Residency Interview Application Inflation and Over-Invitation: Symptoms of the Same Problem

With each passing residency recruitment season, application inflation and over-invitation are becoming more prevalent for residency applicants and programs. These practices are employed by candidates anxious to ensure matching, and by residency programs to make sure PGY positions fill. And they are significantly affecting the application, interviewing, and match strategies at the expense (in both money and well-being) of applicants and programs alike.

For applicants, the zero-sum game of attending the “best” residency program results in a desire to maximize the number of interviews attended. Application inflation is the resulting drastic rise in the number of applications submitted per applicant. Over the last two decades, applications have increased fivefold in both cost and number. In 2018 alone, around four million applications were submitted via the Electronic Residency Application Service (ERAS) for a total cost of $80M, with 60 applications per domestic graduate and 150 applications per international medical graduate (IMG) submitted respectively across specialties.

For residency programs, over-invitation is the practice of inviting an increasing number of candidates—often many more than available positions—to ensure that all available positions will fill. This allows the program the ability to rank the maximum number of candidates in order to ensure a successful match, and hedge against entering the Supplemental Offer Acceptance Program (SOAP).

Unfortunately, while candidates and programs are both struggling for solutions, application inflation and over-invitations are very much the classic “chicken and egg” problem. And with each passing year, fear, anxiety, and concern over not matching fuels programs and candidates to partake in both practices. 

One point is clear, residency application interviews are inflated

Application inflation exists amongst candidates because of program over-invitation.


Over-invitation exists amongst programs because of candidate application inflation.

Currently, applicants have inadequate insight as to the competitiveness of their applications and see classmates applying to many programs while being encouraged to apply to more programs by medical schools looking to optimize internal match rates. Note: While some medical schools are attempting to provide moderate data-driven solutions, candidates remain hesitant and continue to choose to apply to many programs.

Programs, equally, have limited insight as to which candidates are strongly considering their program. As application numbers rise, and all programs invite more candidates to hedge filling interview positions, the chance of any program recruiting anyone applicant decreases. This is because the applicant pool is limited, and an applicant can attend one and only one residency program. So any program which invites an applicant that does not match is “wasting” an interview (in the economic sense). This results in some programs interviewing as many as 20 to 30 candidates for everyone available position. 

These concepts have drawn much attention in the GME world and its accompanying literature, and rightly so. Numerous efforts have been undertaken to find both the cause and a solution. However, the current status quo is difficult to alter, because the process is nuanced and multi-factorial—much to the frustration of both candidates and programs.  

Finger-pointing on both sides only seems to make matters worse. Programs may believe that students are overapplying, overscheduling, and then dropping dates when better programs come along. Applicants are challenged with programs inviting many more applicants than positions available causing interview seasons to fill in a matter of minutes. The sad truth is each side is simultaneously correct and incorrect, and this is only compounding the confusion.

An illustrative example:

An applicant’s perspective:

Let’s say there are 1,000 applicants applying to 100 programs in specialty X.  Each program, on average, matches 10 candidates. In this idealized example, this means that there are 1000 residency positions in specialty X and 1000 residency applicants. So each applicant, assuming an optimized match, should have an available residency position.  

Note, however, that the total number of candidates is fixed at 1000. That means each program can only select from this pool of candidates, and this number never increases or decreases. A program that doesn’t match all 10 positions will go unfilled. And an applicant that does not match to one of the 1000 available positions goes unmatched.

As an applicant, “I worked hard to get through pre-med. I then worked harder to ace my MCAT, applied to medical school and got accepted, took on a tremendous loan burden to learn every nerve in the anatomy lab, struggled through USMLE Step 1, sacrificed to excel during third-year clerkships, and organized a thorough ERAS application to apply to the specialty of my choosing.  

Years of self-sacrifice, learning, and educational costs all led to the residency application process. Since I’d never applied to residency, I looked at Doximity and FREIDA, I asked upperclassmen, I spoke with my advisory dean, academic advisors, and letter writers, browsed over the information provided by ERAS and NRMP, and tried to come up with some strategy—and a number of target programs to which to apply.  

I felt confident with my choices and used data to supplement my decision, but I recognized that even the data provided is incomplete and not targeted to myself or the characteristics of my application. As I spoke with more of my classmates, I saw that they were applying to more programs than I planned, with some even applying to multiple specialties.  

