Counterproductive incentives in transplantation

Which would be better, to have a transplant center transplant 200 patients and have 150 do well, or to have the center transplant 100 patients and have all of them do well? How about if the 100 patients who didn't receive a transplant in the second scenario would all have died?

In Oregon, The Bend Bulletin has a three part series (pointed out to me by Ben Hippen) on the difficulties of regulating transplant centers, and the sometimes counterproductive incentives that are introduced (in an effort to counteract other bad incentives that hospitals may have).

UPMC, the University of Pittsburgh Medical Center, comes out looking very good, incidentally.


Patients denied transplants as donor organs are discarded
PART 1: Centers forced to focus on benchmarks rather than patients
By Markian Hawryluk

"Patients who need organ transplants are dying even while viable organs are being thrown out, as government regulations have forced transplant centers to focus on overall survival rates instead of the well-being of individual patients.

"The rules implemented by the Medicare program, which pays for the vast majority of organ transplants in the U.S., evaluate transplant centers based on the one-year patient and organ survival rates after transplant. Centers that fall below benchmarks could be shut down or forced into a lengthy and expensive remedial process.

That has prompted many centers to choose healthier patients and higher quality organs to transplant. High-risk patients that could pull down a center’s overall survival rate are often unable to get on the transplant list, or end up dying on the waiting list as centers pass on marginal but still usable kidneys, livers and lungs. And the decades-long growth in the number of transplants performed in the U.S. has plateaued since the regulations were implemented.

“The side effect has been to turn people risk averse,” said Dr. Dorry Segev, a transplant surgeon at Johns Hopkins Medical Center, “to the point where patients who would benefit from transplant are being denied transplants, and to the point where organs that are beneficial to patients are discarded.”

The regulations have caused centers to take a hard look at their quality-improvement mechanisms and how they evaluate patients and organs for transplant.

Programs became acutely aware of their survival rates, which began to climb after the rules went in effect in 2007.

The percentage of hearts still beating one year after transplant hit an all-time high of 91 percent in 2010, up from 88 percent in 2003; 85 percent of transplant lungs were still breathing, up from 80 percent in 2003. Deceased donor kidney transplants have achieved an astonishing 93 percent success rate.

"Transplant surgeons routinely credit the regulations with strengthening the transplant system and for the most part, improving the quality of care for transplant patients.

“There is no question we have a healthier system and a more monitored system than we did before these regs came out,” said Dr. Michael Abecassis, director of the transplant center at Northwestern University in Chicago.

The unintended consequence, he said, is that centers are “cherry-picking” their patients.

“Patients might die and not get transplanted even though they may have an 80 percent chance of survival,” Abecassis said. “If the target is 90 percent, or you’re going to get flagged, you may look at 80 and say, ‘I’m not going to do that.’ Well, if you’re the patient, it’s 80 versus zero. Then 80 is pretty good.”

 ...
"Before the regulations went into effect, the number of transplants in the U.S. had grown each year. But in 2007, the first year centers were being held to the survival benchmarks, the number of transplants dropped and has remained flat ever since."
****************

Transplant centers pull back to avoid sanctions
PART 2: High-risk patients can put programs in jeopardy

"The new regulations, known as the Medicare Conditions of Participation, or COPs, were finalized in 2005 after a series of highly public transplant scandals. Reports of wait-list irregularities, diversions of organs and major medical errors spurred CMS officials to step in and establish regulations for what primarily had been a self-regulated field.

"CMS, the largest purchaser of transplants in the world (with the possible exception of China), adopted metrics originally developed by the transplant industry itself. Years earlier, the United Network for Organ Sharing’s Organ Procurement and Transplant Network had set up a flagging system to help identify and improve programs with subpar results.

But while that flagging system relied on peer review and public disclosure, the CMS regulations threaten to shut down centers that don’t improve.

Centers whose number of patient deaths or organ failures exceed 150 percent of what would be expected for their mix of patients are flagged. Multiple flags within a 21⁄2-year period trigger CMS action.

Centers then have 210 days in which to explain the mitigating factors that led to their low survival rates. If programs can improve by the end of that period, they are allowed to continue operating as usual. In other cases, CMS will acknowledge the mitigating circumstances and grant exceptions.

The centers that can’t improve quickly or convince CMS to grant an exception are given three options: shut down voluntarily, shut down involuntarily, or enter into a systems improvement agreement, or SIA. Through August 2012, 127 of the nation’s 330 transplant centers were flagged twice and investigated by CMS, including the programs at Hopkins.

*****************

Transplantable organs go to waste
PART 3: Centers feel regulatory pressure to avoid non-ideal organs

"Transplant programs may have to make do with more marginal organs going forward. Ideal organs generally come from young, healthy individuals who incur a traumatic death. Those types of donors are becoming less common, due to gains in highway safety and medical advances that can save accident victims from brain death. A Canadian study released in October found that the percentage of patients with brain injuries who eventually were declared brain dead fell from a high of 9.6 percent in 2004, to 2.2 percent in 2010.

