Monday, April 30, 2012

Who gets what in medical resident matching

Here's a paper that reports on the many residency positions filled outside of the match, in large part by graduates of foreign medical schools. "In 2007, 1 in 5 positions in primary care was offered outside the match."

    Out-of-Match Residency Offers: The Possible Extent and Implications of Prematching in Graduate Medical Education.
    Wetz RV, Seelig CB, Khoueiry G, Weiserbs KF, J Grad Med Educ; 2010 Sep;2(3):327-33 

    Another paper reports that students from higher prestige American schools match disproportionately into "controllable lifestyle" specialties ("anesthesiology, dermatology, emergency medicine, neurology, ophthalmology, otolaryngology, pathology, psychiatry, and radiology") ...


    And here's my paper on the history of the match...

    Roth, Alvin E., "The origins, history, and design of the resident match, JAMA. Journal of the American Medical Association, vol. 289, No. 7, February 19, 2003, 909-912.

Sunday, April 29, 2012

Home exchanges

In the WSJ, a retired California couple discuss the (approaching three dozen) home exchanges they have done, bartering a stay at their San Diego condo for visits of a few weeks to a month all over the world, via two networks, and

"Exchanging homes involves some negotiating. Your leverage depends on the desirability of the home you have to exchange. In the U.S., New York, San Francisco and Washington, D.C., are favorites of international exchangers. After that, there is a second tier of cities, which includes San Diego.

"(One tricky area in negotiations: whether to exchange cars. Our preference is to rely on local transportation. Some cities, like Paris and London, have great public transportation. Unfortunately, San Diego doesn't, and our exchange partners feel strongly that we need to exchange cars. We've done so about half the time.)
"On each Internet site, we have a page with photos of our home, information about San Diego and ourselves, and details about when, where and with whom we want to exchange. Once we make contact with a potential partner, we discuss details of dates, number of people, transportation, etc. On average, an exchange takes about 20 emails and an occasional phone call to work out all the arrangements.
"We are often asked if we have any problems with our exchanges. We do. The biggest problem is cancellations. Since our first choice is to exchange with other older adults, unexpected medical problems can be an issue. With younger exchangers, you can encounter job or financial problems. In a few cases, my guess is that people cancel because they get a better offer—but they never admit that.

"We have had seven cancellations in connection with our 32 exchanges. Fortunately, all except one were early enough that we hadn't made plane reservations and could arrange alternative exchanges."

Saturday, April 28, 2012

The OPTN Kidney Paired Donation Program --progress report, April 2012

Since opening for business in late 2010, the UNOS/OPTN kidney exchange pilot program has succeeded in enrolling many transplant centers. But so far it has not completed many transplants compared to other kidney exchange networks. (What to do about this was one subject of the recent conference I blogged about here: Competition among kidney exchanges.)

Here's the latest report from UNOS: Everything You Wanted to Know about the OPTN Kidney Paired Donation Program
"The OPTN kidney paired donation pilot program (KPDPP) continues to expand; as of mid-April, 114 transplant centers are participatingview list. The goal is to increase the number of living donor kidney transplants, so center participation is crucial to the success of the national KPD program. Results from recent match runs indicate an upward trend in the number of donors and candidates matched. In the match run conducted in March, 24 matches of compatible potential donors and recipients were found from a pool of 163 donors and 145 candidates, about twice as many matches as the previous month. So far, 19 kidney candidates have received a transplant in the program as of mid-April. "

The page at the link contains a link to some background reports, including the one below which (together with the comments) gives some insights into the kind of debates about kidney exchange policies that have been going on at the program:
And here is some of the evidence that long chains are good for kidney patients, particularly for highly sensitized patients:

Friday, April 27, 2012

Jacob Leshno defends his Ph.D. dissertation

Defense 4

Dr. Jacob Leshno (in suit:), and Al Roth, Susan Athey, and Ariel Pakes (Drew Fudenberg on skype)
And here is what the picture would have looked like if Drew had used photoshop instead of skype:
Drew looks on approvingly (in a photo doctored by the new doctor...)

The title of Jacob's dissertation is Essays in Market Design.  (How cool is that?)

