The article above does a lot of wordsmithing. Using "American Graduates", "People", "Graduates", and "applicants" exactly in ways that help make their point.... since the words are not interchangeable but the sound like they are. Its doubtful the reporter noticed the word smithing in the quotes.
http://www.nrmp.org/wp-content/uploads/2014/09/Charting-Outcomes-2014-Final.pdf
That is all you need to know. That is the actual raw data from the match itself. There are 25,061 spots in the formal match (really closer to 27,000 if you count urology, ophthalmology and the armed forces, all of which function outside of the match itself. and there are 2988 AOA residencies, which are also outside of the match). So we are talking between 25,000 and 30,000 spots depending on what math you want to use. There are only ~17,000 US MD graduates a year. And another 4,500 DO graduates. So if you include EVERY SINGLE u.s. graduate you're only at 21,500. Which is 8,500 spots shy of the number of actual residency spots there are. Even if you chip away as best as you can by adding in re-applicants (1,600) you still dont get anywhere close to saturation.
Let me preface this next statement with: "Exceptions occur but they are rare and they are called exceptions for a reason". When you're dealing with the matching system here, there is a clear and undeniable teiring. US graduates from MD schools go where they want and have no reason to not match somewhere *even if* they screw up their scores UNLESS they make a sh***y match list or they choose to only want to go to a very tiny number of places and would rather burn a year then go somewhere they dont prefer. DO graduates can go almost anywhere they want and have no reason not to match unless they screw up their scores or fit one of the two above categories. Anyone trained abroad is basically treated like leftovers and allowed to fill in the 8,500 spots that are "left over" after the US graduates have their fill. So when the article above lists 26,000 "people" it is combining the US graduates and the "second class citizens" together to allow the math to look scarier while ignoring that any unmatched US graduate would be taken in a heartbeat by any open program by they either screwed up their match list or chose to take a year off if they couldnt get what they want.
example 1: Donnie is a graduate from NYU. He felt that his sh** didnt stink and only applied to major NYC hospitals. He interviewed at 5 hospitals and felt confident only ranking 5. He was "that good". Everyone told him most people rank >10 places. But he didnt care. He didnt make the cut at any of his five hospitals and had "no match". Now as the NRMP guards their "SOAP" (the post-match program to fill remaining open spots) like a national secret, so we dont know hjow many spots are open, but lets say its a few thousand. because it is a few thousand. Donnie gets a list of a few thousnd hospitals which would *almost all* die to have him and take him in a heartbeat over the foreign graduate (generalizing). If he doesnt match its either because 1) He said to himself 'These options are beneath me, I just screwed up my rank list and should rank better next year' or 2) 'I only want to go to these elite NYC hospitals, so I'll do research at one for a year and earn credit in their eyes'
So when you have a clear "class system" and thousands of open spots; if someone from a US school doesnt match its because they want to or they have a criminal record. But when you want to make an argument that we need more spots, you cant admit that there is either a 17,000-25,000 imbalance or a 21,500-30,000 imbalance. It wont fly. You have to play with numbers. You argue on behalf of the american student, but you count the foreign graduate who "fill in" open spots when you make your argument. The federal government is paying for these spots and (internationally) even the worst one is highly desired. So every last spot gets filled. But its disingenuous to combine the US numbers and the total numbers in order to make a point.... but we all do it.