Friday, October 11, 2019

Followup on Robert Kraft: disappearing sex trafficking in Florida

Vanity Fair has a followup on the widely publicized Florida investigation of sex trafficking that included the arrest of Patriots owner Robert Kraft.  The reporter points out that the trafficking charges have evaporated, and in general concludes that much of the concern with trafficking is in fact simply targeted at voluntary foreign sex workers:

“YOU WON’T BELIEVE WHAT HAPPENED”: THE WILD, DISTURBING SAGA OF ROBERT KRAFT’S VISIT TO A STRIP MALL SEX SPA
After the Patriots owner made two trips to Orchids of Asia Day Spa, where a half-hour “massage” costs $59, he was charged with soliciting a prostitute. What happened next was not what anyone expected.
BY MAY JEONG

"Human trafficking is a serious problem: The Department of Health and Human Services calls it the world’s “fastest-growing criminal industry.” But some anti-trafficking groups, in search of funding, routinely overstate the scale of the commercial sex trade. They frequently claim that 300,000 minors are “at risk” for being sold into sexual slavery in America each year—a number that has been debunked by researchers as wildly overinflated. (The Washington Post dismisses it as a “nonsense statistic.”) In 2018, the FBI confirmed a total of 649 trafficking cases in America, adults included.
...
"Florida’s new sex registry is the latest in a long line of similar laws. One of America’s first laws against prostitution, in fact, was the 1870 Act to Prevent the Kidnapping and Importing of Mongolian, Chinese, and Japanese Females for Criminal or Demoralizing Purposes, intended to protect the public from “scandal and injury.” The law was a precursor to the Page Act of 1875, which aimed to “end the danger of cheap Chinese labor and immoral Chinese women,” which in turn was a precursor to the Chinese Exclusion Act of 1882—the first law to bar all members of a specific ethnicity or nationality from immigrating.

"The raids on Orchids and other massage parlors in South Florida were conducted in the name of rescuing women from sex trafficking. But the only people put in jail were the women themselves. A few, like Lulu and Mandy, managed to post bail and were placed under house arrest. But others were transferred to the custody of ICE. Women who migrated to America in search of work—who chose the least bad option available to them—were being punished for what one of their lawyers calls “the crime of poverty.”
...
"Within weeks of the raids, the state’s case had evaporated. There was no $20 million trafficking ring, no women tricked into sex slavery. The things the state had mistaken as markers for human trafficking—long working hours, shared eating and living arrangements, suspicion of outside authorities, ties to New York and China—were, in fact, common organizing principles of many Chinese immigrant communities. As an assistant state attorney in Palm Beach told the court on April 12: “There is no human trafficking that arises out of this investigation.”
********

See earlier post:

Thursday, March 7, 2019

Thursday, October 10, 2019

2019 ESA North American Meetings

The 2019 Economic Science Association (ESA) North American meetings will be held in Los Angeles, California at the campus of Loyola Marymount University from Thursday evening, October 10th through Saturday evening, October 12th, 2019. 
Keynote Speakers:
Panel on Research Methods:


Here's the program.

There's a session on Market Design and Matching:

The Important of Cardinal Information in Matching Clayton Featherstone

Why Do Some Clearinghouses Yield Stable Outcomes? Experimental Evidence on Out-of-Equilibrium Truth-Telling Colin Sullivan

Competition with Indivisibilities and Few Traders Weiwei Zheng

Driving to the Beat Sotiris Georganas

Wednesday, October 9, 2019

Very quick questions and answers about market design and experimental economics--video

Two minutes of Q&A from my recent visit to Lancaster:

Tuesday, October 8, 2019

Transplantation rates for patients in non-profit versus for-profit dialysis centers

From JAMA,September 10, 2019  Volume 322, Number 10:
J::AMA
September 10, 2019 Volume 322, Number 10Association Between Dialysis Facility Ownership and Accessto Kidney Transplantation

Jennifer C. Gander, PhD; Xingyu Zhang, PhD; Katherine Ross, MPH; Adam S. Wilk, PhD; Laura McPherson, MPH; Teri Browne, PhD;Stephen O. Pastan, MD; Elizabeth Walker, MS; Zhensheng Wang, PhD; Rachel E. Patzer, PhD, MPH

