Pandemic disruptions in the market for medical residents

 The coronavirus pandemic and associated lockdowns and limitations have stressed a number of labor markets, including the one for new physicians.  Here's an article from the Journal of Surgical Education that suggests that, in a world of online interviewing, the number of interviews might usefully be capped. They also recommend signalling...

The Case for Capping Residency Interviews

Helen Kang Morgan, MD,*,1 Abigail F. Winkel, MD,† Taylor Standiford, BS,‡ Rodrigo Muñoz, MD,§ Eric A. Strand, MD,║ David A. Marzano, MD,* Tony Ogburn, MD,¶ Carol A. Major, MD,# Susan Cox, MD,⁎⁎ and Maya M. Hammoud, MD, MBA

J Surg Educ. 2020 Sep 14, doi: 10.1016/j.jsurg.2020.08.033 [Epub ahead of print] PMCID: PMC7489264

"As a result of the COVID-19 pandemic, residency programs will make an abrupt shift to virtual interviews in the 2021 residency application cycle.1 ... Medical students, medical schools, and residency programs have needed to react to sudden developments such as cancelled clinical electives, delayed or cancelled United States Medical Licensing Exams (USMLE),2 significant limitations on visiting student elective and sub-internship rotations,3 and changes in Electronic Residency Application Service (ERAS) timelines.4 Given this context, applicants may opt to increase their total number of residency applications as well as interviews accepted and completed, especially since they will no longer be limited by travel and cost deterrents.5, 6, 7 Likewise, residency programs are no longer logistically restrained to configure an applicant's interview schedule on a single day, and will have the ability to schedule interviews throughout multiple days and during non-business hours. In-person interactions provided by traditional interview day experiences have historically weighed heavily in determining mutual compatibility8 , 9; thus, both stakeholder groups will be looking to raise their chances of finding a match, including potentially increasing the number of interviews.

This is particularly troublesome given the current state of residency application processes, "fraught with misaligned stakeholder incentives.10, 11, 12, 13, 14, 15 Although the ratio of positions per applicant is higher now than ever before, the number of applications per applicant have risen.16 , 17 These numbers have increased rapidly in certain specialties, with the mean number of applications per applicant in obstetrics and gynecology (OBGYN) rising from 28 in 2010 to 66 in 2019.18 The consequences of application inflation are numerous and include decreased abilities for residency programs to perform holistic review of applicants with increased reliance on metrics such as USMLE scores. Residency programs also need to devote significant faculty and administrative time for the interview processes.19 The consequence of application inflation that will be of crucial importance this application cycle is the growing awareness that a small percentage of applicants has been receiving a disproportionate percentage of interview offers.20 , 21 In the era of virtual interviews, if these applicants choose to schedule all of their interview offers, there is a real potential for detrimental downstream effects to other applicants. This may also lead to a greater number of unfilled residency spots, with a larger number of programs and applicants required to enter into the Supplemental Offer and Acceptance Program. Given the paucity of data to inform best practices, there is a pressing need to model the potential effects of current application processes and applicant strategies in this disrupted application cycle."

...

"All stakeholders urgently need equitable solutions that address both individual and systems-level problems for this coming application cycle and beyond.30 Capping the number of interviews that an applicant can schedule could remedy 1 pressing flaw in current application interview processes. Implementing caps at the interview scheduling stage is preferable to capping at the application stage given the multiple complexities that must be considered such as DO and IMG status, and overall competitiveness. In addition, exceptions may need to be made for individuals participating in the Couples Match. The potential legal implications of mandatory interview caps are in the infancy of exploration. Capping interviews would likely not violate anti-trust laws given that applicants would still have the choice of where they would like to interview, however these issues would need to be further investigated. New measures such as preference signaling mechanisms30, 31, 32 need to be urgently considered in order for programs to be able to prioritize whom to offer interviews. The use of “tokens” would enable applicants to be able to convey interest to a set number of programs; this has been well-described in graduate PhD economics education literature.33 It will be imperative for “fit” to not become a proxy for decisions guided by unconscious bias,34 but instead, for principles of equity and inclusion to guide change during this time of accelerated change."

HT: Marc Melcher

Organizing tests of drugs to treat, and vaccines to prevent, coronavirus

 Here's a story on some of the difficulties in drug testing: among the difficulties are that positive test results for infection are taking too long after testing to allow prompt treatment...

Clinical Trials of Coronavirus Drugs Are Taking Longer Than Expected  By Katie Thomas

Aug. 14, 2020

"Antibody trials sponsored by Regeneron and Eli Lilly are off to a slow start because of a dearth of tests, overwhelmed hospitals and reluctant patients."

...

"While much of the world’s focus has been on the race to create a coronavirus vaccine, new drugs could also help curb the pandemic by making the disease less deadly. Because drugs are typically tested in sick patients in smaller clinical trials, they can also be developed more quickly than vaccines.

...

"The fast-moving disease has presented opportunities and challenges for the researchers testing antibodies. As the number of infections mounted in states like Florida, Texas and Arizona, there was no shortage of patients who would be eligible for trials. But at the same time, the outbreaks overwhelmed the very hospitals that would be overseeing the studies.

...

"“That doesn’t happen when you’re setting up diabetes trials or cancer trials,” he said. “We’ve had investigators say: ‘Look, I’d love to do research, but I don’t have time to set up a new trial. I’ve got an I.C.U. full of patients.’”

...

"Both the Regeneron and Eli Lilly trials require giving the drug within three days of taking a positive test.

"But with turnaround times in some areas lagging for five days or more, keeping within those time frames has proved difficult."

**********

With conventional (non-challenge) vaccine trials, having a high level of infection is important, to allow the effectiveness of vaccine candidates to be tested.  So, Brazil (and the U.S.):

 Coronavirus Crisis Has Made Brazil an Ideal Vaccine Laboratory  By Manuela Andreoni and Ernesto Londoño  Aug. 15, 2020

"Widespread contagion, a deep bench of scientists and a robust vaccine-making infrastructure have made Brazil an important player in the quest to find a vaccine."

