[sokoloff]

Farther up the page is a comment that rings true to me: companies who have a surplus of applicants are prone to rationally bias their hiring practice to avoid making poor hires even at the risk of rejecting strong candidates.

Just because someone with many years of experience levels lower than typical doesn’t make them a bad person or even a bad hire. But it does make them a more risky candidate to hire (more likely to be "middle of the pack"; less likely to be "undiscovered superstar"), and so some companies choose to pass.

(I’m arguing that this is rational, not that it’s right, fair, or morally sound.)

[lubesGordi]

Well, I'd argue it still isn't rational given that the testing done during the interview is usually not telling much about the candidates engineering prowess.

[sokoloff]

I probably agree with you, but I think that's a different question. Whatever metric you're using to decide which candidates to offer employment, there's a rational reason to hold higher standards if you believe you have an effectively infinite surplus of maybe-excellent candidates if you pass on the current marginal candidate.

If you don't have that endless stream, you are more likely choosing between "this candidate" against "no candidate" vs "this candidate" against "the next qualified candidate".