"MAIN OUTCOMES AND MEASURES: Access to kidney transplantation was defined as time from initiation of dialysis to placement on the deceased donor kidney transplantation waiting list,receipt of a living donor kidney transplant, or receipt of a deceased donor kidney transplant.Cumulative incidence differences and multivariable Cox models assessed the associationbetween dialysis facility ownership and each outcome.
RESULTS: Among 1 478 564 patients, the median age was 66 years (interquartile range, 55-76years), with 55.3% male, and 28.1% non-Hispanic black patients. Eighty-seven percent ofpatients received care at a for-profit dialysis facility. A total of 109 030 patients (7.4%)received care at 435 nonprofit small chain facilities; 78 287 (5.3%) at 324 nonprofitindependent facilities; 483 988 (32.7%) at 2239 facilities of large for-profit chain 1; 482 689(32.6%) at 2082 facilities of large for-profit chain 2; 225 890 (15.3%) at 997 for-profit smallchain facilities; and 98 680 (6.7%) at 434 for-profit independent facilities. During the studyperiod, 121 680 patients (8.2%) were placed on the deceased donor waiting list, 23 762 (1.6%)received a living donor kidney transplant, and 49 290 (3.3%) received a deceased donorkidney transplant. For-profit facilities had lower 5-year cumulative incidence differences foreach outcome vs nonprofit facilities (deceased donor waiting list: −13.2% [95% CI, −13.4% to−13.0%]; receipt of a living donor kidney transplant: −2.3% [95% CI, −2.4% to −2.3%]; andreceipt of a deceased donor kidney transplant: −4.3% [95% CI, −4.4% to −4.2%]). AdjustedCox analyses showed lower relative rates for each outcome among patients treated at allfor-profit vs all nonprofit dialysis facilities: deceased donor waiting list (hazard ratio [HR], 0.36[95% CI, 0.35 to 0.36]); receipt of a living donor kidney transplant (HR, 0.52 [95% CI, 0.51 to0.54]); and receipt of a deceased donor kidney transplant (HR, 0.44 [95% CI, 0.44 to 0.45]).
CONCLUSIONS AND RELEVANCE: Among US patients with end-stage kidney disease, receiving dialysis at for-profit facilities compared with nonprofit facilities was associated with a lower likelihood of accessing kidney transplantation. Further research is needed to understand the mechanisms behind this association.

Here are the figures. "For-profit large chains" seem to give the slowest access to being put on the transplant waiting list, receiving a living donation, or receiving a deceased donation.



HT: Irene Wapnir

Monday, October 7, 2019

Different misconduct in sperm donation

One reason it is rewarding to study unregulated markets is that it gives you some idea of why some regulation might be desirable.  The growth in DNA registries has allowed many children of sperm donors to identify their biological father, and it also allows donors to identify their children, sometimes with unsettling results.

Here's a story from the Washington Post:

Sperm donor says fertility clinic ‘lied’ after discovering he fathered 17 kids ― most in the same area

"It was 1989 when he gave his sperm to the fertility clinic at Oregon Health & Science University, where he was a first-year medical student, believing his donation would help infertile couples and advance science. The facility promised that once his sperm had conceived five babies in mothers living on the East Coast, the rest would be used for research, Cleary said at a Wednesday news conference. He had assured his wife that the donor kids were far enough away that their own four children could never run into them in their Oregon town, or unwittingly befriend them or fall in love with them.

“So you can imagine his shock,” his attorney Chris Best said at the news conference, “when, after 30 years, Dr. Cleary recently [learned] that no less than 17 children have been born from his donations” ― all of whom were born in the state of Oregon and the Pacific Northwest."

Ride sharing and shared rides as public transportation: Via

The ride sharing company in the U.S. that is perhaps closest to a public transportation model is Via, which offers shared rides from near where you are to near where you are going. That is, it's a ride sharing service which has you walk to a nearby corner to be picked up, pooled with other riders who are headed the same way, and dropped off near your destination.

Via is also taking other steps towards deploying their technology in a public transportation way: here's their recent announcement of a contract with New York City schools to manage school buses:

Via Selected to Power New York City’s School Bus System

"August 21, 2019 (New York, NY) — Via, the leading provider and developer of on-demand public mobility, was selected by the New York City Department of Education to provide a revolutionary school bus management system for the nation’s largest school district. “Via for Schools” will be the first integrated, automated school bus routing, tracking, and communication platform in the world. Parents and students will have the ability to track, in real-time, their bus’ whereabouts and receive frequent and reliable communications in the event of service changes, improving safety and bringing important peace of mind to all users of the system. By utilizing Via’s best-in-class algorithms to optimize school bus routing, the Department of Education will be able to achieve operational efficiencies and reduce transportation costs.