...

"Some 5,000 Brazilians have also been recruited to support a vaccine trial conducted by AstraZeneca, a British-Swedish pharmaceutical company in partnership with Oxford University. An additional 1,000 volunteers in Brazil were recruited to test a vaccine developed by New York-based Pfizer.

"Researchers need countries with large enough outbreaks to assess whether a vaccine will work. Some volunteers are given the potential vaccine while others are given a placebo, but they have to be in a place where enough virus is circulating to test the vaccine’s efficacy.

"Brazil, where the virus has infected more than three million people, has clear conditions for these trials. And it will be the only country other than the United States to be playing a major role in three of the leading studies as an unparalleled quest for a vaccine has led to unusually fast regulatory approvals and hastily brokered partnerships."

How will the pandemic affect the medical Match?

Some thoughts on the medical match in JAMA:

Potential Implications of COVID-19 for the 2020-2021 Residency Application Cycle
Maya M. Hammoud, MD, MBA1; Taylor Standiford, BS1; J. Bryan Carmody, MD, MPH2
JAMA. Published online June 03, 2020. doi:10.1001/jama.2020.8911

"Even before COVID-19, calls to reform the residency selection process were becoming more frequent.1,2 Many issues are related to the increasing number of programs to which applicants apply. In 2019-2020, applicants from US medical schools applied to an average of 65 programs, and international medical graduates (IMGs) applied to an average of 137 programs.3 This number of applications likely does not improve match rates and imposes a substantial cost on applicants and a potentially unmanageable load on program directors.

"It is possible that the disruptions caused by COVID-19 may result in an increase in the number of applications and further stress this already challenged system. Due to testing center closures, many applicants have been unable to take portions of the United States Medical Licensing Examination (USMLE). This is especially critical for IMGs, who must pass the Step 2 Clinical Skills Examination to obtain certification from the Educational Commission for Foreign Medical Graduates and apply to US residency programs. Additionally, medical schools have shortened clerkships, shifted to virtual rotations, and canceled away electives, all of which may reduce student opportunities to obtain meaningful faculty evaluations, letters of recommendation, and signal their interest to programs. Students will encounter significant uncertainty regarding how their applications will be evaluated and may respond by applying to even more programs.

Program directors may have difficulty identifying applicants to interview without use of traditional screening metrics. Yet, challenges will persist even after interviews are offered; if travel disruptions and social distancing persist into the interview season, programs may be unable to offer in-person interviews. Temporary solutions, such as conducting virtual interviews or waiving requirements for USMLE scores and letters of recommendation, will be necessary for the selection process to function. But these stopgap solutions may exacerbate existing problems with residency selection and lead to undesirable consequences. For instance, the use of virtual interviews could result in applicants participating in more interviews. Currently, the number of interviews an applicant attends is limited by time and travel expense, but these constraints will be less relevant with virtual interviews. Yet because many programs rely on the same screening metrics, many programs already overinvite the same pool of highly-qualified applicants, with just 7% to 21% of the applicant pool filling half of all interview slots in some specialties.4 The result of those applicants accepting more interview invitations could be an increase in both the number of unmatched applicants and unfilled programs.

Interview congestion in the Ophthalmology Residency Match

An ophthalmology residency program surveyed all its applicants on their experience in the match:

Current Applicant Perceptions of the Ophthalmology Residency Match
Michael J. Venincasa, MD; Louis Z. Cai, MD; Steven J. Gedde, MD; Tara Uhler, MD; Jayanth Sridhar, MD

JAMA Ophthalmology May 2020 Volume 138, Number 5 

"Hundreds of individuals apply for ophthalmology residency positions each year using the Centralized Application Services (CAS), administered by San Francisco Residency and Fellowship Matching Services (SF Match). Although the match rate remains relatively stable at approximately 75%, the mean number of applications submitted has risen from 48 in 2008 to 75 in 2019.1,2 In 2010, highly qualified applicants were advised to apply to between 10 and 20 residency programs,3 but more recent studies suggested a target of 45 applications for these applicants and more than 80 for applicants with less competitive qualification.2 The application process represents a considerable financial burden for applicants; in 2018-2019, the CAS application alone cost $685 to apply to 45 programs, which increased to $1910 for 80 programs. These high costs are not unique to ophthalmology. In emergency medicine, the cost of securing a residency position was estimated at $8312 in 2016.4

These trends also come with increasing administrative burden for residency programs tasked with reviewing rising numbers of applications. As a result, many programs have increasingly emphasized quantifiable cognitive measures, such as clinical grades and the US Medical Licensing Examination (USMLE) board scores.3 The USMLE Step 1 scores and Alpha Omega Alpha Honors Medical Society membership are factors with statistically significant associations with matching into an ophthalmology residency.5

...

"Respondents applied to a mean (SD) of 76.4 (23.5) ophthalmology residency programs, received 14.0 (9.0) invitations to interview (Figure 1), and attended 10.3 (4.4) interviews. 

...

"When respondents received an interview invitation without the involvement of a wait-list, they most commonly reported receiving the invitation between 3 and 4 weeks prior to the interview date (n = 87 [47.8%]). When instead receiving their invitation from a wait-list (n = 92 [49.7%]), the most common lead time was 1 to 2 weeks prior to the interview date (n = 43 [46.2%]), with 20 (21.5%) invitations arriving less than 1 week prior

...

"Many applicants struggled with scheduling conflicts with other residency programs, where interview dates overlapped or the desired date was filled at the time of invitation response. Certain dates were especially popular for residency programs, with 23 of 116 programs (19.8%) holding interviews on a single day during the 2018-2019 interview cycle.

***********

Note that the Ophtalmology residency match is run by SFMatch, not the larger NRMP, but the growing number of applications and interviews are common to both matching platforms.

Confusion in NYC high school wait lists

In August, the New York City Department of Education announced a change in the school choice assignment process--without announcing any details.  But the plan was that after the initial run of the deferred acceptance algorithm, they would institute some sort of wait lists. I blogged about it at the time, and was concerned by the lack of detail.

