If not the match, then what?

[NotAProgDirector]

---

Members don't see this ad.

Having seen how ugly the IM Fellowship application process was before there was a match, I remain convinced that although imperfect, the match is the best way to allocate residents into available spots. However, it's clearly less than perfect. There have been many people calling for changing or discarding the match. If you could change the match, what would you do instead?

In the recent article about the match discussed on another thread, several paragraphs caught my eye:

"The National Resident Matching Program was established in 1952 to make the residency application process less chaotic. Before that, top contenders might sit on offers from multiple programs while holding out for their first choices, leaving less-competitive applicants in limbo for months. For the latter group, this meant eleventh-hour moves across the country as the cards fell where they might. Like pro sports teams, hospitals would compete for prime students by extending offers as early as the sophomore year of medical school. Some sent "exploding" offers that would expire in 24 to 48 hours if the applicant did not reply."​

Any replacement for the match should address these issues.

The system compels participants to accept a binding contract. Unlike, say, third-year law students, Match applicants must train at whatever program chooses them. Conversely, hospitals must accept the doctors they are assigned. But the house has the clear advantage: While a law student can negotiate salary and perks, would-be residents have no say regarding pay and working conditions. "It's not really a free market," Thierman says. "You don't get to negotiate job offers."​

I don't know anything about how L3's get their first job. Theoretically (I think), a student graduating from law school could pass the bar and simply set up their own law practice, although how realistic this is is unclear. All medical students must do some residency training to proceed in clinical medicine.

David Goldenberg favors the MBA model: If the business school lets them, companies extend limited-time job offers to students. "It's not great, but it's better than the Match," he says.​

I also know nothing about the B-school system. This sounds ugly, though. I would take 10 med students for 6 months, and then keep the best 5?

So, if you don't think the match is the best way to do this, what should we do instead? Or if the match needs to be fixed, how to fix it?

 

[EaglesPA]

Good question.
Personally, I definitely prefer the match system.
The only major downside is that you have to accept whatever contract the programs give you and have no negotiating power. The other issue is the minor chance that the computer might screw up and place people in the wrong program - but this is miniscule.
What I know is that, without the match, if you were offered a position by your #3 choice, it would be really tempting to accept that offer rather than waiting and hoping for an offer from your #1. With the match algorithm, you can always shoot for your true #1 without fear of losing a spot at the other choices.

 

[Adam_K]

Any replacement for the match should address these issues.

The system compels participants to accept a binding contract. Unlike, say, third-year law students, Match applicants must train at whatever program chooses them. Conversely, hospitals must accept the doctors they are assigned. But the house has the clear advantage: While a law student can negotiate salary and perks, would-be residents have no say regarding pay and working conditions. "It's not really a free market," Thierman says. "You don't get to negotiate job offers."​

So, if you don't think the match is the best way to do this, what should we do instead? Or if the match needs to be fixed, how to fix it?

The Society for Industrial and Applied Mathematics newsletter on another thread has answers to this question. http://forums.studentdoctor.net/showthread.php?t=376826

You'd figure that these guys would have it.

http://www.siam.org/pdf/news/305.pdf

Peter Winkler of Bell Labs suggested a combinatorial matching in which students would submit their rankings of schools, together with the price differentials that would make the schools equivalent. A student might rank UCLA above Johns Hopkins, for instance, if UCLA's salary were $10,000 higher. The schools, for their part, would submit their lists of students, together with a maximum salary they're willing to pay for each. Winkler's algorithm then works in rounds, starting with the schools bidding the lowest possible salary for the students they want the most. Winkler's algorithm has the advantages of allowing pricing to be a factor and letting hospitals bid for specific students, but it also has drawbacks: It doesn't enable the hospitals to stay within a given budget, for instance, nor does it recognize a bargain in a student who's not a top-choice candidate.

Generally, we decided, a desirable algorithm for residency matching will have the following properties: In addition to having the stability criterion, it will force hospitals to pay the free-market value of each student it gets, allow the schools to stay within a budget and to bid higher for specific students, and motivate all parties to be honest about their preferences.