"Donations after brain death accounts for half of kidney transplants, three-quarters of liver transplants, 90 percent of lungs, and all hearts. But that is changing.

"An increasing proportion of organs are coming from patients whose hearts stopped beating before the other organs could be recovered, or from older, sicker donors. Both categories of organs have a lower chance of surviving one year after transplant.

“We are seeing more and more organs sitting in that category of marginal organs,” Alexander said.

"Expanded-criteria donor kidneys (ECD), for example, have an 82 percent one-year survival rate, compared with 90 percent for a standard-criteria kidney. A kidney procured from a brain-dead donor has a 91 percent one-year survival rate, while those recovered from donors after cardiac death (DCD) have an 89 percent rate.

"When patients are listed for transplant, they are usually presented with a menu of organ types, each with different risk profiles, and asked to choose what types of organs they would accept. When organs become available, doctors and transplant coordinators decide whether they are willing to transplant that organ into that patient and whether to pass on that offer to the patient.

"Studies show that in aggregate, transplant candidates who accept ECD kidneys do better over the long run than patients who wait on dialysis for a more ideal kidney. While standard kidneys last an average of 10 years post-transplant, ECD kidneys average five.
...
"Yet marginal and high-risk organs are routinely turned down. An analysis of organ sharing data by surgeons at the University of California, San Francisco, found that 84 percent of patients who died waiting for a liver had received at least one organ offer and an average of six offers. Most were declined by the surgeons due to donor age or quality of organ.

“Wait-list deaths are not simply due to lack of donor organs as many of us assume,” lead author Dr. John Roberts said, citing the stigma of non-ideal livers.

"When surgeons become more selective about marginal organs, it can quickly result in a snowball effect. When an organ becomes available it is offered electronically to centers in the region and across the country. The more centers decline an organ, the more surgeons with patients lower down the list begin to wonder why the organ has been passed on so many times. And with each refusal, the time the organ sits on ice and degrades in quality increases. Kidneys can still be transplanted up to 48 hours after being put on ice, livers less than a day. If no center is willing to take a chance, eventually the organ is simply thrown out.

“Our system is pretty slow,” Roberts said. “It works well for good organs. It doesn’t work that well to get not-so-good organs broadly distributed.”

Prisoner's dilemma, exam grading, and social media

Long long ago I wrote an undergraduate exam that included a prisoner's dilemma question framed as the question of whether to study or not for a particular exam, with your payoff on the exam depending in part on what other people did.  Here is that story brought to life, complete with how it was organized by social media, and enforced by a contingent strategy that depended on mutual observation. (The url is more informative than the headline:

http://www.insidehighered.com/news/2013/02/12/students-boycott-final-challenge-professors-grading-policy-and-get


"Since he started teaching at Johns Hopkins University in 2005, Professor Peter Fröhlich has maintained a grading curve in which each class’s highest grade on the final counts as an A, with all other scores adjusted accordingly. So if a midterm is worth 40 points, and the highest actual score is 36 points, "that person gets 100 percent and everybody else gets a percentage relative to it,” said Fröhlich.

"This approach, Fröhlich said, is the "most predictable and consistent way" of comparing students' work to their peers', and it worked well.

"At least it did until the end of the fall term at Hopkins, that is.

"As the semester ended in December, students in Fröhlich’s "Intermediate Programming", "Computer Science Fundamentals," and "Introduction to Programming for Scientists and Engineers" classes decided to test the limits of the policy, and collectively planned to boycott the final. Because they all did, a zero was the highest score in each of the three classes, which, by the rules of Fröhlich’s curve, meant every student received an A.

“The students refused to come into the room and take the exam, so we sat there for a while: me on the inside, they on the outside,” Fröhlich said. “After about 20-30 minutes I would give up.... Then we all left.” The students waited outside the rooms to make sure that others honored the boycott, and were poised to go in if someone had. No one did, though.

"Andrew Kelly, a student in Fröhlich’s Introduction to Programming class who was one of the boycott’s key organizers, explained the logic of the students' decision via e-mail: "Handing out 0's to your classmates will not improve your performance in this course," Kelly said.
"So if you can walk in with 100 percent confidence of answering every question correctly, then your payoff would be the same for either decision. Just consider the impact on your other exam performances if you studied for [the final] at the level required to guarantee yourself 100. Otherwise, it's best to work with your colleagues to ensure a 100 for all and a very pleasant start to the holidays."

"Kelly said the boycott was made possible through a variety of technological and social media tools. Students used a spreadsheet on Google Drive to keep track of who had agreed to the boycott, for instance. And social networks were key to "get 100 percent confidence that you have 100 percent of the people on board" in a big class.