“We are delighted to be partnering with the New York City Department of Education to set a new standard of excellence in school transportation,” said Daniel Ramot, co-founder and CEO of Via. “Via was founded in New York, where we are proud to operate the city’s most efficient on-demand shared ride service. We’re thrilled to have an opportunity to further serve the New York community by applying our technology to operate a world-class school bus system.” 

“Through our partnership with Via, we’ll soon have a state-of-the-art app for families to track buses and get real-time automatic updates,” said Schools Chancellor Richard A. Carranza. “We are grateful for the City Council’s advocacy, leadership and partnership. Safe and reliable transportation is critical for all families, and we’re committed to getting it right this year.”

As the largest school district in the nation, the NYC Department of Education (DOE) transports approximately 150,000 students on 9,000 bus routes each and every day to get students safely to and from school across the City. Via for Schools is purpose built to serve the city’s diverse student populations, including General Education, Special Education, Students in Temporary Housing, and others through one integrated school transportation system. The system will utilize a flexible algorithm, which allows for both stop-to-school and home-to-school pickups, accommodating students regardless of their learning style, mobility constraints, or place(s) of residence.

Via’s technology is already in use in more than 50 markets across the globe, by such leading public sector transportation providers as Los Angeles Metro, Transport for London, and Berliner Verkehrsbetriebe (BVG). For Via for Schools, the company’s intelligent routing system will optimize school buses’ daily, fixed routes to be as efficient as possible, while also providing the flexibility to respond dynamically to realities on the ground such as street closures and traffic. The system is expected to significantly improve operational efficiency and aims to ultimately reduce transportation costs for the NYC DOE. Parents and students will benefit from improved visibility and communications regarding system status and unexpected changes, including real-time bus locations, student boardings and alightings, route changes, and vehicle delays.

About Via

Via is re-engineering public transit, from a regulated system of rigid routes and schedules to a fully dynamic and highly efficient on-demand network. Via’s technology is deployed in more than 20 countries worldwide through dozens of partner projects with public transportation agencies, private transit operators, taxi fleets, private companies, universities, and school systems, seamlessly integrating with public transit infrastructure to power cutting-edge on-demand public mobility.  First launched in New York City, the Via platform operates in the United States and in Europe through its joint venture with Mercedes-Benz Vans, ViaVan. For more information, visit ridewithvia.com."

Sunday, October 6, 2019

Conversation on repugnant markets, in French: Quand la science économique sauve des vies, conversation avec Alvin Roth, prix Nobel d’économie 2012

An account of my brief interview with The Conversation is published in both English and French.  It mostly concerned repugnant markets.  Here it is in French:
Quand la science économique sauve des vies, conversation avec Alvin Roth, prix Nobel d’économie 2012
September 23, 2019
Google Translate: "When economics saves lives, talk with Alvin Roth, 2012 Nobel Prize for Economics"

The interview was conducted in English (at the ESA meeting in Dijon), and appears in French translation.  For some reason, Google Translate translates the English word "repugnant" from the  translated French "répugnante" sometimes as "disgusting" and sometimes as "repulsive," both of which make the GT English translation hard for me to read...


And here it is in English, easier for me to read (although for some reason the editors have inserted an odd statement about the legality of same sex marriage in the U.S., which is now in fact legal in all fifty states, following a Supreme Court ruling...)

How economics can help save lives: a conversation with Alvin Roth, 2012 Nobel Prize laureate
September 24, 2019



Saturday, October 5, 2019

Auctions of poker and slot machine licenses in Victoria, by Plott, Cason, Gillen, Lee and Maron


The Design, Experimental Laboratory Testing and Implementation of a Large, Multi-Market, Policy Constrained, State Gaming Machines Auction
Charles R. Plott, California Institute of Technology
Timothy Cason, Purdue University
Benjamin Gillen, Claremont McKenna College
Hsingyang Lee, California Institute of Technology
Travis Maron, California Institute of Technology
(September 2019)

Abstract: The paper reports on the theory, design, laboratory experimental testing, field implementation and results of a large, multiple market and policy constrained auction. The auction involved the sale of 18,788 ten-year entitlements for the use of electronic gaming machines in 176 interconnected markets to 363 potential buyers representing licensed gaming establishments. The auction was conducted in one day and produced over $600M in revenue. The experiments and revealed dynamics of the multi-round auction provide evidence about basic principles of multiple market convergence found in classical theories of general equilibrium using new statistical tests of the abstract properties of tatonnement.