In a late-night panic, I read Reddit and Student Doctor Network (SDN) forums, and all of a sudden my list felt incomplete. The more I read, the more nervous I got and I started adding more programs to my list.  Since the thirty dollars or so to add each additional program is nominal, I rationalized that my whole career is at stake and spent additional money. What’s another $5000 on top of $300,000 in educational debt?  

So I applied to more programs. And the more I spoke with my peers, the more programs we added to our lists. Eventually, many of us were applying to all 100 programs!” 

And similar behavior occurs at all medical schools and all specialties. So, 1000 applicants submit 100 applications each. Every residency program receives 1000 applications, and in total 100,000 applications are submitted across all programs in the entire specialty.  

For those doing the math, of those 100,000 applications, 99,000 of them will not result in a match and be wasted (at best).

But back to our applicant:

“And then interview season began, and I, like all applicants, started paying out of pocket for flights and hotels. Additional loans were disbursed, and credit cards were maxed out. When I was invited to interview, my Fitbit buzzed the unique frequency I set and I signed up within 30 seconds for an open date. Some programs I wasn’t able to get to fast enough, and I added myself to multiple waitlists hoping for a spot. My planning pre-invite went out of the window because spots were filling too quickly to think.”

The program perspective:

“As the program director and/or program coordinator, ERAS opens for us on September 15th, and we’re bombarded with 1000 applications. It is the largest number of applications we have ever received. If we were to spend even five minutes reading every application, it would take 83 hours of consecutive reading.  This is humanly impossible given the tasks that exist and our other job duties, so we set filters.  

For our 10 spots to fill, we want to interview 20 candidates per position, a total of 200 candidates.  Last year, we had problems filling and we want to ensure this doesn’t happen, so this year we are determined to fill all of our interview spots.  We have 20 interview days and will interview 10 candidates each day, so 200 total interview positions.  

But to ensure we REALLY fill them, we invite 300 candidates at one time, in bulk.  

Our dates fill in a matter of minutes, which we are excited about! So many candidates want to interview at our program! And most of our waitlists are 80-100 candidates long…

However, we notice that there is a lot of turnovers. As more candidates cancel, there is even more turnover. Somehow by the time we get to January, our dates are only half full, some even less so. In total, we interview 160 candidates for the season, 40 less than capacity. We still have a strong enough match and just barely fill, narrowly avoiding the SOAP, but it is at the expense of significant administrative efforts. Next year we will likely need to invite even more to give ourselves even more of a buffer!

So, in total, we reviewed 1000 applications, invited 300 applicants, and only interviewed 160.  We took faculty time that wasn’t used, for 40 candidates that never interviewed. That’s GME for ya!”

The way forward

Does this sound familiar?  It should. Because this is the usual story. So how do we fix it?  While this will be discussed thoroughly in a future blog, the formula is as follows:

  1. Provide candidates and programs with data to assess the competitiveness and likelihood of a match.
  1. Give candidates information so they only apply to programs they are actually considering, with real data showing the likelihood of a match. Data that is personalized and evidence-based.  At the very least, stop candidates from applying to a program where they have less than a 5% chance of matching.
  1. Give program information about candidates they are likely to recruit that is personalized and evidence-based. At the very least, stop programs from inviting applicants to interview that they have less than a 5% chance of matching.
  1. Ensure programs ONLY invite the number of candidates for which they have available interview positions (and ideally fewer).
  1. Ensure programs invite candidates in waves: In other words, they invite a small group of “top” (or their most desirable) candidates to start and then follow with subsequent groups over the coming days to weeks. This ensures adequate time for “top” candidates to schedule interviews, gives the candidates flexibility, and signals to them that they are desired by the program. Candidates will come and interview if it fits into their travel schedule. They are interviewing at many programs; offering flexibility at yours signals that your program cares about its residents! Cancellation rates will decrease because candidates will get the dates they want, and completed interview rates will increase. EVERYONE wins.
  1. Remember:  Filling an interview season is not a proxy for optimizing your match!

But how do we ensure such data is available? Thalamus’ Analytics Team is working on features, products, and peer-reviewed articles to provide programs and applicants this information on a meta-data level to ensure better insight while ensuring individual and program privacy.  

Are you a specialty organization or residency program that would like to learn more? Are you an applicant looking for a better solution?  Please contact us at or reach out to us through our website—we’re happy to discuss further.