"Fröhlich took a surprisingly philosophical view of his students' machinations, crediting their collaborative spirit. "The students learned that by coming together, they can achieve something that individually they could never have done," he said via e-mail. “At a school that is known (perhaps unjustly) for competitiveness I didn't expect that reaching such an agreement was possible.”
Although Fröhlich conceded that he did not include such a “loophole” in the policy “with the goal of students exploiting it,” he decided to honor it after the boycott.

"Despite awarding As to all the students who participated in the boycott, the experience has led Fröhlich to alter his long-held grading policy.
“I have changed my grading scheme to include ‘everybody has 0 points means that everybody gets 0 percent,’ ” Fröhlich said,  “and I also added a clause stating that I reserve the right to give everybody 0 percent if I get the impression that the students are trying to ‘game’  the system again.”

School choice in Chicago and England, and some ideas about comparing manipulabiliby

The current (February 2013) issue of the AER has the following paper by Parag Pathak and Tayfun Sonmez, which combines some incisive observation with some new theory:

School Admissions Reform in Chicago and England: Comparing Mechanisms by Their Vulnerability to Manipulation

Parag A. Pathak and Tayfun Sönmez Abstract:

"In Fall 2009, Chicago authorities abandoned a school assignment mechanism midstream, citing concerns about its vulnerability to manipulation. Nonetheless, they asked thousands of applicants to re-rank schools in a new mechanism that is also manipulable. This paper introduces a method to compare mechanisms by their vulnerability to manipulation. Our methodology formalizes how the old mechanism is at least as manipulable as any other plausible mechanism, including the new one. A number of similar transitions took place in England after the widely popular Boston mechanism was ruled illegal in 2007. Our approach provides support for these and other recent policy changes.  "

Here's their story of school choice in England (where I am today):


"In England, forms of school choice have been available for at least three decades.  The nationwide 2003 School Admissions Code mandated that Local Authorities, an operating body much like a US school district, coordinate their admissions practices. This reform provided families with a single application form and established a common admissions timeline, leading to a March announcement of placements for anxious 10 and 11 year-olds on National Offer Day. The next nationwide reform came with the 2007 School Admissions Code. While strengthening the enforcement of admissions rules, this legal code also prohibited authorities from using what they refer to as ‘‘unfair oversubscription criteria’’ in Section 2.13:

In setting oversubscription criteria the admission authorities for all 
maintained schools must not: give priority to children according to the order of other schools 
named as preferences by their parents, including ‘first preference first’ arrangements.

A first preference first system is any ‘‘oversubscription criterion that gives priority to children according to the order of other schools named as a preference by their parents, or only considers applications stated as a first preference’’ (School Admissions Code 2007, Glossary, p. 118). The 2007 Admissions Code outlaws use of this system at more than 150 Local Authorities across the country, and this ban continues with the 2010 Code. The best known first preference first system is the Boston mechanism, and since 2007 it is banned in England.  The rationale for this ban, as stated by England’s Department for Education and Skills, is that ‘‘the ‘first preference first’ criterion made the system unnecessarily complex to parents’’ (School Code 2007, Foreword, p. 7). Moreover, Education Secretary Alan Johnson remarked that the first preference first system ‘‘forces many parents to play an ‘admissions game’ with their children’s future.’’

Iraqi arranged marriages--the suicide constraint

The individual rationality constraint isn't very binding when individuals don't have good outside options, but it doesn't entirely go away. In Iraq, it's taking the form of suicide by unwilling brides in arranged marriages: Where Arranged Marriages Are Customary, Suicides Grow More Common

"In this desolate and tradition-bound community in the northwest corner of Iraq, at the foot of a mountain range bordering Syria, Ms. Merza’s reaction to the ancient custom of arranged marriage is becoming more common. Officials are alarmed by what they describe as a worsening epidemic of suicides, particularly among young women tormented by being forced to marry too young, to someone they do not love.

"While reliable statistics on anything are hard to come by in Iraq, officials say there have been as many as 50 suicides this year in this city of 350,000 — at least double the rate in the United States — compared with 80 all of last year. The most common methods among women are self-immolation and gunshots.

"Among the many explanations given, like poverty and madness, one is offered most frequently: access to the Internet and to satellite television, which came after the start of the war. This has given young women glimpses of a better life, unencumbered by the traditions that have constricted women for centuries to a life of obedience and child-rearing, one devoid of romance.
...

"Ms. Merza’s father, Barkat Hussein, interviewed later in private, said he was aware that the shooting was not an accident.

“We gave her to her cousin less than 20 days ago,” he said. “She accepted him. Like anyone who gets married, she should be happy.”

"He said he would not force her to return to her husband, who lives next door. But, he said: “I hope she will go back to him. His father is my brother.”

