
Originally Posted by
TheDeadFlagBlue
This post is "sipping the tea". Indeed, looking back, I am not sure why I posted this thread - the placement outcomes from my institution are stellar. In fact, based off what I know now, which has been learned via looking at placement options with a lot of context, I'd say that a lot of conventional wisdom found here is misguided.
Typical advice offered here - note that this is different from typical good advice - usually goes something like this "take math, take more math, get good grades, get perfect grades, go to MA program, get perfect grades there, get perfect GRE score". While it's true that taking signal from this advice and conforming to it would boost your chances of getting into a top PhD program, it would not prepare you for a successful career in the economics profession because economics is about publishing novel research papers, not about getting good grades and receiving head-pats. Admissions committees understand this, which is why they seem to heavily weigh LORs, someone obtaining NSF fellowship (mostly due to funding but it's a powerful signal), publications you might have etc. Obviously, bad grades matter, any bad signals matter - you can fail out of a PhD program - but the idea that "grades signal intelligence" or similar nonsense that is proffered about here is embarrassing. It's very, very, very hard to predict who will be a top researcher and who is not cut-out to be a top researcher - it's mostly in the residual. This is what top 10 programs care about - having a cohort of students who can produce many well-regarded research papers.
This is why typical good advice offered here encourages potential PhD-seekers to seek out opportunities to work as a RA. I think this advice is less typical than it should be because there's no performance targets or metrics associated with being a RA - those offering this advice can't say much beyond "working as RA at institution x or for professor y would be helpful". Every RA opportunity is different, even within large institutions. You may get the chance to be a coauthor and to pump out papers or work or you might be a code monkey or both etc. In any case, it seems to matter. I doubt that peers would have placed well without being RAs (can't say this for sure but it seems likely).