Here's a current story from Chalkbeat that suggests that the details are still opaque, but that families are learning that the waitlist position they were given isn't reliable:

‘How can you move back on a waiting list?’: NYC’s high school admissions tweaks spark confusion
By Alex Zimmerman  May 8, 2020

"students vying for the city’s most coveted schools are discovering that their position on high school waitlists can worsen over time, a situation that has come as a surprise to some families — adding anxiety to an admissions process that is already famous for its complexity.
...
"Every student who fills out an application and does not get into their top choice is automatically waitlisted. If you get your third choice school, for example, you’ll be on the waitlist for your No. 1 and 2 choices. Nearly 44,000 students did not get into their first choice high school this year, automatically placing them on at least one waitlist.

"The second way is that students can add themselves to any waitlist once the initial matching process is over, even for schools a student didn’t initially apply to.

"In general, students who initially applied to a school but didn’t get in and are automatically added to its waitlist should be ranked ahead of students who add themselves later on, officials said. But there are exceptions.

"The first major exception is if a student is in a higher priority group than someone who is already on the waitlist. Some schools, for instance, give preference to students who live in certain neighborhoods, which can override a student’s position on the waitlist even if they were added first. (Officials said this is the most common reason a student would see their position worsen.)

"Olga Ramos, the admissions director at Bard High School Early College Queens, pointed to a second reason families can move backward — something that surprised her at first.

"If a student got into their first choice school, and listed Bard as their second choice, they could still add themselves to Bard’s waitlist and be considered as if they had been automatically added — potentially bypassing students who were already on the list."

*********
Here's an earlier story in Chalkbeat by Mr. Zimmerman, indicating that the system was still pretty opaque as the school choice process got ready to announce admissions in March:

NYC high school offers are coming this week with a big change: waitlists. Here’s what you should know.  By Alex Zimmerman  Mar 18, 2020

Here's what was known then...

"What are these waitlists, anyway?

"New York City students must apply to high school, listing up to 12 schools they want to attend. A complicated algorithm, developed by a Nobel prize-winning economist, then matches a student to one of their choices.

"That fundamental algorithm is not changing. But for the first time this year, any student who does not get into their first choice school will automatically be added to the waitlist of every single higher-ranked school they didn’t get into.

"Every school that has more applicants than seats will have a waitlist. It’s a similar model that the education department uses for pre-K, kindergarten, and middle schools — something education department officials said is an advantage."

**********
Here's a story from the time of the initial announcement:

Goodbye round two applications, hello waitlists: NYC announces changes to high school admissions
By Christina Veiga and Alex Zimmerman   Aug 15, 2019

"Starting next year, the city will allow students to sit on waiting lists for schools they wanted to attend, but didn’t get into. The city is also eliminating the second round of admissions, which it now uses to for students who aren’t matched to a school they applied to during the typical process.
...
"“It’s like going to a store and getting the ticket, you know what number you are, and you know how many folks are ahead of you, and you’ll be able to watch the process go,” said Deputy Chancellor Josh Wallack. “You’ll also be able to talk with an administrator in a school who can give you a sense of how much waitlists move each year and that varies a bit by school.”
*****

I'm still confused about a different issue that I haven't yet seen addressed. In the original school choice system using the deferred acceptance algorithm, there was a second round in which students unmatched in the first round were asked for additional preferences over schools, so that they could be matched.  How were those unmatched students assigned to schools this year?

Here's my August post:

Friday, August 16, 2019 

Waitlists in NYC school choice--early reflections on yesterday's initial announcements

Congestion and competition in college admissions (in the WSJ)

Is college admissions ripe for re-design?  (The problems outlined are real, but I'm skeptical that there's the consensus needed for a major overhaul...)

How to Fix College Admissions
Getting into a top school is a stressful, unpredictable process. Here are 10 ways to make it fairer and more transparent.  By Melissa Korn

"We asked college admissions officers, high school and private counselors, parents, students and others for ways to make the system fairer, more transparent and less painful for everyone involved. Here are 10 of their ideas—some easy to implement, others just meant to start a conversation—to reform the status quo.
...
"2. Limit the number of colleges to which students may apply. Thanks in part to the ease of applying online—especially through the Common Application, which allows applicants to use one basic form for hundreds of colleges—36% of students submitted seven or more applications in 2017, up from 10% in 1995. “The number of clicks you can make on the Common App causes congestion in the system,” says Alvin Roth, a Nobel Prize-winning Stanford University economist who helped to design the system that matches new doctors with residency programs.

"Schools pursue aggressive outreach, urging even fairly unqualified applicants to apply, then boast every spring about how many they rejected, as if exclusivity is proof of quality. Ballooning application numbers, combined with stagnant class sizes, cause acceptance rates to slide even lower into the single digits at places like Columbia and Pomona. As a result, high-school seniors apply to more schools just in case, and the vicious cycle continues—creating havoc for schools that can’t predict their yields. The overall yield rate for new freshmen at U.S. colleges fell to 34% in 2017 from 48% in 2007.
...
"Almost nobody needs to submit 20 applications; a reasonable limit would be as low as a half dozen, assuming that students receive meaningful counseling. High schools could enforce the cap by only agreeing to submit a certain number of official transcripts to colleges. The College Board and ACT could also limit distribution of SAT and ACT results, but they have little incentive to do so, since they make money from sending scores.
...
"9. ...Even more radical, schools could try some version of the algorithm used to determine matches for medical residency programs, which involves programs and medical students ranking one another and then being paired up by a computer system. This would be a heavy lift, however, as colleges would need to coordinate their procedures to rank candidates, run the computer program and inform all parties about the outcomes."