"He, too, blamed the Turkish soap opera for his daughter’s unhappiness, and he nodded toward the room where his wife was working. “I got married to my cousin,” he said. “I wasn’t in love with her, but we are here, living together. That’s what happens here, we marry our relatives.”

Regulating prostitution in New Zealand

An incentive issue (although they are not the focus of this story):
"But the Prostitutes Collective warned that outlawing popular streets would encourage sex workers to stop carrying condoms in case they are questioned by police."

Incentives and organ donation in Germany

Bettina Klaus writes:  There is an article about a recent organ donation scandal at the university hospital Goettingen. 

One of the transplant surgeons there seems to have manipulated patient files (25 cases are currently under investigation) in order to push his patients up on the waiting list. Interestingly his salary depended on the numbers of transplants he did and the article states that under his lead the number of liver transplants at the university hospital Goettingen had significantly increased in 2009 and 2010. 

Here's an English account of that, and organ sales as well: Desperation, Greed and the Global Organ Trade 
************

Something similar  to the Goettingen events led to a change in the deceased-donor liver allocation system in the U.S.
*******************************

See also http://www.spiegel.de/gesundheit/diagnose/organspende-skandal-goettingen-montgomery-fordert-schaerfere-kontrollen-a-847985.html, for a related story (also in German), which focuses on recent changes in the German transplant and organ donation systems. (HT Rosemarie Nagel)
************

And Markus Mobius points to this article (also in German, but the headline is clear): Transplantationsskandal

He writes "The article says that an increasing share of transplants in Germany are now bypassing the waiting list and go straight to in-house patients.
"10 years ago, 9.1% of livers, 8.4% of hearts, 10.6% of lungs and 6.3% of pancreas bypassed the waiting list. Now the shares are 37.1%, 25.8%, 30.3% and 43.7%, respectively. "

Incentives and privacy

A new paper by three computer scientists and an economist reports on some connections between privacy and incentive compatibility.

MECHANISM DESIGN IN LARGE GAMES: INCENTIVES AND PRIVACY
by
MICHAEL KEARNS, MALLESH M. PAI, AARON ROTH and JONATHAN ULLMAN
July 18, 2012


ABSTRACT
We study the design of mechanisms satisfying two desiderata— incentive compatibility and privacy. The first, requires that each agent should be incentivized to report her private information truthfully. The second, privacy, requires the mechanism not reveal ‘much’ about any agent’s type to other agents. We propose a notion of privacy we call Joint Differential Privacy. It is a variant of Differential Privacy, a robust notion of privacy used in the Theoretical Computer Science literature. We show by construction that such mechanisms, i.e. ones which are both incentive compatible and jointly differentially private exist when the game is ‘large’, i.e. there are a large number of players, and any player’s action affects any other’s payoff by at most a small amount. Our mechanism adds carefully selected noise to no-regret algorithms similar to those studied in Foster-Vohra [FV97] and Hart-Mas-Colell [HMC00]. It therefore implements an approximate correlated equilibrium of the full information game induced by players’ reports.
*********

As I understand it, adding appropriate randomness to regret learning algorithms doesn’t harm their long term equilibration properties, and gives them good privacy properties, which together give them good incentive properties.

Kidney exchange pilot program: progress report

The official name is the OPTN national kidney paired donation pilot program (KPDPP). The latest news release reports that in March 2012 there were 110 transplant centers enrolled.

However the report also notes that "as of this writing" there are only 129 patients enrolled. That's just about one patient per program. And since 2010 a total of 19 transplants have been performed to date.

I'm hoping that the program will take a look at this participation problem.

Here's a bit of good operational news from a recent announcement:

"Match run frequency increased to twice a month

The goal of the OPTN Kidney Paired Donation Pilot Program (KPDPP) is to increase the number of transplants for candidates who have a willing but incompatible living donor. One way to do this is to increase the number of match runs to more than one a month, which was recently requested by the KPDPP work group and many of the 118 participating transplant centers."

School choice in San Francisco, reports on first year

I've written before about school choice in San Francisco, and about how Muriel Niederle and Clayton Featherstone led the effort by a group of our colleagues to design a strategy-proof choice algorithm (explained here at a Board meeting in 2010), based on transfer cycles ("top trading cycles" to game theorists...). The school board adopted the plan, but then the staff of the school district decided to implement it themselves, without making the details public. Fast forward to 2012, when the first children have been assigned by the new plan.

Rachel Norton's blog has a post about it here: They're out! School Assignment Letter 2012, and an earlier one here, with a link to a March 5, 2012 SFUSD report on Student Assignment. As in the previous SFUSD reports, this does not describe the choice algorithm, it only describes the "tie breakers" that are used whenever the algorithm would otherwise try to assign more students to a school than it has room for.