It turns out, of course, that economists have already addressed the price-competition issue. In a 1982 paper, Alexander Kelso and Vincent Crawford described a matching of workers to firms, by way of an auction, that satisfies most but not all of the above properties. And just this year, Paul Milgrom of Stanford University has put the auction-style theory and the matching theory into one over-arching framework.

As the economists have shown, it's possible to devise an algorithm that results in stable matchings, and allows hospitals to bid higher for specific students. The matching won't be stable, however, if the hospitals have to stay within a certain budget, and there are always opportunities for "gaming the system."

Oh, no, programs would have to start to pay real wages to residents. The horror of it.

In another review published by the same folks, an economist speculates:

www.siam.org/pdf/news/336.pdf

...that the wage depression in the medical market may result not just from the match, but
also from the limited number of good slots available and the lack of fluidity in the resident labor market.

Well, look at that. This happens to be exactly what Panda and I have been arguing about on the Gen Res forum. http://forums.studentdoctor.net/showthread.php?t=379820

 

[filter07]

How about multiple rounds of matching? After the first round you get a job/salary offer and you can take it or leave it. Subsequent rounds would ensue, with the catch that you can only go down your ranklist, not back up.

 

[deleted109597]

Seriously. How is would having residency applications be similar to medical school applications adversely affect the universe? Why is it that I can accept or decline medical school and job offers, but I am locked into a residency offer? Name any other system that works this way (apart from the military).

 

[EaglesPA]

Seriously. How is would having residency applications be similar to medical school applications adversely affect the universe? Why is it that I can accept or decline medical school and job offers, but I am locked into a residency offer? Name any other system that works this way (apart from the military).

Well, in the match you DO inherently "accept or decline offers" when you make your rank order list. If your first choice wants you there you'll get it. If they prefer other people, you move on to your 2nd choice and so on, until your choices are exhausted. The only difference here is that we make the decision earlier (at the time of rank list submission).

 

[deleted109597]

The only difference here is that we make the decision earlier (at the time of rank list submission).

That and we have no inherent way of knowing if those programs like us. So if it were like the real world, you could stop at one, if you found a job you wanted. However, in the match, no matter what the PD tells you, most people don't stop at 1. Or in my case, 11.

 

[NotAProgDirector]

I personally don't think that somehow linking salary differentials to ranks spots will really improve things. Do most people really choose a residency spot based on salary? I would expect that quality of training and location are probably much bigger drivers. And I don't think the match is the primary driver of resident salaries, as stated above in this thread.

Multiple match rounds is interesting. The NRMP recently floated a "second match" idea -- the idea would be to have the first match as is, but then instead of a scramble all unmatched applicants and open programs would have a chance to submit second rank lists and try again, over a 2 week period. I doubt this would address most of the concerns raised in these forums, although it would bring some additional order to the scramble. The major problem with this plan was that the NRMP planned to move match day to April to deal with the extended timeline, which no one liked.

A first match, which you are then able to decline and try to match again but can only go down your rank list doesn't make sense to me. In that case, why doesn't the applicant just rank programs lower on their rank list at higher positions? And if someone gives up their spot, then wouldn't it be more fair to allow a different applicant to "move up" their rank list if they now get a spot? Mathematically, I think this is an unstable problem -- i.e. the solution keeps changing through each round of the match.

Seriously. How is would having residency applications be similar to medical school applications adversely affect the universe? Why is it that I can accept or decline medical school and job offers, but I am locked into a residency offer? Name any other system that works this way (apart from the military).

Simply allowing programs to offer spots to people is where the whole process started, and was felt to be horrible. Perhaps times have changed so this will be less of an issue. Perhaps not. Per reports, the students felt that the match was a large improvement, although I have no primary data to back up that statement.