Friday, October 4, 2019

Eric Maskin on Rank Choice Voting (RCV) in Massachusetts and beyond

Eric Maskin, the Harvard University Professor and Nobel laureate in Economics, gave two talks at Stanford this week.  The first was the joint Economics/GSB economic theory seminar, in which he talked about his theoretical work on Condorcet voting and Borda count voting, two different voting systems in which voters record a ranking of candidates (rather than voting for a single candidate).  His work leads him to think that these two voting systems have some properties that recommend them to democracies.

His second talk was a political meeting, at the home of Susan Athey and Guido Imbens, in support of the Massachusetts Ranked-Choice Voting Initiative (2020)#.
He argues that, although in an ideal world he'd prefer Condorcet voting, he thinks that Ranked Choice Voting is a pretty good alternative, that has a greater chance of being adopted, and that we shouldn't let the perfect be the enemy of the good.

Voting is complicated, and Arrow's theorem assures us that no voting system is perfect, but I concluded the evening by making a modest contribution to the RCV initiative in Massachusetts, since RCV reduces the problems caused when multiple similar candidates split the vote in our current system of plurality voting.



Eric Maskin explains Rank Choice Voting

#the initiative:
*********
See my earlier post:

Monday, September 30, 2019

Thursday, October 3, 2019

Parag Pathak celebrated in Science News

Science News advises us that Parag Pathak is worth watching...

Parag Pathak uses data and algorithms to make public education fairer
After designing school choice systems, he’s studying student performance

(https://www.sciencenews.org/article/parag-pathak-sn-10-scientists-to-watch)


"If he could be granted one wish, he says he would design a school system from scratch. How, he asks, would you set up that system to be as equitable as possible?"


Wednesday, October 2, 2019

Repeal of motorcycle helmet laws increases deceased donor transplants: Dickert-Conlin, Elder and Teltser in AEJ:Applied

I imagine that a law that anyone who dies while riding a motorcycle without a helmet is automatically considered to be registered as a willing deceased organ donor would increase the voluntary use of helmets.  Here's a paper that investigates the relationship between helmet laws and transplants under current laws, which vary by state and over time.

Allocating Scarce Organs: How a Change in Supply Affects Transplant Waiting Lists and Transplant Recipients
By Stacy  Dickert-Conlin, Todd Elder, and Keith Teltser
American Economic Journal: Applied Economics 2019, 11(4): 210–239 https://doi.org/10.1257/app.20170476

Abstract: "Vast  organ  shortages  motivated  recent  efforts  to  increase  the  sup-ply  of  transplantable  organs,  but  we  know  little  about  the  demand  side  of  the  market.  We  test  the  implications  of  a  model  of  organ  demand using the universe of US transplant data from 1987 to 2013. Exploiting variation in supply induced by state-level motorcycle helmet  laws,  we  demonstrate  that  each  organ  that  becomes  available  from a deceased donor in a particular region induces five transplant candidates to join that region’s transplant wait list, while crowding out    living-donor  transplants.  Even  with  the  corresponding  demand  increase,  positive  supply  shocks  increase    post-transplant  survival  rates."