The three papers he chose to include in his dissertation are

In many assignment problems items arrive stochastically over time. When items are scarce agents form an overloaded waiting list and items are dynamically allocated as they arrive; two examples are public housing and organs for transplant. Even when all the scarce items are allocated, there is the efficiency question of how to assign the right items to the right agents. I develop a model in which impatient agents with heterogeneous preferences wait to be assigned scarce heterogeneous items that arrive stochastically over time. Social welfare is maximized by appropriately matching agents to items, but an individual impatient agent may misreport her preferences to receive an earlier mismatched item. To incentivize an agent to avoid mismatch, the policy needs to provide the agent with a (stochastic) guarantee of future assignment, which I model as putting the agents in a priority buffer-queue. I first consider a standard queue-based allocation policy and derive its welfare properties. To determine the optimal policy, I formulate the dynamic assignment problem as a dynamic mechanism design problem without transfers. The resulting optimal incentive compatible policy uses a buffer-queue of a new queueing policy, the uniform wait queue, to minimize the probability of mismatching agents. Finally, I derive a robustly optimal policy which uses a simple rule: giving equal priority to every agent who declines a mismatched item (a SIRO buffer-queue). This robustly optimal policy has several good properties that make it a compelling market design policy recommendation.
A Supply and Demand Framework for Two-Sided Matching Markets, with Eduardo Azevedo
(Extended abstract published in EC11 under the former name: "The college admissions problem with a continuum of students" )
We give a version of the Gale and Shapley (1962) college admissions problem where colleges have a large capacity and show that the resulting model allows for tractable analysis of matching markets. When colleges are large stable matchings can be described concisely by cutoffs, the admission thresholds at each college. Under broad conditions the model corresponds to the limit of large discrete matching problems.

Will a Decrease In The Minimum Wage Improve Training?, with Michael Schwarz
We show that firms that have a cost advantage in providing training can recoup training cost even in an almost frictionless labor market. Lowering the minimal legal wage can reduce the efficiency of training and harm welfare. In our model training contracts give positive surplus to workers that is not competed away. This explains the existence of intermediary services that essentially sell internship positions to college graduates.
Jacob is one of the group of job market candidates I blogged about here: Five Harvard candidates for the Economics job market this year (2011-12)

He will be going next year to a postdoc at Microsoft Research in Cambridge, after which he'll take up a position at Columbia GSB.

This concludes my defenses for the week (and the Offense never even got a point up on the board...)

Welcome to the club, Jacob.

Thursday, April 26, 2012

Eduardo Azevedo defends his Ph.D. dissertation

Defense 3

Eduardo Azevedo (in suit:) having just fended off his committee: from left, Eric Budish, Al Roth, Oliver Hart, Susan Athey, Andrei Shleifer

Eduardo chose the following three of his papers to constitute his dissertation:

A Supply and Demand Framework for Two-Sided Matching Markets (Job Market Paper #1)
with Jacob Leshno
extended abstract published in EC11

Evolutionary Origins of the Endowment Effect - Evidence from Hunter-Gatherers (PDF available upon request)
with Coren ApicellaNicholas Christakis, and James Fowler

I earlier blogged about two of those papers:

A supply and demand model for stable matchings, by Eduardo Azevedo and Jacob Leshno

Market design in a future of trusted smart markets: paper by Eduardo Azevedo and Eric Budish

Eduardo is one of the group of job market candidates I blogged about here: Five Harvard candidates for the Economics job market this year (2011-12)

As I write it isn't clear whether he'll be working next year in Philadelphia, NYC, or Chicago, which will depend on his fiance's jobmarket, which is still to be concluded.

Welcome to the club, Eduardo.
Update: May 11--It's Wharton.

Wednesday, April 25, 2012

Boston kindergartens and preschool places

Boston is short of good public kindergarten places. This in turn creates a demand for preschool places, since preschool kids move up automatically to kindergarten in the same school, so even parents who wouldn't ordinarily send their kid to preschool may be tempted to do so since it gives them two chances at a good kindergarten.

Rise in kindergarten demand leaves many in city scrambling: Hundreds of students remain unassigned

"Demand for kindergarten seats in the Boston Public Schools for this fall has risen by more than 25 percent, an unanticipated increase that has left hundreds of students without an assigned school and has prompted officials to add more classrooms.
The enrollment boom surfaced in the past few months during the first round of registration for kindergarten classrooms that will serve students who will be 5 by Sept. 1. The School Department received 2,306 such applications, up from 1,823 during that same period last year.
"Predicting kindergarten enrollment in a city as large as Boston can be a tricky endeavor, school officials say. The transient nature of the city - with many young, middle-class families moving out and a number of immigrant families moving in - creates volatility in relying on birth rates.
"The city school system also faces immense competition for kindergarten students from dozens of private and parochial schools and a growing number of independently run public charter schools. City school officials often do not know until after the school year begins if all the kindergarten students offered a seat will show up
"The School Department is trying to respond to the rising demand by adding a kindergarten classroom at five schools: Umana Academy in East Boston, Harvard-Kent in Charlestown, Mission Hill in Jamaica Plain, and Haley Elementary and Sumner Elementary, both in Roslindale.

