<p>Good stuff OP.</p>
<p>Further to the discussion, reliable data is going to be hard to produce. Instead of grad school acceptances, one could look at how many LAC grads (top 5 math LACs) there are on current research faculties (pick top 20). For comparison, youd also have to look at how many of the universities own undergrads show up.</p>
<p>The big problem is how you model the data to decide whether the numbers are proportional. You would need to estimate how many grads from each school youd expect to find, based on PhD production rates from each school, plus the career intentions of the students. Hopelessly difficult. And this would be one of the easier ways to assess the OPs question with a bit of objectivity.</p>
<p>So its tough. As tk21769 mentioned, all were left with in the end is a collection of opinions and anecdotes and testimony from the elusive b@r!um. I still believe that LAC math grads are under-represented on these faculties, but yes, its just conjecture. Maybe if we knew what their career aspirations were, their relatively low numbers in research faculties may be very acceptable.</p>
<p>For a good math student who is very interested in research, though, its probably not worth the risk to go beyond the very best LACs for math. LACs never seem to receive the benefit of the doubt.</p>