I am a second-year graduate student in applied mathematics looking to apply for fellowships for this upcoming year. It also happens to be the case that I am a Caucasian male. Would it be more advantageous for me to select "Prefer not to respond" in applications that ask for my race and gender (especially applications that give preference to women and minority students), or is that generally considered to do more harm than good?

  • 16
    It shouldn't matter - their systems should split up application and demographic info, and anonymize the latter. No one on the review committee should be able to tie them together. That said, I'll let someone else answer because that's all hypothetical and I don't know if all systems do that well Commented Sep 14, 2017 at 20:29
  • 20
    I always select "prefer not to respond", not because I think it benefits me, but because it's the closest match to "that's none of your business."
    – Thomas
    Commented Sep 14, 2017 at 23:12
  • 4
    @Azor-Ahai OP does not specify a location, but in the US what you say is actually wrong. US law prohibits race/gender discrimination in employment, but race/gender discrimination (for the purpose of enhancing diversity) is legal both in college admissions and in awarding scholarships and fellowships. This has been reaffirmed as late as 2016 by our Supreme Court. en.wikipedia.org/wiki/Fisher_v._University_of_Texas_(2016)
    – David
    Commented Sep 15, 2017 at 2:14
  • @David These are fellowship applications, not admissions. Unless the fellowship is targeted toward women or minority groups, then the information is probably separated. But, that's why I left a comment. Commented Sep 15, 2017 at 2:32
  • 4
    @Azor-Ahai I'm not sure what you're saying. OP specifically asks about fellowships that express a preference for women and minority applications. These programs have been upheld as legal- i.e. there is no requirement for separating and anonymizing applicant demographics. The review committee absolutely can make decisions based on applicant gender and race.
    – David
    Commented Sep 15, 2017 at 2:39

2 Answers 2


Technically, I see nothing wrong with not reporting your race/gender. However, I suspect that if a fellowship gives preference to a specific group, putting "prefer not to respond" puts you outside of that group by default. Therefore, it would make no difference if you were a Caucasian or unidentified, you would still not gain the selection benefit for minorities. Hence the net benefit would be zero.

  • 2
    I don't think that's how those things work. Diversity information is separate from application data by law in the US, as Azor-Ahai said. This information is simply meant to be used to demonstrate compliance with various non-discrimination laws. Specifically inviting minority/female applicants does not mean they will, or legally can, look at this information. Commented Sep 15, 2017 at 0:58
  • 3
    Given the lack of region specifying information I have commented generally.
    – user44476
    Commented Sep 15, 2017 at 1:41
  • 2
    @Co3O4 Everything I've ever applied for in the UK has also made it clear that the diversity questionnaire is for statistical purposes only and the people making the hiring decisions will never see it. Commented Sep 15, 2017 at 12:06
  • And the UK legal situation can be summed up as "You can choose a candidate who has a protected characteristic over one who doesn’t if they’re both suitable for the job, but only on a case by case basis, not through general rule". gov.uk/employer-preventing-discrimination/recruitment
    – origimbo
    Commented Sep 15, 2017 at 14:59
  • 1
    @zibadawa-timmy As pointed out in comments responding to Azor-Ahai, in the US this data can be incorporated into decision making for scholarships and fellowships. It's only employment applications where its use is forbidden. Commented Sep 15, 2017 at 17:19

Please don't do this. Race/gender information has two purposes: 1) understanding the demographics of the applicant population, and 2) ensuring that strong candidates from underrepresented groups are not overlooked (i.e. some degree of affirmative action).

Most descriptions of 2) that I have heard essentially provide benefits to people who are underrepresented, or have in the past been discriminated against. Listing yourself either as a Caucasian male or as [decline to state] is likely to take you out of that population. [Caveat: I don't review fellowships myself, this is only based on what people have told me about admissions committees, etc.; for a given fellowship, the details of this may be publicly available.]

However, if a large fraction of Caucasian males decline to state their race/gender, the statistics that the NSF/whoever collects on their applicant pool will be biased. This will lead the NSF to have erroneous conclusions about whether their efforts to encourage diversity are working.

In essence, doing this is poisoning a well - corrupting a public source of data for negligible personal benefit.

  • 9
    It's probably zero personal benefit, not negligible personal benefit. Commented Sep 15, 2017 at 0:02
  • 26
    It's not like "prefer not to specify" carries no useful information. It carries the information "this is a person who prefers not to specify." If there's a lot of those people, it begs the question(s): why do so many prefer not to, and what can we do about it? Commented Sep 15, 2017 at 1:01

You must log in to answer this question.

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