"Kindergarten is not the only grade experiencing a rise in applications. The city’s preschool program for 4-year-olds also has an increase, with 2,518 applications filed during the first registration cycle, compared with 2,070 during the same period last year. That has left 745 4-year-olds without preschool assignments, an increase of 513 from last year.
The city is not obligated under state law to make a seat available for all preschool students who apply, as it is with kindergarten.
"Margaret Day’s 4-year-old son is near the bottom of waiting lists at three schools for preschool, leaving little chance of admittance. Now, the Jamaica Plain mother is resigned to going through the lottery again next year for kindergarten, even though many seats will be snatched up as preschoolers move up. She said she does not understand how the School Department was “blindsided’’ about the enrollment increase, and is pushing for changes.
“We are going to be picking through the bones to get a good seat next year,’’ Day said. “The reason we went through the [preschool] lottery is because of the difficulty of getting into kindergarten.’’

Tuesday, April 24, 2012

Game theory in Operations Research

The venerable journal Operations Research has a new subject area, with distinguished area editors:

David Gamarnik and Asu Ozdaglar, editors

Networks have become pervasive in every aspect of our lives and have emerged as a crucial tool for understanding the world we live in. The World Wide Web, which links us to and enables information flows with the rest of the world, is the most visible example. Physical networks play a similarly vital role in wireless communication, data transmission, and energy production and distribution. Equally central are social and economic networks that shape opinions, information flows, and transactions. A fundamental understanding of network behavior, including the nature of interconnections, issues of stability, and decision making place operations research methodology at the core of this emerging research program.
We invite high-quality submissions that advance the theory and application of operations research methods, including graph theory and optimization, economic analysis, game theory, and stochastic models, in the emerging field of network science. In addition to works contributing to the analysis of networks, we also encourage submissions using tools and concepts from network analysis in other application domains.
Associate Editors: Mung Chiang, Ayalvadi Ganesh, Ali Jadbabaie, Ramesh Johari, Andrea Montanari, Yannis Paschalidis, Devavrat Shah, Sanjay Shakkottai, and Adam Wierman

I'm hopeful that this may mark a return to a high level of interaction between game theorists and the operations research community. (Market design seems like an area that is ripe for this kind of interaction.)

Yuichiro Kamada defends his Ph.D. dissertation

Defense 2 (Offense 0)

Yuichiro (in suit:), with Tomasz Strzalecki and Al Roth (and Drew Fudenberg and Attila Ambrus via skype)
Yuichiro with the full defense team

Yuichiro had to choose three out of his many papers for his dissertation, which he called "Essays on Revision Games." Those papers all concern the difficult problem of analyzing incentives in non-stationary environments.

Multi-Agent Search with Deadline (joint with Nozomu Muto), December 31, 2011.(An earlier version of this paper referred to in the new version is here)

Revision Games (joint with Michihiro Kandori), December 31, 2011

Asynchronicity and Coordination in Common and Opposing Interest Games (joint with Riccardo CalcagnoStefano Lovo, and Takuo Sugaya), March 2012, Revise and Resubmit, Theoretical Economics(This paper is a result of a merger between two independent papers: Preopening and Equilibrium Selection by Calcagno and Lovo, and Asynchronous Revision Games with Deadline: Unique Equilibrium in Coordination Games by Kamada and Sugaya)

I earlier blogged about another of his papers, on the design of the market for new Japanese doctors:

Matching Japanese Doctors: problems with the current mechanisms, and suggestions for improvement by Yuichiro Kamada and Fuhito Kojima

Yuichiro is one of the group of job market candidates I blogged about here: Five Harvard candidates for the Economics job market this year (2011-12)

 He will be going next year to a postdoc at Yale, after which he'll take up a position at Berkeley-Haas.

Two more defenses are coming up this week.

 Welcome to the club, Yuichiro.

Monday, April 23, 2012

Katie Baldiga defends her Ph.D. dissertation

Defense 1 (Offense 0)

Katie Baldiga and her committee of admirers: Iris Bohnet, Al Roth, Jerry Green (and Drew Fudenberg via skype)
(Alternative caption: Katie B. got her Ph.D.)

The three papers in Katie's dissertation are so wide ranging that she characterizes them together as "Essays in Microeconomics." (I blogged about her experimental paper here.)