"We find that transplant candidates respond strongly to local supply shocks, along two  dimensions.  First,  for  each  new  organ  that  becomes  available  in  a  market,  roughly five new candidates join the local wait list. With detailed zip code data, we demonstrate that candidates listed in multiple locations and candidates living out-side of the local market disproportionately drive demand responses. Second, kidney transplant recipients substitute away from  living-donor transplants. We estimate the largest crowd out of potential transplants from living donors who are neither blood relatives  nor  spouses,  suggesting  that  these  are  the  marginal  cases  in  which  the  relative  costs  of  living-donor  and    deceased-donor  transplants  are  most  influential.  Taken together, these findings show that increases in the supply of organs generate demand behavior that at least partially offsets a shock’s direct effects. Presumably as  a  result  of  this  offset,  the  average  waiting  time  for  an  organ  does  not  measurably  decrease  in  response  to  a  positive  supply  shock.  However,  for  livers,  hearts,  lungs, and pancreases, we find evidence that an increase in the supply of deceased organs increases the probability that a transplant is successful, defined as graft survival. Among kidney transplant recipients, we hypothesize that living donor crowd out mitigates any health outcome gains resulting from increases in  deceased-donor transplants.
...
"The  SRTR  data  show  that  multi-listing  is  not  common,  with  only  6  percent  of  all  candidates  choosing  to  do  so  at  a  point  in  time  (online Appendix A describes how we identify multi-listed candidates and spells in the data). However, those who multi-list are systematically different from those who do not, with higher probabili-ties of having attended some college (46 percent versus 36 percent), higher rates of employment (44 percent versus 33 percent), and lower rates of insurance coverage via Medicaid (5 percent versus 11.5 percent). Not surprisingly, they are also more likely to register outside their own or a bordering DSA (12 percent) than candidates with a single listing (4 percent).
...
"the  percentage  of  liver candidates who receive a transplant within 5 years of listing ranged from 30.5 percent in New York to 86.1 percent in Arkansas (Israni, et al. 2012, 70). Similarly, “a striking (but not new) observation is the tremendous difference ... in the percent-age  of    wait-listed  patients  who  undergo  deceased  donor  kidney  transplant  within  5  years,”  varying  from  roughly  25  percent  in  California  DSAs  to  67  percent  in  Wisconsin (Israni, et al. 2012, 13).
...
"in  the  early  1970s  most  states  had  universal  helmet  laws  because  the  federal  government  tied  state  highway  construction  funds  to  such  laws  (Insurance Institute for Highway Safety (IIHS) 2018). By the  mid-1970s, states successfully lob-bied  Congress  to  break  that  link,  and  states  began  repealing  their  universal  helmet  laws (IIHS 2018).
...
"Using    state-level  OPTN  data  from  1994  to  2007,    Dickert-Conlin,  Elder  and  Moore (2011)—henceforth, DCEM—uses 6  state-level repeals and 1 enactment of a universal helmet law to estimate that repealing universal helmet laws increases the supply of organ donors who die in motor vehicle accidents by roughly 10 percent."
*********
This paper is part of an exciting line of work that I've blogged about earlier:

Thursday, August 1, 2019  How much do Kidney Exchanges Improve Patient Outcomes? Keith Teltser in AEJ-Policy


Tuesday, October 1, 2019

College athletes can be paid for endorsements in California: an end to an American repugnance?

College sports in the U.S. has long been an anomaly--our overseas colleagues must run colleges for reasons other than sports, but in the U.S. college sports are a big deal. And being a big-time college athlete leads to lucrative professional careers for some, but while in college the powers that be have long ruled that athletes can't be paid.  That may now be changing.

The NY Times has the story:

California Governor Signs Plan to Let N.C.A.A. Athletes Be Paid
Gov. Gavin Newsom signed a bill to allow college athletes to hire agents and make money from endorsements. The measure, the first of its kind, threatens the business model of college sports.

"The governor’s signature opened a new front of legal pressure against the amateurism model that has been foundational to college sports but has restricted generations of students from earning money while on athletic rosters.

"If the law survives any court challenges, the business of sports would change within a few years for public and private universities in California, including some of the most celebrated brands in American sports. So, too, would the financial opportunities for thousands of student-athletes, who have long been forbidden from trading on their renown to promote products and companies.

“Every single student in the university can market their name, image and likeness; they can go and get a YouTube channel, and they can monetize that,” Newsom said in an interview with The New York Times. “The only group that can’t are athletes. Why is that?
...
"In a statement on Monday, the N.C.A.A., which had warned that it considered the measure “unconstitutional,” said that it would “consider next steps in California” and that “a patchwork of different laws from different states will make unattainable the goal of providing a fair and level playing field.”

The state’s rebuke of a system that generates billions of dollars each year went against powerful universities, including California, Stanford and Southern California. The schools said the law would put their athletes in danger of being barred from routine competitions and showcase events like the College Football Playoff and the men’s and women’s N.C.A.A. basketball tournaments, made-for-TV moments that help some universities log more than $100 million each in annual athletic revenue."

Monday, September 30, 2019

Ranked-choice voting in Maine, and elsewhere

From the Atlantic:

A Step Toward Blowing Up the Presidential-Voting System
Maine’s adoption of ranked-choice voting for the 2020 general election could upend a close race for the White House.

"The 2016 presidential election pitted the two most disliked candidates in the history of public polling against each other. In the race between Donald Trump and Hillary Clinton, millions of Americans found themselves forced to vote for a major-party nominee they plainly couldn’t stand or to risk electing the candidate they hated even more by casting their ballot for a third-party contender.

