I don't think you have any real evidence for whether the association of white-collar job and degree is post-hoc or not? Do you? Without evidence, it's kind of a cheap shot.
Stipulating the fact that jobs are just statistically going to be filled by people who automatically got degrees, I don't think it's a stretch (or a cheap shot) to suggest that the relationship between jobs and school is tenuous.
Moreover, given the fact that white collar jobs in large companies tend to be managed conservatively, and the fact that people without degrees are outliers (because almost every applicant for any job went to college by default), some of the "credentialing effect" of a degree is actually just an observational bias.
† Note also the preponderance of college attendees who attain degrees that are almost objectively without credentialing merit, such as the liberal arts.††
†† "Without credentialing merit" being intended as an objective statement and not as an argument that people shouldn't get liberal arts degrees; I think most degrees are without credentialing merit.
You say college has no causal effect, I say college has a causal effect (granting that we aren't talking about credentialing degrees, blah, blah).
The best you have done is destroy my argument in favor of causation; however, this doesn't imply that you are correct for NO causation, just that my argument doesn't work (maybe correct). To prove no causation you would have to evaluate all possible arguments for it, and show them all to be wrong. (good luck)
However, I do have partial evidence for causality based on the association between college and white collar jobs. I am basing this on the "warrant" that P(causation| association) >= P(causation) -- if there is association, the likelihood of causation is higher or the same.
So how about this: neither of us can say for sure whether college helps, but I have a little bit of faulty evidence and you have no evidence whatsoever.
We are in the same situation as most social scientists -- the evidence for anything really sucks, and is nowhere like the evidence for hard science propositions. However, very important policy decisions have to be made based on this faulty evidence. What most people do is pretend be sure and so guarantee they act foolishly.
Umm, wrong. I am arguing with the "minimal value of a degree as a credential outside the Professions". I have made that clear, by statements to the effect of "granting the professions are a different case."
I am saying -- I have limited evidence for causality in the effect of a degree (outside the professions), and you have no evidence for the absence of causality .
It's all good though...
Ok? Stipulated? Now what?
So ... retract your argument that college has no credentialling value in getting a white collar job (outside of the professions) ;) You have no evidence.
(Heuristically, I have evidence, but it is poor. But poor is better than none.)
Not much directly. Even in engineering, entry-level work is usually scutwork, so they probably won't get the chance to break anything important.
But hiring a flake can cause a lot of damage indirectly. If they lose an important document or attach the wrong appendix to the marketing proposal, you could lose a lot of money.
So if you're a big company with complicated procedures and a lot to lose -- what if that lost document is important because of a lawsuit? -- hiring dependable people trumps almost everything else. Bachelor's degrees are a great way to certify that someone can show up on time, follow directions, meet deadlines, and pay attention to one thing for four years at a time.
People make this observation all the time, but I'm guessing that even the most diligent students who make it showed up late to class often enough that they'd have gotten disciplined had class been a real job.
My evidence, to be fair, is the graphs I have vaguely seen associating educational outcomes with salary and "higher" job classifications.