A Failure of Representative Democracy (Job Market Paper 1) 
In this paper, we study representative democracy, one of the most popular classes of collective decision-making mechanisms, and contrast it with direct democracy. In a direct democracy, individuals have the opportunity to vote over the alternatives in every choice problem the population faces. In a representative democracy, the population commits to a candidate ex ante who will then make choices on its behalf. While direct democracy is normatively appealing, representative democracy is the far more common institution because of its practical advantages. The key question, then, is whether representative democracy succeeds in implementing the choices that the group would make under direct democracy. We find that, in general, it does not. We analyze the theoretical setting in which the two methods are most likely to lead to the same choices, minimizing potential sources of distortion. We model a population as a distribution of voters with strict preferences over a finite set of alternatives and a candidate as an ordering of those alternatives that serves as a binding, contingent plan of action. We focus on the case where the direct democracy choices of the population are consistent with an ordering of the alternatives. We show that even in this case, where the normative recommendation of direct democracy is clear, representative democracy may not elect the candidate with this ordering.
Gender Differences in Willingness to Guess and the Implications for Test Scores (Job Market Paper 2) 
Multiple-choice tests play a large role in determining academic and professional outcomes. Performance on these tests hinges not only on a test-taker's knowledge of the material but also on his willingness to guess when unsure about the answer. In this paper, we present the results of an experiment that explores whether women skip more questions than men. The experimental test consists of practice questions from the World History and U.S. History SAT II subject tests; we vary the size of the penalty imposed for a wrong answer and the salience of the evaluative nature of the task. We find that when no penalty is assessed for a wrong answer, all test-takers answer every question. But, when there is a small penalty for wrong answers and the task is explicitly framed as an SAT, women answer significantly fewer questions than men. We see no differences in knowledge of the material or confidence in these test-takers, and differences in risk preferences fail to explain all of the observed gap. Because the gender gap exists only when the task is framed as an SAT, we argue that differences in competitive attitudes may drive the gender differences we observe. Finally, we show that, conditional on their knowledge of the material, test-takers who skip questions do significantly worse on our experimental test, putting women and more risk averse test-takers at a disadvantage.
Assent-Maximizing Social Choice with Jerry R. Green, forthcoming in Social Choice and Welfare
We take a decision theoretic approach to the classic social choice problem, using data on the frequency of choice problems to compute social choice functions. We define a family of social choice rules that depend on the population's preferences and on the probability distribution over the sets of feasible alternatives that the society will face. Our methods generalize the well-known Kemeny Rule. In the Kemeny Rule it is known a priori that the subset of feasible alternatives will be a pair. We define a distinct social choice function for each distribution over the feasible subsets. Our rules can be interpreted as distance minimization -- selecting the order closest to the population's preferences, using a metric on the orders that reflects the distribution over the possible feasible sets. The distance is the probability that two orders will disagree about the optimal choice from a randomly selected available set. We provide an algorithmic method to compute these metrics in the case where the probability of a given feasible set is a function only of its cardinality.

Katie is one of the group of job market candidates I blogged about here: Five Harvard candidates for the Economics job market this year (2011-12).

She and her significant other LC solved the two-body problem this year (!), and will be together at The Ohio State University, which is now more than ever a hotbed of experimental economics.

Three more defenses are coming up this week.

Welcome to the club, Katie.

Sunday, April 22, 2012

How to evaluate school choice: Daily Telegraph

The Daily Telegraph has an article about school choice in Britain that uses what strikes me as a tricky standard to declare that something is rotten in Britain:

Children 'forced to accept unpopular secondary schools': Almost 75,000 children have been rejected from their preferred secondary school amid a desperate scramble for the most sought-after places, official figures show.

The article goes on to say
"More than one-in-seven pupils across England are being forced to accept second, third or fourth-choice schools this September, it emerged.
"According to figures, some schools in parts of London received as many as nine applications for every place.

"Mr Gibb said: “I want us to reach a position where it is parents choosing schools, not schools choosing parents."
And here's a similar article about younger kids: More children rejected from first-choice primary school

"Among councils that provided year-on-year figures, some 90 per cent reported an increase in applications in 2012 compared with 2011.
"In those areas, almost 14 per cent of four and five-year-olds failed to get into their first choice school
Britain would not be alone in having a shortage of good schools, but the point I want to raise about these articles is that the statistics mentioned could instead indicate that Britain has a few remarkably good schools that are oversubscribed. That would be something quite different, but it would still mean that many students didn't get one of their top choices.

Saturday, April 21, 2012

Matching Mechanisms, Market Design Seen as Tools to Address Inequality

A conference in Chicago, this past February, organized by Scott Kominers:
Here's the press release-- --and here's the list of talks:


Download schedule» (.pdf)

9–9:40 a.m. "Consumption and Income Inequality in the U.S. Since the 1960s"
Bruce D. Meyer (joint work with James X. Sullivan)
9:40–10:20 a.m. "Dismantling the Legacy of Caste: Affirmative Action in Indian Higher Education"
Dennis Epple (joint work with Surendrakumar Bagde and Lowell J. Taylor)
10:20–10:35 a.m. Break
10:35–11:15 a.m. "Promoting School Competition Through School Choice: A Market Design Approach"
John William Hatfield (joint work with Fuhito Kojima and Yusuke Narita)
11:15–1:55 a.m. "School Admissions Reform in Chicago and England: Comparing Mechanisms by Their Vulnerability to Manipulation"
Tayfun Sonmez (joint work with Parag Pathak)