This outraged Stan Goldberg (who reports about SF schools as "Senior Dad"), and he posted a video about the lack of transparency called Assignment System Fraud?

He must be an influential guy, because this prompted SFUSD to post some new information, including this "fact sheet" dated March 23, called How does the student assignment computer program work?  It still doesn't come close to explaining the actual algorithm they use, but it does include a diagram of "transfer cycles."

Which raises a question. If they in fact implemented the plan we proposed and the Board adopted, you would think they would want to make this clear. The benefits of a strategy-proof assignment procedure can only be realized if parents know that they can safely list their true preferences.

On the other hand, if the algorithm isn't correctly implemented, or if some other assignment algorithm is implemented (whether or not it includes some use of transfer cycles) then it would most likely not be strategy-proof, that is, it might not be safe for parents to reveal their true preferences, and it might be in the interest of some to "game the system" in some way. That might account for a desire to keep the algorithm secret. (So might a desire to avoid revealing any inadvertent mistakes in implementation...)

I should say that SFUSD's brief description of their algorithm doesn't look to me like it describes one that is strategy-proof...:(On the contrary, it looks like it might be patched together from something like Boston's old immediate acceptance algorithm followed by some trading...but then again, it isn't a complete enough description to make me confident that it is a description of whatever they are in fact doing...)

Anyway, one point of this post is to say that, unlike the case of the systems in New York and Boston and the work that IIPSC is doing around the country, my colleagues and I don't know what algorithm SFUSD is using, even though we know what we proposed and the Board adopted. So...this post is a bit like the ads that sometimes appeared in the financial sections of newspapers when I was young, which, following a divorce, would announce that Mr John Doe was henceforth no longer responsible for any debts incurred by the former Mrs John Doe...

Incentive compatibility in a giant outdoor 18th century restaurant

"the waiters were prevented from cheating the management by being made to pay cash from their own pockets for everything they ordered from the kitchens, reimbursing themselves when the customers settled their bills."

That detail is from The English pleasures of Vauxhall, a large, long lived "pleasure gardens" on the banks of the Thames that was both an outdoor concert venue and an innovative early large scale caterer to its customers, founded in the 1700's.

Medical (and pre-medical) culture: cheating on tests

While some of the very best motivated and most talented students find their way into medicine, it is also a large, well-compensated profession fed by a stream of undergraduate "pre-med" majors who grow up in a culture of exam-taking. It shouldn't be surprising if this carries over into their post-graduate years (in fact I can't even tell if this CNN headline is meant to be ironic): Exclusive: Doctors cheated on exams

"For years, doctors around the country taking an exam to become board certified in radiology have cheated by memorizing test questions, creating sophisticated banks of what are known as "recalls," a CNN investigation has found."

School choice in Denver: communication, communication, and communication

Shannon Fitzgerald, director of choice and enrollment services for Denver Public Schools, explains what parents need to know about a strategy proof system:

“All you have to put on the form is what you really want for your kid. There is no strategy that you can really employ … All parents needs to do is tell us is what they really want.”

Seattle backs away from transparent (strategy-proof) school choice

It appears that, after a decade of using a school choice system based at least in part on a deferred acceptance algorithm, in the course of making some other changes, Seattle schools may have backed into using a non-strategy-proof immediate acceptance algorithm instead.

Here's a Q&A on How to Rank Schools for the Seattle Public Schools Choice Process for 2010-2011

"My true first choice option school is very popular and hard to get in to. If I list it first on my School Choice form, am I throwing away my chances of getting in?

"For the 2010-2011 school year, you would not be throwing away your chances.

"This year the district is using what it refers to it as the Barnhart/Waldman processing algorithm, named after the school board members who suggested it. It is more properly called the Gale/Shapley “Stable Marriage” algorithm. This algorithm is well studied and understood by computer scientists and mathematicians. It has been specially designed so that there is never an advantage to listing your school choices in any order other than your true preference. The Gale-Shapley paper referenced below is highly recommended. It is well written and does not require special software or mathematics skills to understand. In addition to being used for school assignments in many districts, a variation of this algorithm is used for the yearly matching of graduating medical students to residency programs.

"Note that this will change when the NSAP is complete. The final plan calls for using a simpler algorithm in which families who list a school as their first choice have priority over families who list the school second choice or later. This can result in cases where your best chance for getting into one of your choices is to “lie” and list your second choice first."
********

And here's an analysis by a concerned parent, Chris MacGregor, who happens to be a software engineer and took the trouble to examine the school district's school choice algorithm source code: Serious Problem With New Student Assignment Plan
********

It sounds like the plan may have been passed hastily, with the change in algorithm being a detail in a more politically charged discussion: Student-assignment plan passed behind closed doors.
********

Here's some background on school choice in Seattle, in the form of court documents from 2001, about the use of race as a tie-breaker in school assignment.
********

Here's a 1998 story about the adoption of the deferred acceptance algorithm, with some discussion touching on strategy-proofness and making it safe to submit true preferences: New Student-Assignment Plan Set
"The change, devised by board members Scott Barnhart and Nancy Waldman, will extend the district's "hold harmless" rule to all choices. Currently, when students apply to alternative schools, they don't lose their place in line for a regular school.

