I routinely receive inquiries from potential graduate students, who send me their Curriculum Vitae, asking me whether I am planning to take new graduate students under my supervision. They state to me their interest for the research that I conduct, and inquire as to whether I would consider them as members of my research group. My response to them, in these cases, is usually “boiler plate”, almost.
The basic point is, I am fairly limited in what I can promise them. I cannot take students under my supervision, unless they are formally accepted into the physics graduate program in my department. So, all I can really tell them, is to submit an application to said program, including all relevant information, as well a required undergraduate (and graduate, if appropriate) transcripts, formal English test score records, and letters of recommendation from previous supervisors. There is also an application fee that they are required to pay upfront, which at my institution is $100 (CAD) — I believe in line with that at most comparable institutions (possibly slightly higher than average, but not by much). This is a fee whose purpose is none other than for an applicant to have his/her application considered by the institution. Presumably, it is applied toward the cost associated with the bureaucracy required to collect, sort and process all of the applications.
The application has to be reviewed by the graduate admission committee (GAC) of my department, whose difficult task is that of selecting, out a pool of typically around one hundred applicants, the twenty-five, thirty (not many more) who will form the entering graduate physics class of 2011. I can surely offer a prospective applicant my opinion on his/her chances of being admitted; I can definitely, and in fact I am encouraged to forward my recommendation to the GAC, if I feel that a particular student is worthy of special consideration. I can also inform them of my willingness to support the student financially, possibly even during the student’s first year, even though it is something that most of us are reluctant to do (we are talking people whom we have usually never met, and who will scarcely have any time for research in their first year, as they will be busy taking courses ).
Ultimately, however, applications from students in different areas of physics are received, practically every faculty has one or two “excellent” applicants, and it is often essentially impossible to compare applicants in different areas. Considerations of balance, fairness and equity among faculty and/or departmental research areas, besides obvious budgetary and infrastructural restrictions, inevitably play a major role in the admission process. Thus, no individual faculty can “guarantee” admission to an inquiring student, no matter how strong (s)he may appear on paper.
So, what would be my recommendation to an applicant ? Well, I would simply tell them to apply to many different departments (obviously not any one — see here for more on this subject), in order to maximize the chances of being admitted at least into one. But I have to confess to having overlooked, until very recently, an obvious problem, namely the financial one: to how many departments can a cash-strapped applicant realistically apply, especially a person from a third world country, if each application will cost him/her a hundred bucks ? That is an enormous amount of money for some.
And so, here is where the “catch-22” arises: the student is reluctant to apply, unless some concrete assurance is given that the application will be successful. On the other hand, no such assurance can be given, and a student can honestly be only told to apply .
It is inevitable for one to wonder how many possibly good applicants never even bother to apply, due to their objectively limited financial resources. There simply has to be a way to overcome this problem.
Doing away with the application fee altogether does not seem viable. Aside from the fact that there are probably objective costs associated with processing applications, I have heard several times, in America the opinion that, without an application fee, the reputation of the school is perceived as lesser — “Anything that you get for free, is worthless”, which reflects of course basic supply-and-demand mentality.
I think that a simple and quick “fix”, that most schools would be willing to implement and would entail only a minuscule financial effort for each one of us, is that of allowing individual faculty who wish to see applications of specific students examined by the departmental GAC, to pay upfront for the corresponding graduate application fee. Typically, these will be the applications of students who have contacted those faculty directly, before applying. Of course, this would be done with the full understanding, on the part of both the applicant and the “sponsoring” faculty, that the application may not be successful in the end, for a number of reasons (some mentioned above). I think it is an investment of a few hundred bucks every year, that many of us would be more than willing to make.
It would not be a “solution”, in the sense that it would still require applicants to establish an individual line of communication with a faculty, something that many, perhaps most students are unable and unwilling to do, for many valid reasons. But, it would still be a first step. Any other ideas ?
 Yes, I know, why not just get rid of those courses ? Eh, don’t get me started…
 A popular opinion is that, as long as a faculty is willing to commit to shouldering the full cost of supporting the student through his/her studies, by tapping into his/her individual research grant, admission should essentially always be granted by the department. I think that that is a bad idea for many reasons, primarily because it “binds” from the start a student to an individual faculty, something that may turn out later on not to be in the best interest of either one. Adopting such a policy would also likely bring about non-trivial problems with ensuring uniform standards of admission, and even of evaluation of students throughout their studies.