1–2 p.m. Invited Address:
"Matching with Evolving Human Capital"
Lones Smith (joint work with Axel Anderson)
2:00–2:15 p.m. Break
2:15–2:55 p.m. "Effective Affirmative Action in School Choice" M. Bumin Yenmez (joint work with Isa E. Hafalir and Muhammed A. Yildirim)
2:55–3:35 p.m. "Matching with Slot-Specific Priorities: Theory and Applications" Scott Duke Kominers (joint work with Tayfun Sonmez)
3:35–3:50 p.m. Break
3:50–4:30 p.m. "Do Parents Choose under the Boston Mechanism?" Caterina Calsamiglia (joint work with Maia Guell)
4:30–-5:10 p.m. "Homophily and Transitivity in Dynamic Network Formation"
Bryan Graham 

Friday, April 20, 2012

Online labor markets

John Horton, who has an online labor blog, has compiled a great list of online labor markets. Here's what it looked like when I recently checked:


Major General / Programming Sites

  • oDesk is a rapidly growing site that encourages buyers to offer work on hourly terms. The site provides very robust monitoring tools: workers download software that tracks the time that they spend working on projects. The software also logs workers' keystrokes and even periodically takes screenshots of their screens. Because this monitoring allows buyers to observe workers so closely. Workers are guaranteed payment for their hourly work. 
  • allows buyers to post jobs and solicit bids from workers. 
  • Elance is probably the largest and most well-known general site. The main work categories are Web & Programming, Design & Multimedia, Writing & Translation, Administration Support, Sales & Marketing, Finance & Management, Engineering & Manufacturing and Legal. 
  • focuses on computer programming. Programmers work on a strictly fixed-price basis. After a project is awarded to a bidder, the buyer places the agreed upon amount in an escrow account. Payment is not released until the buyer approves the work. If there are disputes, the company provides arbitrators.
  • iFreelance is a general site with the same basic model as Guru, oDesk and Elance. Very little data is available without registering with the site.
  • ScriptLance
  • Solvate - TBD 
  • Hypios - TBD

Software Testing

  • uTest provides software testing using a pool of workers who are paid to find bugs. Buyers (software companies) seek testers with certain demographic characteristics, including software sophistication and the platforms used. uTest selects matches from its pool of registered testers. The testers report bugs and make suggestions and are paid for verified bugs and helpful reviews.


  • 99 Designs allows buyers to sponsor design contests, usually for logos or specialty events. The buyer states a price that they will pay for the winning design and a date when they will choose the winner. Workers then submit designs as small, thumbnail images which the entire community can view and vote on (these votes are not in any way binding). One interesting feature of this market is that it is one of the few real-life examples of non-political all-pay auctions. As of January 2009, the community had 25,100 registered designers who submitted 1,152,786 designs for 15,993 contests and received $3,465,314 in prizes.
  • is just like 99 Designs, but for tattoos.
  • CrowdSPRING
  • iStockPhoto is a royalty-free photo exchange that allows photographers to upload images and sell them to individual buyers. Over 50,000 photographers have contributed over 4 million photos.

Advice & Search

  • BitWine provides a network of freelance advisors who charge clients a per-minute rate for consultations. Advisors are ostensible experts in some field, such as nutrition, travel, coaching, technology, etc. Consultations occur over the voice-over-IP (VoIP) technology Skype and payments are made through PayPal. BitWine takes a percentage of transactions fees.
  • InnoCentive
  • ChaCha



  • Amazon's Mechanical Turk is unique in allowing workers to perform simple, piece-rate tasks like captioning writing captions for photographs, extracting information from scanned documents and transcribing audio clips for money or store credit. Unlike other sites, workers may begin a task as soon as it is posted, and the buyer reserves the right to accept or reject work.
  • LiveWork
  • Short Task
  • Clickworker

Patent Validation / Invalidation

  • Article One Partners is a private firm that helps corporate clients invalidate patents. A network of "AOP Associates" who have registered with the site can earn up to $50,000 for providing evidence of patent-invalidating "prior art."
  • Peer-to-Peer Patent Project also provides patent assessment, but it is not a true contest or market. The project is sponsored by the US Technology and Patent Office and NYU Law School and patent reviewers are not paid.


  • LendingClub facilitates unsecured lending between individuals.
  • Kiva
  • is a for-profit peer-to-peer lending site similar to LendingClub. As of January 2009, the site claims 830,000 members and $178,000,000 of loans.

Enabling Techologies and Organizations

Interested Third Parties

  • Turk Opticon is a website and Firefox plug-in that allows Amazon Mechanical Turk workers to rate "Requestors" who post work on the site. The Firefox plug-in allows workers to view the ratings in real time before deciding whether or not to accept a task from the requestor.