Here's what happens (I think): As the process starts, the top candidates will be offered spots at multiple programs. As each program has a limited number of spots, it is unlikely that programs will offer many more slots than those they have available (or will simply offer 30 people for 20 slots, and then simply accept the first 20 to respond, an "exploding offer"). Less competitive candidates will get no offers early, and will need to wait until the "first round draft picks" make their choices. This could be very ugly for those in the middle and bottom of the pack. Waiting for match day was frustrating for many. Remember what waiting for a med school letter of acceptance was like, and not being sure when or if it would arrive? It would seem (to me) that this sort of system is one big scramble, and anyone in the scramble (successful or not) can tell you how horrible it is.

From the program standpoint, I'd be offering to people and collecting acceptances and rejections. I'd start offering spots early in the application season (unless there was a date before which I could not offer anything), which means that theoretically I could fill my slots before the end of recruiting. Students would figure this out, and there would be a rush for people to interview early.

People would accept my offers, only days/weeks/months later to turn down those offers for better ones. I would then call the next person on my list, who might then accept my offer and turn down someone else's, etc. This leads to an interesting problem -- I might be tempted to offer a spot to less competitve candidates, figuring that this would be the best offer they could get and that they will at least be reliable in not withdrawing in the future. Hence, students in the middle could get squeezed.

It would be very difficult to coordinate couples, especially if trying for different specialties (although not impossible). Let's say both halves of a couple have an offer at Location A, but both would rather train at Location B. One of them gets an offer at B with an answer expected in 48 hours, the other does not (but might, if they wait). Does the person try to accept both offers? Drop A and take B and hope for the best?

Interestingly, mathematically, the match theoretically results in the ideal result without all the acceptances / rejections.

However, there would be some clear benefits to such a system:
1. If you applied, and were offered a spot at one of your top choices, you could cancel all interviews in your "lower ranked" programs and only interview in places that you were more interested in, saving time and money.

2. You might be able to negotiate for a higher salary (i.e. "well, MGH just offered me a spot but I think I'd like to stay here, but you'll need to increase my salary by $10,000"). Whether this would happen is unclear. Is it legal to pay different people at the same level hired into the same job different salaries? I have no idea.

3. You might know earlier where you were accepted compared with the current system, especially if you are highly competitve for your chosen specialty.

Is there another system like this? No, there isn't that I know of (short of professional sport drafts, which are not much of a comparison). Does that mean that this system is ridiculous and needs to be dismantled, or does it mean it's innovative and needs to be replicated elsewhere? No idea also.

 

[emtji]

2. You might be able to negotiate for a higher salary (i.e. "well, MGH just offered me a spot but I think I'd like to stay here, but you'll need to increase my salary by $10,000"). Whether this would happen is unclear. Is it legal to pay different people at the same level hired into the same job different salaries? I have no idea.

Hi Doc,

Programs are required to give us salary information at the interview. And people can vote with their feet. For my prelim, the program I ranked #2 on my list was in philadelphia and paid 42K. The program I landed at and ranked first is in new york and paid 52K. I didn't think i'd save 10K by living in philly and both programs were equally appealing otherwise, so i went for the new york program.

As most programs fill these days, i doubt any program would pay a premium to get one applicant over another. IF you went to MGH and said that program x is paying you 5K more, MGH would tell you to buzz off.

 

[Adam_K]

I personally don't think that somehow linking salary differentials to ranks spots will really improve things. Do most people really choose a residency spot based on salary? I would expect that quality of training and location are probably much bigger drivers. And I don't think the match is the primary driver of resident salaries, as stated above in this thread.

Take this for what it is worth, but the table I constructed to compare the residency programs I interviewed at included these items.

• Salary
• Cost of living
• Cost of medical insurance
• Moonlighting possibilities
• Vacation/Sick leave

One of the programs that offered exceptional training in the subspecialty I am interested in came out very poorly based on these factors, as it is located in a major metropolitan area, pays poorly and has a very high medical insurance copay with limited moonlighting possibilities. I ended up ranking this program near the end of my list for those reasons.

 

[LHCF]

Seriously. How is would having residency applications be similar to medical school applications adversely affect the universe? Why is it that I can accept or decline medical school and job offers, but I am locked into a residency offer? Name any other system that works this way (apart from the military).

So true! I feel like I enlisted in the military when I signed up for medical school.