The common app: reduced friction and increased congestion by Knight and Schiff

Reducing Frictions in College Admissions: Evidence from the Common Application

Brian G. Knight, Nathan M. Schiff

NBER Working Paper No. 26151
Issued in August 2019

Abstract: "College admissions in the U.S. is decentralized, with students applying separately to each school. This creates frictions in the college admissions process and, if substantial, might ultimately limit student choice. In this paper, we study the introduction of the Common Application (CA) platform, under which students submit a single application to all member schools, potentially reducing frictions and increasing student choice. We first document that joining the CA increases the number of applications received by schools, consistent with reduced frictions. Joining the CA also reduces the yield on accepted students, consistent with increased student choice, and institutions respond to the reduced yield by admitting more students. In line with these findings, we document that the CA has accelerated geographic integration: upon joining, schools attract more foreign students and more out-of-state students, especially from other states with significant CA membership, consistent with network effects. Finally, we find some evidence that joining the CA increases freshmen SAT scores. If so, and given that CA members tend to be more selective institutions, the CA has contributed to stratification, the widening gap between more selective and less selective schools."

"The CA began with just 15 colleges in 1975 but grew rapidly thereafter, with increases in member-ship in every year since 1975 and a significant acceleration of membership starting around 2000(Figure 1). It currently includes over 700 institutions, which, taken together, receive approximately4 million applications from 1 million students annually."

More chaos in the medical resident interviewing process

Not only do graduating medical students go on (too) many interviews for residency positions, but the process by which interviews are offered and accepted is chaotic. (I'm reminded of the process of offering and accepting actual positions in the 1940's, before a matching clearinghouse was first developed...)

Here's a paper from the August 2019 issue of the Journal of Graduate Medical Education, by three concerned doctors at Northwestern University, who describe the situation and then offer their own suggestions. (I admire the description, and am skeptical that the suggestions are radical enough to change the incentives responsible for the current congestion...)

A Challenge to Disrupt the Disruptive Process of Residency Interview Invitations
Matthew R. Klein, MD, MPH
Sandra M. Sanguino, MD, MPH
David H. Salzman, MD, MEd

"Residency programs may also contribute to the sense of urgency to secure an interview by offering
more invitations than available interview slots. The rise in the number of applications per
applicant, while certainly multifactorial, is facilitated in part by the ease of online interview scheduling...
...
"In the current environment, applicants who receive an interview invitation while they are in a situation where interruption is impossible—such as taking a clerkship examination, assisting in an operating room, performing a bedside procedure, participating in a difficult conversation with a patient or family, rounding on the wards, or simply being asleep as a result of time zone differences—are effectively penalized. A delay in response may result in the inability to schedule a favorable date or an increase in the cost to travel to an interview at a less convenient time, or it may preclude an interview altogether if all interview slots have been taken.
...
"We propose a 2-step process that we believe balances the need for residency programs to recruit
applicants and schedule interviews with the goal of preserving an environment that allows students to
maintain their commitment to their educational and patient care responsibilities. We recommend the
following process
Step 1: A program informs an applicant of an invitation for an interview. In that communication,
the program shares the interview dates and indicates the date and time when online interview
scheduling will become available. This allows medical students to plan in advance when they
need to be briefly excused from clinical or educational responsibilities to attend to interview
scheduling
Step 2: At the time indicated in the initial communication, and no earlier than the day after
that communication, the online interview scheduling system opens to applicants
...
"In an effort to further decrease pressure to reply immediately to an interview invitation, programs
should also ensure that the number of initial invitations does not exceed the number of interview
slots available. Additional interview offers from a wait-list could occur as needed based on unclaimed
interview slots or subsequent cancellations."

Waitlists in NYC school choice--early reflections on yesterday's initial announcements

Yesterday the New York City Department of Education announced a change in the school choice assignment process--I gather that after one round of deferred acceptance, they will do something else, involving interim assignments  and wait lists.  (The original design included a subsequent round of deferred acceptance, after disseminating to unmatched students a list of schools with vacancies, and eliciting new preference lists for this second round.)
The details of the new plan for the second round aren't yet clear (at least to me).

Here's the press release from the city:

Mayor de Blasio, Chancellor Carranza Announce Easier and More Transparent Middle and High School Admissions Process
August 15, 2019
Families will now have one form and one deadline for middle and high school admissions

"“We are changing the middle and high school application processes so families don’t have to go through the gauntlet just to get a placement. There will be one application round and one deadline to make everyone’s lives easier.”

“We’ve heard from families and educators that they want a simpler, more transparent, and more accessible system of school choice, and today we’re taking a step forward,” said Schools Chancellor Richard A. Carranza. “This common-sense change will make a real difference for families across the five boroughs, and improve our middle and high school choice process for years to come.”

The DOE is eliminating the second application rounds for middle and high school. The main round application process and timeline will remain the same, with middle and high school applications opening in October with a December deadline. Students will receive their offer in March. Families can still appeal for travel, safety, or medical hardships; if families have any hardship, they will be able to access in-person support at Family Welcome Centers, rather than wait to participate in a second process. The waitlists will open after offers are released and will be a simpler, clearer process for families, increasing:

• Transparency:  By knowing their waitlist position, families have a better understanding of their chances of getting into a preferred school option in the event that seats become available.
• Ease: This is a shorter process that requires less paperwork. Rather than having to complete a second application and wait weeks—often into May or June—for a second decision or offer, families will complete one process, receive one offer, and receive any additional offers based on waitlist position.
• Consistency:  Families will now have one admissions system at all grade levels, with the changes to the middle and high school process making it more similar to the elementary school admissions process. Currently the elementary school process has one round, and the middle and high school processes have two rounds with different names; now, families will not need to learn a different process each time a child applies to a new school—allowing them to focus on school options and not process."

************
From the WSJ:

New York City Introduces Wait Lists for Students Unhappy With School Placements
City’s complex school-choice system, in which students hope to be assigned to a top pick, has long been daunting for many families
By Leslie Brody

"In recent years, applicants who didn’t like their middle school assignments—given in spring for entry the next September—would have to go through an appeals process. High school applicants who didn’t like their offers would have to try a second round of applications and then appeal if need be.

"Under the new Department of Education system for fall 2020, students will be placed on wait lists for each school listed higher on their applications than the schools they were admitted to. They will be informed of their positions on wait lists and may be offered seats if they open up."