Thursday, April 19, 2012

College admissions committees and how they work

Inside Higher Ed has a story: How They Really Get In, based on surveys of selective colleges and universities conducted by by Rachel B. Rubin, a doctoral student in education at Harvard.

"Almost all of the colleges that provided information first do a winnowing of one of two sorts that yields the group that gets a more thorough review. The most common winnowing process (used by 76 percent of the colleges that answered Rubin) is some measure of academic merit. This may be based on grades, rigor of high school courses, test scores and so forth. While there is some difference in the relative weight given to various factors, there is a straightforward value on doing better than others in whatever formula the college uses.

"A minority of elite colleges and universities (21 percent) starts off on measures of "institutional fit." These colleges do the initial cut based on student essays, recommendations and specific questions of whether particular students will thrive at and contribute to the college in various ways. In an interview, Rubin said she believed that these colleges also valued academic merit, but that the vast majority of applicants had an appropriate level of academic merit, so that could be weighed later, while other parts of "creating a class" needed to dominate at the point of first cut.
"For those colleges that look at institutional fit first, the two most favored factors are underrepresented minority status and "exceptional talent" (which she said could mean many things: "lacrosse recruits, flautists, etc.").
"Rubin's paper says that whether the first cut is done through academics or fit, most colleges then report a more formal system in which two readers review the application portfolio, with a third reader or an entire team involved in difficult calls. At this stage, academic issues are discussed at institutions that started with "fit" issues, and "fit" is discussed at places that started with academics. "

Wednesday, April 18, 2012

Long college waiting lists are long shots...

In case you didn't know: Colleges' Tough Waiting Game: Schools Keep Hundreds of Applicants on Reserve Lists, but Very Few of Them Get In

"Carnegie Mellon University in Pittsburgh admitted just six of 5,003 applicants invited onto its waitlist last year. At Cornell University, in Ithaca, N.Y., not one of 2,998 students offered a spot on last year's list was admitted. Princeton, which accepted 2,095 students for a record-low 7.86% admission rate this year, offered 1,472 applicants places on its waitlist. In the past six years, it has taken as few as zero from the list, or as many as 164.

"Many colleges are reluctant to disclose the number of students on their waitlists. Harvard, which admitted a record-low 5.9% of applicants this year, doesn't release the size of its list. A Harvard spokesman said it accepted 31 from the waitlist last year, and between 49 and 228 in the four years prior to that."

"Most schools know by May 1 who has accepted their initial offers of admission. They then turn to the waitlist to fill any remaining slots, a process that is supposed to wrap up by Aug. 1, a deadline set by the National Association for College Admission Counseling. On average, 45% of students offered spots on a waitlist accept, said the Princeton Review."

Tuesday, April 17, 2012

School Choice in New Orleans with top trading cycles

Something exciting is happening in New Orleans Recovery School District: a novel school choice algorithm is being tried out for the first time. And it's being rolled out with some pretty good communications to the community. Here's a story by Andrew Vanacore in The Times-Picayune yesterday: Centralized enrollment in Recovery School District gets first tryout

The newspaper story even comes with a graphic meant to convey an important piece of the top trading cycle algorithm:

"Somewhere inside the jumble of narrow beige corridors on Poland Avenue in the Upper 9th Ward, in the state-run Recovery School District's headquarters, someone will hit a button on a computer in the next few weeks, and presto: Almost all of the roughly 28,000 students in New Orleans who applied for a seat this fall at one of the district's 67 schools will be assigned a place. With that final keystroke, the school system will move for the first time from a frustrating, ad-hoc enrollment process handled at individual campuses to one centralized at district offices.

"Every child will get a spot. But because schools don't have an infinite number of seats and some schools will doubtless prove more popular than others, not every family will get their top pick. More than likely, some small percentage won't get any of the eight choices they ranked when they filled out their application earlier this year.

