The dramatic shift in the past few years has been the online application process. Now, training directors can sort through applicants in a spreadsheet form without needing to read cover letters. I'm not saying that this is happening everywhere, just that it is doable (a TD on the listserv discussed this a few months back) and some sites probably choose their interviewees that way (i.e., top 50 clinical hours, top 50 publications, etc.). I'm sure that this same comparison/ranking/process happened in the past, it's just a LOT easier to do now because of the computerized application process.
So the difficulty with great applicants not matching comes from, I think, the interviewee selection process.
Sites where I have interviewed (these past 2 years) have tried to illustrate their great training opportunities, definitely. But this is after they've already cut hundreds of applicants out.
If every applicant applied to every site (not feasible, obviously), the unmatched percentage would still be about the same. However, it seems that something is "lopsided" when applicants (and their programs, colleagues, etc.) are often left thinking, "Jeez, what happened? I'm a great candidate for internship but I didn't [match OR get any interviews]." Obviously, not every unmatched student is in this group but far too many are. If there are no apparent flaws in the applicant and people with less experience/qualifications/etc. do match (match rates at professional schools are relatively low compared to the overall rate but still surprisingly high if there's the assumption that they offer poor training/preparation), then it seems like the fault must lie with the sites' interview criteria or with the entire process (and the latter is too vague for me, so I assume it's the former
)
Based on my experiences (faulty/biased/etc.), Phase I interviews are offered like invitations to a middle school dance. Please bear with me on this. In middle school, boys generally ask the prettiest girl to go to the dance with them (if they ask anyone at all). What they don't realize is that there are multiple boys asking the same girl to the dance. The girl can only go with one, so she picks him and all the rest are disappointed (except for the boys who threw in a few "safety" date requests to girls that they've known since preschool or who live next door to them, i.e., the girls with whom they have similar/shared experiences). This is the Phase I interview and match process. Many sites ask the same 500-1000 prettiest applicants to interview with them (with a few idiosyncratic choices at most sites based on perception of fit and previous experience), ignoring that there are 15 other sites who this applicant also wants to dance with. On Phase I match day, the sites realize that they all asked the same prettiest applicants and some have a date for the dance and some were passed over. Then Phase II arrives (this is another big change/shift from previous years). Having been turned down by the prettiest girls (except for those boys who immediately asked the old standby friend after their initial rejection), the boys can stay home and not go to the dance, quickly ask the prettiest girl left to go to the dance with them (this was the old Clearinghouse version), or reevaluate their definition of pretty and act accordingly with it. So in Phase II, sites either drop out are forced to reevaluate their interview selection criteria. If you're going to match, you usually do it by the end of Phase II, though there are a few sites still left over the spring and summer due to a variety of scheduling, funding, or other changes. (Corresponding to the same last-minute pre-dance changes that happen to middle schoolers.)
Thanks for sticking with me through that. I don't mean to trivialize the importance of internship or the imbalance issues (or developmental issues or gender roles) but this conceptualization helps me to understand why 1) great sites end up in Phase II and 2) many great applicants don't get interviews in Phase I.