*******************
I got some emails about this. Here's my reply to a reporter...

"at this point I’m only an observer of NYC schools from a distance—I haven’t been involved in advising them for over a decade, and even then I worked only on the high school match, not anything involving middle schools.

So I don’t know anything about the current plans besides what I’ve read today ...

So I don’t have comments so much as questions.

1. How is the NYCDOE going to handle the timing of moving waitlists?  Many vacancies don’t become visible until just before (or just after) the official start of school, which means that there could be some complications right around that time, for families and schools.
2. How will students on multiple waitlists be dealt with?  Suppose a student waitlisted at multiple schools is admitted off one of them in the summer—if he or she accepts that new assignment, do his/her other waitlist positions remain?  
1. (If other waitlists have to be given up, this could be a complicated decision whether to accept a somewhat preferred school, or wait for an even more preferred one…  If other waitlist positions can be maintained, then the process may move slowly, as some students accept for one waitlisted position, and then a better one when it becomes available, and maybe another…)
3. How long will a student have to consider whether to accept a given waitlist position?

As with many questions of market design, the devil is in the details…"

and I added this in replies to followup emails asking for my thoughts on waitlists:

"I’ve always been cautious about waitlists, because some of the questions I asked you just don’t have good answers.  There’s a tension between wanting waitlists to move early and fast—to make planning easy for families and schools, and avoid disruption of the first week(s) of school, and wanting to give students the best chance at the schools they like best…"

"my colleagues and I never recommended waitlists to nyc, back at the turn of the century.:)
We thought it was important to reduce the number of “unmatched” students who had to be assigned to a school over which they  hadn’t had an opportunity to express preferences. This is why we had a second stage of the matching algorithm, in which lists of schools with still available places were disseminated to students unmatched in the first round, so that they could express preferences over these.

Another question about the new system is, how will such students now be assigned?  E.g. they might be assigned to the closest school to their home that has unfilled places.  In what order?  i.e. after some students are assigned this way, some schools will no longer have unfilled places, and students will have to be assigned to other schools.  The things I read today didn’t address that issue, but I gather that these interim assignments of unmatched students, which will turn out to be final assignments for students whose waitlists don’t move enough, will be made without having the students express preferences.

Another question about the waitlists: how will they be ordered?  According to the school priority/preferences that were used in the first round of matching?  Or perhaps unmatched students will be given preference? (that might sound attractive but I think it would be a bad idea, because it might make it seem desirable to be unmatched after the first round, which would interfere with eliciting student preferences altogether….)

My point is not to try to guess what design decisions have been made, but rather that there are lots of important decisions that have to be made to have a working system, and the initial announcements and news reports don’t reveal these. And they will have consequences.  So I hope that the system has been carefully designed."

Why is it hard for securities exchanges to restore price competition (instead of speed competition)?

Many stock exchanges earn rents by giving privileged access to high speed algorithmic traders.  Why doesn't a new exchange enter the market with a design that would privilege price competition over speed competition?  Budish, Lee and Shim have some thoughts on that:

Will the Market Fix the Market?A Theory of Stock Exchange Competition and Innovation
 Eric Budish, Robin S. Lee and John J. Shim
February 27, 2019

Abstract As of early 2019, there are 13 stock exchanges in the U.S., across which over 1 trillion shares ($50 trillion) are traded annually. All 13 exchanges use the continuous limit order book market design, a design that gives rise to latency arbitrage—arbitrage rents from symmetrically observed public information—and the associated high-frequency trading arms race (Budish, Cramton and Shim, 2015). Will the market adopt new market designs that address the negative aspects of high-frequency trading? This paper builds a theoretical model of stock exchange competition to answer this question. Our model, shaped by institutional details of the U.S. equities market, shows that under the status quo market design: (i) trading behavior across the many distinct exchanges is as if there is just a single “synthesized” exchange; (ii) competition among exchanges is fierce on the dimension of traditional trading fees; but (iii) exchanges capture and maintain significant economic rents from the sale of speed technology—arms for the arms race. Using a variety of data, we document seven stylized empirical facts that align with these predictions. We then use the model to examine the private and social incentives for market design innovation. We show that the market design adoption game among exchanges is a repeated prisoner’s dilemma. If an exchange adopts a new market design that eliminates latency arbitrage, it would win share and earn economic rents; perhaps surprisingly, the usual coordination problems associated with getting a new market design off the ground are not central. However, imitation by other exchanges would result in an equilibrium that resembles the status quo with competitive trading fees, but now without the rents from the speed race. We conclude that although the social returns to adoption are large, the private returns are much smaller for an entrant exchange and negative for an incumbent that currently derives rents from the inefficiencies that the new design eliminates. Nevertheless, our analysis does not imply that a market-wide market design mandate is necessary. Rather, it points to a more circumscribed policy response that would tip the balance of incentives and encourage the “market to fix the market.” 

Congestion in resident and fellowship applications and interviews: Plastic Surgery

Medical graduates can enter plastic surgery through a residency immediately upon graduating.  As in many other specialties, there is lots of applying and interviewing before the residency or fellowship Match.  Here's a proposal to limit the number of applications:

Solving Congestion in the Plastic Surgery Match
Molina Burbano, Felipe, B.A.; Yao, Amy, B.S.; Burish, Nikki, M.D.; Ingargiola, Michael, M.D.; Freeman, Matthew, M.D.; Stock, Jeffrey, M.D.; Taub, Peter J., M.D.
Plastic and Reconstructive Surgery: February 2019 - Volume 143 - Issue 2 - p 634–639