"So it's worth considering: Who will decide the fates of these 28,000 students? In a district unique for giving children the right to pick the school they attend, instead of the right to attend the school down the street, who decides what's fair?
"The short answer is nobody, or at least nobody with a pulse. As the technical experts working with the Recovery District explained to a group of community groups and reporters last week, a computer algorithm will go to work trying to match every student with a school that's as high on their list as possible. Students with siblings at a particular school will get a first shot at open seats at that school, followed by those living nearby. But much will hinge on a randomly assigned lottery number.
"This, officials say, will give the greatest possible number of students as close to their top choice as possible in a way that's fair and transparent.
"For parents, there are some key ideas to keep in mind. The experts who developed the algorithm -- folks from Duke, Harvard and MIT -- say there is no way to game the system. If what you really want is a seat at KIPP Renaissance High School, you should not rank Sci Academy first, thinking that you're more likely to get your second choice. Ranking KIPP as your top choice gives you your best shot at getting in. 
"There's also no way to lose a seat that your child already has. If a second-grader at School A applies for a transfer to School B and doesn't win a spot, he will automatically remain where he is.
"For now, there's no way of telling which percentage of families will get their top choice, but a small number will almost certainly get none of the eight schools they ranked on their application. That's simply because they were unlucky in the assigning of lottery numbers, picked eight very popular schools or both.
"Until all of our schools are equally amazing, not every family is going to get everything they want," said Neil Dorosin, founder of the Institute for Innovation in Public School Choice, which helped develop the new process."
"Another important caveat: The 17 schools that are still overseen by the Orleans Parish School Board will not be included. "That's the missing link, " said Debra Vaughan, a researcher at the Cowen Institute. It means students who want to reach for a spot at one of the city's prestigious magnet schools still have to navigate two separate enrollment systems. The Recovery District and the School Board have been in talks but haven't reached an agreement on how to merge the systems."
Previous posts on New Orleans schools (most recent first):

New Orleans launches its new school choice process

Monday, April 16, 2012

How to finance kidney exchange?

In kidney exchange (also called kidney paired donation), potential donors have to be evaluated without knowing what patient, if any, their kidney might go to. should those donor evaluation costs be paid for, since they can't be billed to a particular patient's insurance? (This is particularly so for donors who don't proceed to transplant, e.g. because their costly CT scan reveals an anatomical abnormality: so that's a hard to recover cost.) Below, from Pub Med, is the abstract of a paper proposing a solution like the one reached for deceased donors. (The full paper is still under embargo as it hasn't appeared yet in the print version of the American Journal of Transplantation (but you can see the abstract for free here); I'll post a version when I can...)

Call to Develop a Standard Acquisition Charge Model for Kidney Paired Donation.
Rees MA, Schnitzler MA, Zavala EY, Cutler JA, Roth AE, Irwin FD, Crawford SW, Leichtman AB., Am J Transplant. 2012 Apr 9. doi: 10.1111/j.1600-6143.2012.04034.x. [Epub ahead of print],

We propose a Medicare Demonstration Project to develop a standard acquisition charge for kidney paired donation. A new payment strategy is required because Medicare and commercial insurance companies may not directly pay living donor costs intended to lead to transplantation of a beneficiary of a different insurance provider. Until the 1970s, when organ procurement organizations were empowered to serve as financial intermediaries to pay the upfront recovery expenses for deceased donor kidneys before knowing the identity of the recipient, there existed similar limitations in the recovery and placement of deceased donor organs. Analogous to the recovery of deceased donor kidneys, kidney paired donation requires the evaluation of living donors before identifying their recipient. Tissue typing, crossmatching and transportation of living donors or their kidneys represent additional financial barriers. Finally, the administrative expenses of the organizations that identify and coordinate kidney paired donation transplantation require reimbursement akin to that necessary for organ procurement organizations. To expand access to kidney paired donation for more patients, we propose a model to reimburse paired donation expenses analogous to the proven strategy used for over 30 years to pay for deceased donor solid organ transplantation in America.
© Copyright 2012 The American Society of Transplantation and the American Society of Transplant Surgeons.

Sunday, April 15, 2012

Attacks on the Defense of Marriage Act

The 1996 Defense of Marriage Act, designed to 'defend' marriage from those States that have legalized same sex marriage, is under renewed attacks in the courts. (The Obama administration has indicated that it does not believe the Act is constitutional, so it will not defend it's key provisions, but this doesn't mean that it will die an easy death...)

Mass. leads fight on right to marry
"Massachusetts will once again take center stage in the national debate over same-sex marriage as the state becomes the first to go before a United States appeals court to challenge a federal law that defines marriage as a union only of a man and a woman.

"The US Court of Appeals for the First Circuit in Boston on Wednesday will hear two cases that challenge the 1996 Defense of Marriage Act: The first is a lawsuit brought by 17 local plaintiffs who say it deprives them of the federal benefits that other married couples receive.

"The second, brought by the state, alleges the Marriage Act forces Massachusetts to discriminate against gay couples when the state’s highest court has already declared their marriages constitutionally protected."

Noncitizens Sue Over U.S. Gay Marriage Ban
"Five legally married same-sex couples filed a lawsuit on Monday to challenge the 1996 law that bars the federal government from recognizing  same-sex marriages, arguing that its impact is particularly harsh on couples that include an American citizen and a foreigner.

"The lawsuit, filed in United States District Court for the Eastern District of New York, was brought by Immigration Equality, a gay rights legal organization that focuses on immigration issues. Same-sex marriage advocates said it was likely to become the most prominent suit seeking to overturn the law, known as the Defense of Marriage Act, based on its effect on gay or lesbian immigrants who want to gain legal residence through marriage to American citizens.

