I don't understand why considering degree and reputation of the university when hiring so frowned upon ? It seems perfectly logical to me. Afterall, when you have 1000s of resume a day to sift through (not unusual for big companies like Google, etc), what criteria should you use to filter the candidates ? It's easy to fill a resume with lot of crap which is often fake or exaggerated. A degree and university reputation is a very objective measure that can used to quickly identify a potential good set of candidates, and then use actual interview to find bad apples from them. This approach optimizes toward having as few False positives as possible when hiring (False negatives on the other hand are plenty, and that's fine) - which is a stated goal of hiring for all big companies (Google, FB, etc), who are at luxury of having more candidates than they need, hence don't care much about False negatives.
Is it fair to people without degrees, esp, for cases where the only reason was that they couldn't afford it (considering how damn expensive higher education is in US) ? No, of course it is not, but then who said life has to be fair ? And of all things, why should we expect for-profit (and I use that word with respect, not derision) companies to hurt themselves to "just" try and make it fair to everyone ?
IMHO, education (both degree, and even more importantly the selectiveness of the university) serve as a very good proxy measure for how likely a candidate will be good. This is similar to say work-experience at a very reputable company, which is known for its selectivity. In both (university or top company) the scenarios you know that candidate had been filtered among a very large pool of candidates in past, so very likely to be bright.
As an example say, you want to select a dream team to attack the "Riemann Hypothesis" problem. Will you consider a candidate's credential (like education, university they are working at) as one of the big factors, or will you be just trying to interview anyone who applies, even if they don't have any exposure to higher math ? Say you get 100k applications, and can only interview at most 100 of them, does your answer change ? I know, that I will look at credentials, past work (which all correlate insanely with education & university affiliation) as the first measure to filter things out. Of course their will be outliers, and there will be exceptional people who are very good at Math without any degree (like Ramanujan), but there is no scalable way to find them in this scenario.
Similar example: What is the probability of some random dude on internet claiming he has a proof of P != NP being right vs say it coming from a reputed researcher (for the sake of argument say "Richard Karp") ? Whose proof are you going to read more carefully ? I guess the answer is obvious is this case. Now let's say it is between a random no-credential dude vs a PhD student of Richard Karp's (at Berkley) - who are you going to take more seriously ? Is it fair, probably not ? But statistically you will be much better with having these conscious & unconscious biases (backed up by data of course), then sampling uniformly from the world.
Is it fair to people without degrees, esp, for cases where the only reason was that they couldn't afford it (considering how damn expensive higher education is in US) ? No, of course it is not, but then who said life has to be fair ? And of all things, why should we expect for-profit (and I use that word with respect, not derision) companies to hurt themselves to "just" try and make it fair to everyone ?
IMHO, education (both degree, and even more importantly the selectiveness of the university) serve as a very good proxy measure for how likely a candidate will be good. This is similar to say work-experience at a very reputable company, which is known for its selectivity. In both (university or top company) the scenarios you know that candidate had been filtered among a very large pool of candidates in past, so very likely to be bright.
As an example say, you want to select a dream team to attack the "Riemann Hypothesis" problem. Will you consider a candidate's credential (like education, university they are working at) as one of the big factors, or will you be just trying to interview anyone who applies, even if they don't have any exposure to higher math ? Say you get 100k applications, and can only interview at most 100 of them, does your answer change ? I know, that I will look at credentials, past work (which all correlate insanely with education & university affiliation) as the first measure to filter things out. Of course their will be outliers, and there will be exceptional people who are very good at Math without any degree (like Ramanujan), but there is no scalable way to find them in this scenario.
Similar example: What is the probability of some random dude on internet claiming he has a proof of P != NP being right vs say it coming from a reputed researcher (for the sake of argument say "Richard Karp") ? Whose proof are you going to read more carefully ? I guess the answer is obvious is this case. Now let's say it is between a random no-credential dude vs a PhD student of Richard Karp's (at Berkley) - who are you going to take more seriously ? Is it fair, probably not ? But statistically you will be much better with having these conscious & unconscious biases (backed up by data of course), then sampling uniformly from the world.