"Summary: Plastic and reconstructive surgery is among the most competitive specialties in the residency match. Applicants seeking to maximize their chances of a successful match often submit numerous applications to the National Residency Matching Program. It is not uncommon for those applying to plastic and reconstructive surgery to apply to every program. The high application volume imparts significant time and financial burden for applicants and programs alike. Furthermore, it makes distinguishing between applicants with a genuine interest in a specific program and those who are merely hoping to improve their chances vastly more difficult. The authors sought to characterize trends in the match rate, as the number of integrated plastic and reconstructive surgery programs continues to increase. Furthermore, they reviewed the literature on game theory for possible solutions to residency application congestion. The authors propose the use of the game theory model to explain the observed results and show why an application limit is the most reasonable approach to address this issue.
...
"it is not uncommon for those applying to plastic and reconstructive surgery to apply to every program. In fact, of a total of 73 available training programs in the 2017 season, senior U.S. applicants applied to a median of 70 programs

"Such a high application volume imparts significant time and financial burden for all parties involved. Applicants spend an average of $6073 and up to $15,000 on applications and interview travel.3 Residency program directors, in turn, must review a greater number of applications and conduct additional interview dates. Furthermore, distinguishing between applicants with a genuine interest in a specific program, versus those who are merely hoping to improve their chances, has become vastly more difficult, with some preferred applicants possibly getting overlooked in the process.4

"Weissbart et al. observed the same trend in the urology match and other competitive specialties.
...
"Introducing an application limit would guide applicants to apply to programs based on fit (i.e., where they felt they were viable applicants). Similarly, the reduction in applications would allow programs to focus more carefully on reviewing all received applications. A more thoughtful application process, with increased focus on each individual, will likely produce “better” matches. Further research would need to be conducted on the correct application limit and the possibility that less competitive students could be left unmatched if not allowed to apply as widely.
***********

Plastic surgery was also attractive at Hogwarts, although not to Harry:

Harry Potter and the Resident Match | ZDoggMD.com

Colleges harvest signals of interest in more ways

Congested markets--those in which there are more potential transactions than can be easily processed--promote signaling, and the search for signals, about which transactions to pursue.  College admissions is a famously congested market, particularly since it became easy for students to submit many applications. So all but the most elite colleges have long searched for signals of "demonstrated interest."  The WSJ has an update on how technology is changing that search.

The Data Colleges Collect on Applicants
To determine ‘demonstrated interest,’ some schools are tracking how quickly prospective students open email and whether they click links  By Douglas Belkin

"Enrollment officers at schools including Seton Hall University, Quinnipiac College and Dickinson College know down to the second when prospective students opened an email from the school, how long they spent reading it and whether they clicked through to any links. Boston University knows if prospective students RSVP’d online to an event—and then didn’t show.
...
"At Seton Hall University, in South Orange, N.J., students receive a score between 1 and 100 that reflects their demonstrated interest, said Alyssa McCloud, vice president of enrollment management. The score includes about 80 variables including how long they spent on the school’s website, whether they opened emails and at what point in high school they started looking on the website (the earlier the better).
...
"In 2017, 37% of 493 schools surveyed by the National Association of College Admission Counseling said they consider demonstrated interest to be of moderate importance—on par with teacher recommendations, class rank and extracurricular activities. It carried less weight than grades, class rigor or board scores.
...
"Colleges also have low-tech means to help determine demonstrated interest. Last year, one third of students who applied to American University either visited its Washington, D.C., campus or attended an information session about the school, said Andrea Felder, assistant vice provost for undergraduate admissions. Two thirds of those admitted took part in either the campus tour or offsite information session.
...
"Mary Hinton, a senior at Dickinson College, benefited from demonstrated interested without knowing it. After she toured Dickinson in high school, she sent a thank-you note to her tour guide, at her mother’s suggestion.

Now a tour-guide herself, Ms. Hinton has learned those notes are forwarded from tour guides to admissions officers. Her advice to prospective students about thank-you notes: “Write them. It just takes a minute and it can make a difference.”

Professional line sitters: Giving the gift of time

Have an urgent need to shop Black Friday bargains, but don't like waiting on lines in the cold and dark?  There's a business that will take care of that for you...

Professional Line Sitters Make Up to $35 an Hour — And This Is Their Busiest Time of the Year

Some links if you're too busy to read the story:
Skip the line (STL) inWashington DC
Same Ole Line Dudes in New York City
InLine4You an app for both sides of the market

"The Supreme Court website says security starts admitting people to oral argument sessions at 9:30 a.m., but “visitors may begin lining up on the Front Plaza as early as they feel comfortable” — which sometimes means four days in advance. The average SCOTUS wait time Goff gets tapped for runs about five hours (she charges $35 per hour), with occasional overnight requests for big cases like the travel ban and Masterpiece Cakeshop v. Colorado Civil Rights Commission."

Universal enrollment--embracing both district and charter schools--was once on the agenda in NYC

One cause of congestion in school choice systems is that if some students receive multiple offers of admissions, other students must wait for admission to a school they want, particularly if the system is so decentralized that a student is only discovered to have rejected an admissions offer after he or she doesn't show up for the first week of class. So a lot of the school choice work that Atila Abdulkadiroglu, Parag Pathak, Neil Dorosin and I have done through IIPSC is aimed at 'universal' enrollment systems, in which all schools take part.

This hasn't happened yet in NYC. So it is interesting that a lawsuit has brought to light emails which suggest that universal enrollment was as one point seriously considered by the city.

Chalkbeat has the story:

Mayor de Blasio almost proposed a universal enrollment system for district and charter schools, emails show  BY ALEX ZIMMERMAN

"Common — sometimes known as “universal” — enrollment systems exist in cities from Newark to Indianapolis. Backers of the approach argue it can simplify the often complex and time-intensive process required to apply to either district or charter schools in cities that allow parents to choose among both. Streamlining the process can put parents on equal footing instead of allowing those with more time, knowledge or resources from automatically getting a leg up
...
"Common enrollment systems have gained traction in recent years as some cities have embraced a “portfolio model” of schools, a new way of organizing school districts that has developed strong backing. This enrollment approach is central in New Orleans and Denver, which received input from Neil Dorosin, who created and once ran New York City’s high-school application system."

