<p>
</p>
<p>I’d like to see some figures on this, if you have them. I’m not sure it works in such a straight-line fashion as you suggest. Grad students typically don’t get their own labs or other facilities, and research funds available to them are usually quite limited. Faculty get those things, and it can be quite expensive in some fields; less so in others. A philosopher, for example, typically needs nothing more than an office, books, a computer, and maybe some limited travel funds. Sciences and engineering, on the other hand, are going to be quite costly, and I suspect the US News “expenditures per student” category is going to favor science- and engineering-heavy schools over humanioties-oriented schools. </p>
<p>But other things equal, wouldn’t a school that gets to charge all (or a higher percentage) of these fixed faculty research costs to undergraduates as opposed to graduate students have an advantage, the way US News calculates it? It’s going to cost a physics professor the same money to run his lab in either case, regardless of the mix of graduate and undergraduate students he’s teaching. And perhaps more fundamentally, I’m not sure graduate students are the driver of educational costs, so much as they are the cheap labor that allows professors to carry out their costly high-end research.</p>