"Under immigration law, a citizen can apply for a foreign spouse to obtain legal permanent residency, with a document known as a green card. Since unlike many other visas, there are no limits on the number of green cards available to spouses of citizens, those applications are among the fastest and most straightforward procedures in the immigration system. "Under the marriage act, which is called DOMA, federal authorities do not recognize same-sex marriages, even from states that allow them. In recent years, as same-sex marriage became legal in several states, gay and lesbian couples have come forward to say they were facing a painful choice: either deportation for the immigrant or exile to life in a foreign country for the American. ...

"In February 2011, the Obama administration announced that it regarded the central provision of the marriage act as unconstitutionally discriminatory, and said officials would no longer defend it in the courts.

 "On Wednesday, the United States Court of Appeals for the First Circuit in Boston will hear arguments in the first marriage act case to advance to the appeals level. That case contends that the act is unconstitutional because it denies federal benefits to same-sex couples married in Massachusetts, the first state to make same-sex marriage legal."

"Justice Department officials have said that they will not defend the core provision of the marriage act in that hearing, but will dispute other claims in the case. A conservative legal group appointed by the House of Representatives will argue in favor of the act."

Saturday, April 14, 2012

Market design lecture today

If you're on the Harvard campus this afternoon, perhaps because you are an alum of the Graduate School of Arts and Sciences and are here for Alumni Day, you could hear me talk about market design at 1:15:

1:15-2:45 p.m.
Symposia I 
Who Gets What: The New Economics of Matchmaking and Market Design
Alvin Roth
George Gund Professor of Economics and Business Administration

Emerson 205

Harvard admissions

Some statistics: Harvard Accepts Record Low of 5.9 Percent to the Class of 2016

"In total, including the 772 students admitted in December in the early action acceptance round, a total of 2,032 initial offers, more than 100 fewer than last year, were extended to applicants. The admit rate for those considered under regular decision, including the 2,838 early action candidates who were deferred to the original round, was 3.8 percent.
"Fitzsimmons said in a press release that the return of early action admissions, discontinued at Harvard in 2006 and revived for the Class of 2016, made it more difficult for the admissions office to predict the number of students who will ultimately matriculate. Thus, officers were conservative in the number of students they accepted.
"Harvard placed an unspecified number of students on the waitlist, and these students will receive notice of their status later in the spring. Fitzsimmons said that the office expects to admit more waitlisted students than usual this time around, since it was cautious in its initial offers."

Friday, April 13, 2012

Starting a new university

Inside Higher Ed has an article on the Minerva Project, named after the Roman goddess of wisdom, about an attempt to start a new, elite, for-profit university: An Elite University ... From Scratch?

Some of the ideas sound interesting:
"Minerva plans to define elite differently, he said. Its students -- roughly 200 in the first year, planned for 2014, and it hopes many more in later years -- will be selected through a rigorous two-step process based on academic credentials (grades, test scores, essays, etc.) at a first, statistical level and then -- for those who move on -- an interview process focused on testing an applicant's drive, analytical skills and goals.

"Factors that may help a student get admitted to Princeton or Williams -- athletic skill, alumni connections, money -- won't play any role at Minerva, Nelson said. And because the process won't factor in geography, either, he said he expected that the vast majority of its students will come from outside the United States. "We won't discriminate based on state or country of origin, and the idea that the majority of the smartest English-speaking kids live in the U.S. is absurd," he said. "We're going to have the most racially diverse student body of any elite institution."

"But the real test of its eliteness will come in its curriculum, which Nelson compared to "1950s University of Chicago." It aims to hire top professors to create their own online lectures and course materials, and students will also dig into that material in 25-student interactive seminars led by instructors (Ph.D.s who favor teaching over research, for instance, not grad students). While the formal curriculum will be delivered online, students who choose to will live in dorms in major cities around the world, where they will gain from the same kind of peer encounters that enhance the education at liberal arts and other residential colleges."

That being said, it's a big task to found a new university, let alone an "elite" one.  As I wrote in an earlier post, Rankings of universities: vintages and coordination of expectations, regarding the rankings of American universities:
"the first place goes to the oldest American university, founded in 1636, and spots 2 and 3 go to universities established in 1746 (as the College of New Jersey), and 1701. Two universities that opened more than a century later, in 1861, and 1891 , are tied for 4th place. The two universities tied for 6th are of different vintages, 1891 and 1755, as are the three tied for 8th place, 17541838, and 1892. The top dozen ranks are filled out by universities open for business since 17691855, and 1856.

At number 17, Rice University, opened in 1912, seems to be the highest ranked university on the list to have begun in the 20th century."