Take a break: avoiding congestion in ant colonies and elsewhere

A recent paper in Science is introduced by the NY Times under this headline:

The Secret to Ant Efficiency Is Idleness

Here's the paper:
Collective clog control: Optimizing traffic flow in confined biological and robophysical excavation
J. Aguilar, D. Monaenkova, V. Linevich, W. Savoie, B. Dutta, H.-S. Kuan,  M. D. Betterton, M. A. D. Goodisman, D. I. Goldman

Here's the abstract of the paper (preceded by another description, perhaps written by a writer for Science(?) meant to be more intelligible...

When fewer workers are more efficient:
"A narrow passageway can easily become clogged or jammed if too much traffic tries to enter at once or there is competition between the flow of traffic in each direction. Aguilar et al. studied the collective excavation observed when ants build their nests. Because of the unequal workload distribution, the optimal excavation rate is achieved when a part of the ant collective is inactive. Numerical simulations and the behavior of robotic ants mimic the behavior of the colony."

Abstract: Groups of interacting active particles, insects, or humans can form clusters that hinder the goals of the collective; therefore, development of robust strategies for control of such clogs is essential, particularly in confined environments. Our biological and robophysical excavation experiments, supported by computational and theoretical models, reveal that digging performance can be robustly optimized within the constraints of narrow tunnels by individual idleness and retreating. Tools from the study of dense particulate ensembles elucidate how idleness reduces the frequency of flow-stopping clogs and how selective retreating reduces cluster dissolution time for the rare clusters that still occur. Our results point to strategies by which dense active matter and swarms can become task capable without sophisticated sensing, planning, and global control of the collective.

Modern congestion pricing, by Cramton, Geddes and Ockenfels

Here's a short paper in the July 31 Nature:
Set road charges in real time to ease traffic
Track vehicles to link tolls with demand and cut congestion, urge Peter Cramton, R. Richard Geddes and Axel Ockenfels

And here's the longer working paper:
Markets for Road Use: Eliminating Congestion through Scheduling, Routing, and Real-Time Road Pricing
Peter Cramton, R. Richard Geddes, and Axel Ockenfels

Their vision:

"Efficient pricing of network capacity is not new. Indeed, wholesale electricity markets have been dynamically priced for over a decade. Communications markets are adopting dynamic pricing today. Efficient pricing of road use, however, has only recently become feasible. Advances in mobile communications make it possible to identify and communicate the location of a vehicle to within one cubic meter—allowing precise measurement of road use. User preferences can be communicated both in advance to determine scheduled transport and in real time to optimize routes based on the latest information. Computer advances also facilitate efficient scheduling and pricing of road use. Consumer apps help road users translate detailed price information into preferred transport plans. Computers also allow an independent system operator to better model demand and adjust prices to eliminate congestion and maximize the total value of road infrastructure. An independent market monitor, distinct from the
operator, observes the market, identifies problems, and suggests solutions. A board governs the market subject to regulatory oversight."

The job market(s) for professional psychologists: history and current issues

Various parts of the psychology job market resolved problems of thickness, congestion and incentives by adopting centralized clearinghouses. For other parts of the job market, those issues persist. Here's a recent article, and some snippets from it regarding this history.

Recruitment and selection in health service psychology postdoctoral training: A review of the history and current issues. 
Bodin, Doug; Schmidt, Joel P.; Lemle, Russell B.; Roper, Brad L.; Goldberg, Robert W.; Hill, Kimberly R.; Perry-Parrish, Carisa; Williams, Sharon E.; Kuemmel, Angela; Siegel, Wayne.
Training and Education in Professional Psychology © 2017 American Psychological Association

2018, Vol. 12, No. 2, 74 – 81

"Pursuing a career in health service psychology involves navigating three broad stages of training: graduate school, doctoral internship, and postdoctoral training. Each stage involves distinct procedures for recruitment and selection. The purpose of this article is to review the history of and current issues involved in the recruitment and selection process for postdoctoral training in health service psychology. In this review, we will discuss the specialty of clinical neuropsychology separately as that specialty has a formal computerized match and therefore faces subsequent challenges that are distinct from, but in some ways mirror, the  faced broadly by health service psychology postdoctoral training programs.
...
"The current computerized matching psychology internship selection process began in 1998–1999, 10 years after an earlier trial, rejection, and reintroduction (Keilin, 2000). Prior to the introduction of computerized matching, the internship selection process relied on the “uniform notification day” (UND) system for more than 25 years. This process of internship offers and acceptances shortened from an initial offer and acceptance period of 9 days to 4 hours (Keilin, 1998). In 1988–1989, there was a trial period to evaluate the effectiveness of a computer-based algorithm match process. Although the outcomes showed a significant improvement over the UND, there was a lack of full participation by internship sites that appeared to have a detrimental impact on the outcomes observed (Keilin, 1998). Recurrent discussion within the Association of Psychology Postdoctoral and Internship Centers (APPIC) reflected gradually shifting approval, from a vote of 55% to 45% to retain the UND in 1990, to 77% approval to move to a computer-based match system by May 1998 for immediate implementation for the 1998–1999 recruitment year.
...
"There were wide-spread reports of both applicants and programs accepting less preferred choices due to the fear of not securing or filling a position within the UND window. APPIC responded by clarifying the rules with a prohibition against sites soliciting rank information (Constantine & Keilin, 1996). APPIC voted to proceed with a computer-based match process in 1998 using a match algorithm similar to the National Residency Match Program (Roth & Peranson, 1999). The computer-match system transcends difficulties on both sides by providing an equitable solution designed to provide the best possible match for all parties involved (Keilin, 1998). This solution reduces the pressure associated with short decision-making times and has the benefit of reducing the gridlock associated with UND (Roth & Xing, 1997).
...
"The computerized match has been used every year since, and it is now widely considered the optimal process for doctoral internship selection.

"The recruitment and selection process at the postdoctoral level remains in an evolutionary stage at the present time. Coordination of the psychology postdoctoral selection market is complicated by the fact that it is actually made up of multiple submarkets. Psychology interns may consider several options for their next professional experience, including an entry-level job, a research fellowship, a general clinical psychology fellowship, or a postdoctoral fellowship in a specialty practice area (e.g., neuropsychology, health psychology, clinical child psychology, or forensic psychology). Several of these submarkets have different timelines. For example, entry-level jobs and research fellowships are dependent on idiosyncratic funding streams and employment needs and, as such, may not realistically be expected to adhere to a coordinated timeline. 
...
"In 2003, there was renewed effort to establish more systemization to protect applicants from being forced into making decisions about early offers that did not reflect their true preferences. 
...
"As part of a 2009 APPIC biennial conference presentation, Lemle proposed a modified selection process with an additional feature intended to make participation more attractive to programs. The additional feature would allow applicants who received an early offer from a nonparticipating site to contact a preferred program to pledge they would accept a “reciprocal offer” should one be extended. A UND with reciprocal offer option (UNDr) assured programs that their top applicants who also wished to come to their site would not be lost to competing offers.
...
" Beginning in Fall, 2016, the APPIC postdoctoral workgroup took strides to address several proximal recommendations borne in the summit. First, the workgroup further refined the Postdoctoral Selection Guidelines (PSG) for positions starting in 2017–2018 as a step toward developing a clear and transparent system that would allow applicants sufficient time to consider offers (applicants were allowed 24 hr to consider offers) and allow programs the time needed to interview applicants. A primary aim of the PSG is to decrease the significant stress on both applicants and training directors. The PSG address multiple stages of the selection process impacting trainees including reasonable interview notification timing, consideration of remote interview formats, postponing offers until the UND, and proper use of the reciprocal offer option (UNDr).
...
"The APPIC postdoctoral workgroup also followed the summit recommendation to spearhead the development of a centralized listing of postdoctoral positions, the Universal Psychology Postdoctoral Directory (UPPD), that is free for both APPIC and non-APPIC member programs. 
...
"Clinical neuropsychology has a distinct history that is informative when considering the issues discussed in this article. Most notably, there is currently a formal computerized match for postdoctoral selection in clinical neuropsychology (Belanger et al., 2013; Bodin, Roper, O’Toole, & Haines, 2016). Prior to 1994, there was no organized process for recruitment and selection in clinical neuropsychology. 
...
"In 2001, APPCN partnered with NMS to conduct a computerized match for postdoctoral selection in clinical neuropsychology that has continued to the present day. The APPCN match employs the same algorithm used for the psychology internship match, but dissimilarities across the two recruiting environments have contributed to differences in the success of the two systems over the years. Since its inception, the internship match has enjoyed virtually universal participation by programs. Importantly, the APPCN match was introduced without a clear understanding of the total universe of positions being offered by neuropsychology postdoctoral programs, and those outside of APPCN were not a part of an organized group interested in centralized recruitment. Years later, Belanger et al. (2013) estimated that roughly two thirds of available positions were offered in the APPCN match. Withdrawal rates for applicants initially registering for the APPCN match have ranged from 26% to 37% over the past 10 years and have been stable from 34 and 36% for the last five years. Although some applicants withdraw for personal reasons, the most common reason has been to accept offers from independent programs prior to the rank-order list (ROL) deadline for the APPCN match.

Too many traffic jams: an interview with Stanford's Mike Ostrovsky (and a paper by Ostrovsky and Schwarz)

Mike Ostrovsky at Stanford GSB is interviewed on technology and traffic:
An End to Traffic Jams? It Might Not Be a Dream

And here's the paper on which the interview is based:

Carpooling and the Economics of Self-Driving Cars
Michael Ostrovsky and Michael Schwarz
February 12, 2018

Abstract: We study the interplay between autonomous transportation, carpooling, and road pricing.We discuss how improvements in these technologies, and interactions among them, will affect transportation markets. Our main results show how to achieve socially efficient outcomes in such markets, taking into account the costs of driving, road capacity, and commuter preferences.  An important  component  of  the  efficient  outcome  is  the  socially  optimal  matching  of  carpooling riders.  Our approach shows how to set road prices and how to share the costs of driving and tolls among carpooling riders in a way that implements the efficient outcome

Colleges that get a lot of applications read each one quickly (WSJ)

How to deal with congestion?  Move fast...  Here's the story from the WSJ:

Some Elite Colleges Review an Application in 8 Minutes (or Less)
With so many applying, fewer schools have one person read a whole application; plowing through 500 files in a day

"As application numbers surge, admissions officers at some elite colleges say they don’t have time to read an entire file.

"Instead, staffers from more schools—including the Georgia Institute of Technology, Rice University and Bucknell University in Pennsylvania—now divvy up individual applications. One person might review transcripts, test scores and counselor recommendations, while the other handles extracurricular activities and essays.

"They read through their portions simultaneously, discuss their impressions about a candidate’s qualifications, flag some for admission or rejection, and move on. While their decision isn’t always final, in many cases theirs are the last eyes to look at the application itself.

"The entire process can take less than eight minutes.
...
"Efficiency is crucial, since more students are using the Common Application, which allows them to submit material to multiple schools. Nearly 902,000 students used it last year. As of Jan. 15 this year, the number was already 898,000 students submitting to an average of 4.8 schools.

"Applications to Georgia Tech jumped by 13% for the coming academic year, to 35,600. The current freshman class has roughly 2,800 students."

Evolution of dating apps: OK Cupid requires mutual consent for messaging

Slate reports (last month):
OkCupid Users Have Long Had to Put Up With Unwanted Messages. That's About to Change
"OkCupid announced a big change to its messaging system in an email to users on Friday afternoon. Starting next week, the email said, “Only the people you like or have responded to will remain in your messages. Messages from people you're not interested in, or people you haven't liked yet, will be moved to their profile.”
"This shift will bring the platform more in line with other online dating platforms, such as Tinder and Bumble, on which users can’t message one another at all until both have shown interest in the other. The new OkCupid way won’t be quite so strict—users can still send messages to whomever they want, though those messages will only appear in the recipients’ mailboxes if they indicate that they like the sender—but engineers hope it will help seed more connections while filtering out some messages that will never get a response."