But, when students apply to popular regular schools and don't get in, the computer bumps them down the list until it finds a school with a vacancy - even though they may have siblings in their second-choice school or may live right across the street.

Barnhart and Waldman's system will ensure that more students get their real second choice and eliminate bumping down to lower and lower choices, which has discouraged parents and driven some to send their children to private schools, according to Waldman.

At last night's meeting, Waldman said that had happened to her.

Space wasn't available at her first-choice kindergarten, so her daughter was assigned to a school Waldman hadn't even listed. So, Waldman put her daughter in private school for three years."

******
Note that Scott Barnhart is a medical prof at U. of Washington, who was matched to a number of residency positions, and so would have been familiar with the medical match.
*********

And here's a 1996 story (short on detail, but gives some of the political atmosphere) about the run-up to the adoption of the deferred acceptance algorithm that is apparently on its way out: Vote On Reform Plan -- School Board To Tackle Student-Assignment System
*******
For those of you new to discussions of making school choice simple and safe for parents to reveal their preferences over schools, see the papers here about New York and Boston, and see the Institute for Innovation in Public School Choice which has more recently been involved in the design of school choice programs in Washington D.C., Denver, and New Orleans.

HT: Parag Pathak

An Incentive System with Heart: Wharton celebrates Judd Kessler

That's the subheading of an article in which Wharton celebrates the work of Judd Kessler: How to Encourage People to Become Organ Donors: An Incentive System with Heart

"The decision to be an organ donor may seem easy for some: You sign an agreement that will let your heart, kidneys, liver, pancreas and other organs be used after your death in a way that helps the recipients lead fuller, healthier lives.
But for other people, the choice is harder. Some fear that a doctor may not work as hard to save them because he or she wants their organs for other patients, or that their organs might be removed prematurely (although there is no evidence to support either of these concerns). There may also be a psychological cost of having to think about your own death at a time when you are still relatively healthy. Other people may simply not want to bother with a program that doesn't directly benefit them.
It is against this backdrop that Wharton business and public policy professor Judd Kessler and Harvard economics professor Alvin Roth set out to see whether changes in the management of organ waiting lists could increase the number of donors."
...


The paper is here:
Kessler, Judd B. and Alvin E. Roth, '' Organ Allocation Policy and the Decision to Donate,'' American Economic Review, forthcoming.


Update: when the Financial Times covered the story (scroll down here) they gave Judd a new first name...

Incentives for cheating and for cheating-detection

A chilling and surprisingly complicated story from Panos Ipeirotis at NYU Stern (who blogs under the title "A Computer Scientist in a Business School): Why I will never pursue cheating again (since deleted, but that URL now has related links). Apparently he got a cease and desist letter, and the story has since attracted some press.

Here's one story about it: NYU Professor Catches 20% Of His Students Cheating, And He's The One Who Pays For It,

and here's another: NYU Prof Vows Never to Probe Cheating Again—and Faces a Backlash

Ultimately he draws a market design conclusion:
"In Mr. Ipeirotis’ view, if there’s one big lesson from his semester in the cheating trenches, it’s this: Rather than police plagiarism, professors should design assignments that cannot be plagiarized.

"How? He suggested several options. You could require that projects be made public, which would risk embarrassment for someone who wanted to copy from a past semester. You could assign homework where students give class presentations and then are graded by their peers, ratcheting up the social pressure to perform well. And you could create an incentive to do good work by turning homework into a competition, like asking students to build Web sites and rewarding those that get the most clicks."

Further followup on school choice in San Francisco

My Thursday post, Followup on school choice in San Francisco, has generated some followup on its own, in the form of an audio interview yesterday of School Board member Rachel Norton by Stan Goldberg who follows the SF school system under the name Senior Dad.  He summarizes the interview as "Straight answers from Commissioner Norton “because people have a right to know”."

The issue is whether the algorithm adopted by the board last year was in fact implemented correctly by the district staff. It's an important question because the correctly implemented algorithm would be strategy proof, and if parents had confidence in this it would vastly simplify the school choice system from parents' point of view.

Here is my very incomplete and possibly imperfect transcript to give the flavor of the last 5 minutes of the interview (starting just after minute 39) in which Stan Goldberg (SG) raises this issue, and Rachel Norton (RN) replies. It's worth listening to.

SG “The school district was supposed to release the algorithm they were assigning students on, and so far they have not released that algorithm.”

RN “you’ve been reading Al Roth’s blog” ...“I’ve advocated for that, and will continue to advocate for that. I don’t think the staff right now wants to do that. [laughter] But short of 5 votes, 4 votes, they don’t have to.

…

SG ‘why should the public trust the school district?  “I’ve had the deputy Superintendent say ‘you guys shouldn’t trust us, we haven’t been reliable’. He said that; I believed him.”

RN “I don’t know what to tell you Sam, I think we should release the algorithm, and I’ve said that to staff, I’ve said that to the Superintendent”…short of 3 other board members joining with me and demanding that it be released the superintendent can do what he thinks is best,  unless he’s ordered by the board to do something else…”

SG “not releasing the algorithm makes everybody think something funny is going on…”

RN” well, not everyone…”

Reverse scholarships reversed at KAIST

A recent article concerning several suicides at the Korea Advanced Institute of Science and Technology pointed out that a high powered incentive system might possibly be implicated.

Elite South Korean University Rattled by Suicides
"Mr. Suh also engineered a system that required students to pay extra tuition for each hundredth of a point that their grade point average fell below 3.0 (based on a 4.3-point system). All students pay a token fee each semester, Kaist administrators said, but otherwise their tuition is free, financed by government scholarships.

"Under the so-called punitive tuition program, a bad semester could cost a student’s family thousands of dollars.

"The program, which was applauded at first, has since led to deep humiliation and anxiety among many students. Those who struggled and lost their full rides suddenly saw themselves as losers. Some critics, calling it ruthless, even blamed the program for the recent suicides."

Non-simultaneous kidney chains are getting longer

Mike Rees' revolution in non-simultaneous extended altruistic donor (NEAD) chains continues to grow in importance in kidney exchange. Since the 2009 NEJM paper reporting the first NEAD chain, which accomplished 10 transplants (and therefore required 20 surgeries, ten of them nephrectomies), chains have been getting both more frequent and often longer.

The venerable transplant center at the University of Pittsburgh reports its recent involvement in a long chain:
A UPMC First: Transplant Team Participates In Large Multi-State Kidney Chain Involving 32 Operations and 16 Transplants
"The paired kidney exchange, coordinated by the National Kidney Registry, involved 32 patients and 16 transplants performed at 12 U.S. hospitals over two months. It marked the first time that UPMC participated in a kidney chain."

The National Kidney Registry is one of the relatively nimble networks that, following the New England Program for Kidney Exchange, and Rees' Alliance for Paired Donation, have grown by recruiting sometimes overlapping networks of transplant centers.
Sean Hamill at the Pittsburgh Post Gazette gives the story a competitive angle by contrasting the NKR with the national exchange that is slowly gearing up (emphasis added, on which I'll comment after):

32-person chain nets 16 kidney transplants: 2 national groups in big competition

"This most recent chain started Dec. 17, 2010, and by the time it ended Feb. 11 it involved 16 donors and 16 recipients at 12 transplant centers in nine states from New York to California. Mr. Johnson's transplant and Ms. Dolezal's kidney removal both took place Feb. 10 in Pittsburgh and both of them are doing well.
...
"A national program that could push the number of paired donations from nearly 400 last year nationally to 3,000 or more a year has been the dream for the last half decade. That's significant when 93,000 people are on the national waiting list for kidneys and only about 17,000 a year get transplants

"Last fall, the United Network for Organ Sharing, an organization that oversees the nation's organ and transplant network, finally began its long-awaited pilot program that will attempt to do just that and link all of the eligible patients from the nation's more than 200 transplant centers.

"But Garet Hil, founder and president of the National Kidney Registry in Babylon, N.Y., which organized the 32-person chain, said UNOS's pilot program "is a failure."

"Mr. Hil, who runs a consulting and software development company, sits on the Kidney Paired Donation Pilot working group that has been trying to make the program work.

"I've witnessed a program with a flawed design, working in a bureaucratic way that's not going to get many people transplanted," he said Friday in a phone interview. "The program has been out since October and it's only done two transplants and we've done 60 since then -- including this chain" of which UPMC was a part.
...
"He has a host of criticisms with the pilot program, including that it allows only small chains that would give just two or three people new kidneys at a time, while the National Kidney Registry runs chains, like this recent one, where dozens of people get new kidneys.

"Ken Andreoni, an Ohio State University transplant surgeon who chairs UNOS's kidney committee, has heard Mr. Hil's arguments and concerns before and he believes he's just being too impatient.

"I'm in this for the 50-year- and 100-year-long issues," he said, adding that they will take time to solve.

"He concedes the pilot program is slowed by the bureaucracy of having to follow rigid rules that aim to maximize successful transplants and minimize risk.

"But that's the price you pay when you're creating what is to be a real national program that answers to everyone, unlike Mr. Hil's registry.

