^The problem with this is that it’s combining two errors: that correlation = causation and the ecological fallacy.
- **Correlation =/= Causation**. Yes, attendance at elite schools is correlated with acceptance (or at least attendance) to elite programs. But that could be for a variety of reasons. It could be, for instance, that students from elite undergrads are simply more likely to get accepted.
It could also be that students from elite programs are simply more interested in graduate school in the first place, and more likely to apply. This makes sense for a variety of reasons: They tend to come from wealthier backgrounds and can afford the undertaking of graduate school. (Yes, I do know that PhD porgrams are funded, but they still often require at least a minimal outlay of money.) Since they come from wealthier backgrounds, they are also more likely to have had family members who have attended graduate school, or family friends who are professors or whatnot, and thus more likely to have been exposed to the idea of going to graduate school. They thus may have a clearer idea of what graduate school actually entails.
In addition, professors at those elite schools may be more likely to assume that their students will want to or be interested in attending graduate school, and mention it more often. This may have the effect of making those students more interested or at least more aware of the possibility of attending graduate school. And let’s face it: the students who go to elite schools are the ones who are on average more interested in school in general anyway. They’d have to be, right? They were at the top of their classes in high school, and probably were nerdy academics their entire lives in some shape or another. Then they go to a school full of OTHER elite nerds who were also at the top of their classes in high school, and they talk about their philosophical ideas and scientific interests and big plans to take over the planet (in some way, shape, or form) with each other for four years. Of COURSE they are more interested in school, so maybe they are also more likely to apply to graduate school.
And even when the student from Regional State U applies to graduate school, maybe they are less likely to even apply to a top graduate program - because they believe that they can’t get in, or they believe that only students from UCLA and Stanford go to top programs. So it becomes a self-fulfilling prophecy.
Your average working-class student from Regional State U is not necessarily interested in getting a PhD. The students who are attracted to and attend regional state universities tend to be less well-off, and attend college often with more vocational goals in mind. Maybe they have student loan debt and do not realize that PhD programs are funded, or they want to work right away to support families they have, or they were never really into school in the first place (because of a variety of OTHER reasons) and just want to start working.
- **Ecological fallacy**. The ecological fallacy is "a logical fallacy in the interpretation of statistical data where inferences about the nature of individuals are deduced from inference for the group to which those individuals belong."
In other words, you say that “because this generally happens in Group X it must also the case that this will happen to Specific Person X1 within Group X.”
It may be true that generally - for example - CSU-Northridge students don’t get into the PhD program at Stanford. But - particularly since we don’t know why that happens - we cannot say whether or not any specific CSU-Northridge student’s chances are higher or lower than a UCLA or UCSB student’s chances of getting in. It all depends on a variety of complex factors.
And again, I’m not saying that elite school origination doesn’t make a difference in admissions. I suspect it does. Professors are humans who have their own biases and beliefs. If the question is whether it matters, it probably does - but if the question is whether it matters enough to compel an individual student to uproot their life to transfer schools, including spending an entire additional year in college (and the concomitant tuition money)? I’d say probably not.