"For the first time next November, a slice of the American electorate will have a way out of that lesser-of-two-evils scenario.

"With a law set to take effect in 2020, Maine will become the first state to adopt ranked-choice voting for a presidential election—a method in which people list candidates by order of preference rather than bubbling in just one circle. Maine controls only four electoral votes and splits them in half by congressional district, but the change could have huge consequences if the national presidential race to 270 electoral votes is close.
...
"The format works like an instant runoff: If no candidate receives more than 50 percent of the first-choice votes, the candidate with the least support is eliminated. Whomever that person’s voters picked as their second choice is then added to the tallies, and the process repeats until one candidate reaches a majority.
...
"The format played a big role in San Francisco’s high-profile mayoral race last year, and voters in New York City will decide in a ballot referendum this November whether to implement ranked-choice voting in future citywide elections.
...
"Kansas, Alaska, and Hawaii all plan to use ranked-choice voting in their primary or caucus, and Wyoming Democrats are considering it as well, says Rob Richie, the president and co-founder of FairVote, an advocacy group that has pushed for ranked-choice voting across the country.
***********

And here's a story from the Guardian:

America needs ranked choice voting – here's why
If more swing states introduced ranked voting, progressive candidates could challenge centrist Democrats without fear of aiding Trump

Sunday, September 29, 2019

In Memoriam: Martin Shubik

Ben Mattison and Kerry DeDomenico at Yale have sent around an email pointing to some brief remembrances of Martin Shubik from his friends, that was distributed at his memorial service.

In memoriam: Martin Shubik

 It's short, read the whole thing and think of Martin. 

I would have loved to read what Lloyd Shapley would have said about Martin, but I was delighted to read, in Pradeep Dubey's loving account, what Martin once said about Lloyd, after beating him in a game of Go: “No matter what happens in the game, Lloyd always wins the analysis.”

If you're in a hurry, here's what Bob Aumann had to say about Martin's professional contribution:

"I think the most important thing that can be said about Martin scientifically is that he is simply the father of the application of game theory to modern economic theory—an immensely important contribution. Von Neumann & Morgenstern’s book is called “Theory of Games and Economic Behavior,” and indeed they must be credited with the fundamental idea of studying economics with Game Theory tools; but their approach—Stable Sets—never really caught on. It was Martin who made the highly successful connection of the Core with the Competitive Equilibria, and this is what got started the ball rolling."
Robert Aumann
***************

Some earlier posts:

Thursday, December 20, 2012 Martin Shubik

Thursday, August 23, 2018  Martin Shubik, 1926-2018

Saturday, September 28, 2019

Automatic algorithmic affirmative action, by Ashesh Rambachan and Jonathan Roth

There's been some justified concern that algorithms that make predictions and choices based on previous choices made by humans might replicate the human biases embedded in the historic data.  Below is a paper that points out that the opposite effect could happen as well.

As explained here: "Imagine a college that has historically admitted students using (biased) admissions officers, but switches to an algorithm trained on data for their past students. If the admissions officers unfairly set a higher bar for people from group A, then assuming student performance is fairly measured once students arrive on campus, students from group A will appear to be stronger than students from group B. The learned model will therefore tend to favor students from group A, in effect raising the bar for students from group B."*

Here's the paper itself, and its abstract:

Bias In, Bias Out? Evaluating the Folk Wisdom
Ashesh Rambachan, Jonathan Roth

Abstract: We evaluate the folk wisdom that algorithms trained on data produced by biased human decision-makers necessarily reflect this bias. We consider a setting where training labels are only generated if a biased decision-maker takes a particular action, and so bias arises due to selection into the training data. In our baseline model, the more biased the decision-maker is toward a group, the more the algorithm favors that group. We refer to this phenomenon as "algorithmic affirmative action." We then clarify the conditions that give rise to algorithmic affirmative action. Whether a prediction algorithm reverses or inherits bias depends critically on how the decision-maker affects the training data as well as the label used in training. We illustrate our main theoretical results in a simulation study applied to the New York City Stop, Question and Frisk dataset.
**********

* I'm reminded of the saying "To get the same reward as a man, a woman has to be twice as good.  Fortunately that's not hard..."

Friday, September 27, 2019

More on the shortage of transplantable kidneys

Here are some snips from the transcript of the Undark podcast,
Solving the Deadly Transplantable Kidney Shortage
This month: A penetrating look at the trials of patients with kidney failure, and the doctors working to make more lifesaving transplants possible.