"The pilot program has had a difficult time because, first, the patients who have signed on are typically the most difficult patients to match, with significant variables that make them hard to match.

"And while UNOS would love to run long chains like the registry, currently it believes that shorter, two- and three-person-paired chains are safer because they're less complicated, Dr. Andreoni said.

"He doesn't see it as the competition Mr. Hil does.

"In the end, if you get people off the waiting list, that's great," Dr. Andreoni said, "no matter who is doing it."

*******
About the controversy about long versus short chains, the concerns about long chains are based in part on a modeling error, see this earlier post: Nonsimultaneous kidney exchange chains produce more transplants than simultaneous chains

And the issue about having only hard-to-match patients enrolled so far in the national exchange, that is likely to be a problem that will need to be addressed at a fairly fundamental level, because the current incentives make that a tempting strategy for transplant centers. See this earlier post: Kidney exchange when hospitals are the players

Finally, regarding time horizons, I sure hope that in 50 and 100 years we'll have better cures for kidney disease than transplantation. But for the next 10 or 20 it's likely to be the best solution by far for patients with end stage renal disease, and so we'd better keep figuring out how to make the best use of it.

Relative manipulability of school choice systems

That's the topic of a new NBER working paper by Parag Pathak and Tayfun Sönmez:

SCHOOL ADMISSIONS REFORM IN CHICAGO AND ENGLAND: COMPARING MECHANISMS BY THEIR VULNERABILITY TO MANIPULATION, Working Paper 16783.

Abstract: In Fall 2009, officials from Chicago Public Schools changed their assignment mechanism for coveted spots at selective college preparatory high schools midstream. After asking about 14,000 applicants to submit their preferences for schools under one mechanism, the district asked them re-submit their  preferences under a new mechanism. Officials were concerned that "high-scoring kids were being rejected simply because of the order in which they listed their college prep preferences" under the abandoned mechanism. What is somewhat puzzling is that the new mechanism is also manipulable.
This paper introduces a method to compare mechanisms based on their vulnerability to manipulation. Under our notion, the old mechanism is more manipulable than the new Chicago mechanism. Indeed, the old Chicago mechanism is at least as manipulable as any other plausible mechanism. A number of similar transitions between mechanisms took place in England after the widely popular Boston mechanism was ruled illegal in 2007. Our approach provides support for these and other recent policy changes involving matching mechanisms.

Can we make school choice more efficient?

A new paper by Eduardo Azevedo and Jacob Leshno addresses a recent puzzle in school choice: Can we make school choice more efficient? An Example

Here's their abstract:
"The deferred acceptance mechanism, currently used in the New York City and in the Boston public school systems, can produce Pareto-dominated assignments. When students are non-strategic a Pareto-dominating and efficient mechanism can be achieved by allowing some students to trade schools (e.g. Stable improvement Cycles, Erdil and Ergin 2008). We find that when students are strategic the equilibrium assignments of these mechanisms can be unstable with respect to the true preferences and can be Pareto inferior to the deferred acceptance equilibrium assignment."

The paper follows up on the Erdil and Ergin 2008 AER paper, and also on the paper by Atila Abudlkadiroglu, Parag Pathak and me which discusses the NYC high school match and shows that about a thousand NYC high school students could in principle be made better off
Abdulkadiroglu, Atila , Parag A. Pathak, and Alvin E. Roth, "Strategy-proofness versus Efficiency in Matching with Indifferences: Redesigning the NYC High School Match,'' American Economic Review, 99, 5, Dec. 2009, pp1954-1978.

That paper sets the stage for Azevedo and Leshno by observing that about 1,000 of the approximately 90,000 new high school students each year could have been given better matches if strategy-proofness were not a desiderata, and if their true preferences could be recovered without strategy proofness. So that raises the open question of what kind of preferences would be submitted to a system that tried to get these apparent welfare gains, and whether welfare gains could be achieved at equilibrium:
"Nothing is yet known about what kinds of preferences one could expect to be strategically submitted to such a mechanism, or what their welfare consequences would be. Consequently, there is room for more work to further illuminate the tradeoff between efficiency and strategy-proofness."

Azevedo and Leshno show that there's at least a possibility that attempting to improve welfare while sacrificing strategy-proofness could, in equilibrium, reduce welfare.

One of the exciting things about the modern school choice literature is how it has given rise to questions about matching that didn't arise in other applications (like to labor market clearinghouses). The special feature of school choice is the importance of how ties are broken, between between students who have equal priority for a given school that doesn't have capacity for all of them.

Azevedo and Leshno advance the discussion of how we should evaluate school choice algorithms, in the complex settings in which we're asked to design them. Their paper shows how strategy-proofness and welfare can be intertwined in subtle ways at equilibrium. There's a lot more to learn about this, so their short paper answers one question and raises many others.