is not the black skin that should be in any way interesting/selected for (is this a a correlation/causation fallacy or a category error?) and so will ultimately fail in its social justice mission...
    – 52d6c6af
    Nov 17, 2017 at 12:16
  • ... and that any such policy is actually ironically socially regressive in that it re-enforces and creates division where there previously was none ("sorry you can't attend this learn-to-code session. I don't care if you are poor and disadvantaged. You're white.")
    – 52d6c6af
    Nov 17, 2017 at 12:17
  • @Ben I understand your point but I'm not sure I agree. You're essentially saying that, because an action doesn't resolve the cause, you shouldn't act on the symptoms. Questionable. Your second comment, assuming you can substantiate it, is more powerful i.e. that the actions make things worse. But you'd need to show that the evident negative effects outweigh any positive effects. The simple existence of negative effects isn't really sufficient.
    – Alex
    Nov 17, 2017 at 12:56
  • “You’re essentially saying that...” - this is an inaccurate summary. We should act. But we should act in a way that does not discriminate along lines of social identity. The very fact you take action to ameliorate an injustice just not make that act just.
    – 52d6c6af
    Nov 17, 2017 at 13:07
  • My argument is that discrimination using social identity (either way) is usually unethical (exceptions like being employed to test seatbelts for women do, of course, exist). The argument for this can be made as easily as any argument against racism or bigotry. Furthermore if the intention is to enact social justice then discriminating in this way is like taking one step forward (sure the Hispanic at the learn-to-code session will benefit) and another back (I’m disadvantaged and can’t attend because I am Chinese?!). Less discriminatory and more effective approaches exist.
    – 52d6c6af
    Nov 17, 2017 at 13:19

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .