<p>Ok, this business of grades comes up often. Let me give you my thoughts on this (LONG POST).
In an ideal world, there would be a way of comparing the performances of students across the board. The GRE was partly designed to do this, but it is inadequate. Sadly the trend is to have LESS testing of students not more. Only a universally administered test that distinguished between high levels of student ability would equalize different schools.</p>
<p>In the absence of such tests one is faced with a dilemma that is mentioned often in the Caltech and MIT forums (and Chicago and Cornell). A mediocre grade from a tough program may be worse than a good grade from an easy university, no matter how unselective. This is a more serious problem for med school and law school than for PhD programs. On the other hand, it is less critical to graduate from the very top law and med schools just to get a decent job.</p>
<p>For the PhD the issue is complicated. We know which undergrad programs are tough, we just dont know what the appropriate conversion factors are. The upside is that if we see a 3.9 from MIT or Caltech and the recs are good, we usually dont care if the courses taken match the exact requirements of the program. That applicant moves to the head of the pile. A 4.0 from a mediocre school will keep you in the running when a 3.3 from Caltech/MIT might get you cut in some programs. But sometimes, there is nothing that the 4.0 from an undistinguished school can do to raise his/her visibility against the other applicants. A 4.0 with letters from teachers we dont know may not do any good.</p>
<p>The student at a top school like Cornell has the possibility of compensating for a lower grade by doing research as an undergrad and getting a strong and detailed letter of recommendation from a well known professor. Or even of publishing papers in serious journals. Students at elite schools have more chances to meet profs who are doing cutting edge research and who will know exactly what the top programs want.</p>
<p>The unfairness comes because schools minimize false positives not false negatives. Phd programs often want the kid who would have gotten an A at the very toughest schools. That sometimes means they mistakenly accept a weaker student with an A from a state school who might not perform as well as the 3.3 from a tough college. </p>
<p>On the other hand, some programs (for example Chicago in economics) take advantage of these errors by accepting many students who would be rejected by HYPSM and then flushing most of them in their early years of the PhD. Chicago has been extremely successful with this strategy of finding diamonds in the rough.</p>
<p>You also get to meet more top students in elite schools and these can be useful for business even if you do poorly in class. </p>
<p>Finally there are intangibles. There is the satisfaction of finishing in a top program. It is not about a career, its about how you view yourself. If you just want to maximize your career choices, go to the most famous school with the easiest grading but be aware that if you dont impress any profs, even that strategy may not pay off. So you pay your money and you take your chances.</p>