In the U.S., there are 58 local organ procurement organizations, more commonly known as OPOs. When someone is dying in the hospital with no chance of recovery, doctors will call their local OPO and set the organ procurement process into motion. The donor will go into surgery, their organs will be collected and the OPO will work to distribute the organs to people on the local waitlist. But for years, journalists and independent researchers have said these OPOs are not getting as many organs as they should be. Numerous studies and investigations have claimed OPOS could be recovering more than twice as many organs as they do now, if they were to opt for organs that were less-than-perfect, but likely still good enough.
Just a few weeks ago, a study published in the journal JAMA Internal Medicine looked at discard rates in France and in the United States. It found that OPOS in the U.S. discard at least 3,500 kidneys a year, nearly 20 percent of all deceased donor kidneys, as compared to 9 percent in France. These discarded kidneys are often from donors over 50 years old, or with curable diseases. But it’s hard to tell just how many organs we are missing out on because OPOS self-report their own numbers. And, according to a 2017 study published by the American Society of Transplantation and the American Society of Transplant Surgeons, some OPOS have even manipulated their numbers to appear better than they are.
In July, President Trump signed an executive order to launch an initiative called “Advancing American Kidney Health.” One of the plans is to order Secretary of Health and Human Services Alex Azar to reform the organ procurement process to increase the supply of transplantable kidneys
...
Kaitlin Benz: Highly complex because the transplant surgeons who decide whether or not to accept a less-than-perfect kidney have a lot to consider. The government evaluates the 261 transplant centers in the U.S. by their one-year post-op success rate, which generally ranges between 90 to 95 percent. Ideally, all of a program’s transplanted patients are still alive and well after a year, but that’s just not always going to happen. For their program to be considered successful, doctors need to have a high success rate, which means they have to closely consider how much risk they’re able to take on donor organs. What if they accept a less-than-perfect kidney and the patient dies six months later? Here’s Ron Gill.
Ron Gill: And so, ok, if I’m being measured on a one-year survival, I don’t want to take a kidney that has a greater risk of not working in a year. However, what’s dawning on us all is the comparator can’t be them not working as well. The comparator is the waiting list.
Kaitlin Benz: He says measuring success by one-year survival rates can disincentivize surgeons from even trying on those borderline, suboptimal kidneys that may not be perfect, but might give their patient a few more years of health and freedom than dialysis would.
Ron Gill: It kind of puts a stranglehold on innovation in my view. And many of us in the field feel that if you’re going to hold people to a very high standard and we keep losing so many people on the waiting list every year, what is it going to take to make that change? There are groups that are probably being punished for some of their lesser outcomes because they’ve taken greater risks. And again, we all probably made a mistake if we’re comparing their outcomes with other centers rather than comparing their outcomes with the waiting list.

Thursday, September 26, 2019

Padua Meeting on Economic Design and Institutions, September 27-28

Here is the web site:
Padua Meeting  on  Economic  Design  and  Institutions

Invited Speakers:
Salvador Barberà (Universitat Autonoma de Barcelona)

Dolors Berga (Universitat de Girona)

Guadalupe Correa Lopera (University of Malaga and University of Padova)

Albin Erlansson (University of Essex)

Matthew Jackson (Stanford University)

Louis Philippos (University of Cyprus)

Paolo Pin (University of Siena)

Massimo Morelli (Bocconi University)

Bernardo Moreno (Universidad de Malaga)

Paolo Roberti (Free University of Bozen-Bolzano)

Riccardo Saulle (University of Padova)

Fernando Vega Redondo (Bocconi University)

Sonal Yadav (Umea School of Business)

Wednesday, September 25, 2019

Lost Wages Support for Living Organ Donors Demonstration Project

HRSA (the Health Resources & Services Administration) has now funded a
Lost Wages Support for Living Organ Donors Demonstration Project.

It will be run by a consortium of organizations and administered through NLDAC (the National Living Donor Assistance Center).

NLDAC has also been running a randomized control trial sponsored by the Arnold Foundation:
Effect of Lost Wage Reimbursement to Kidney Donors on Living Donation Rates

***********
This is a developing story:)
Here are some of my earlier related posts

Tuesday, July 16, 2019

Tuesday, September 24, 2019

Algorithms and intelligence at Penn

From Penn Today:
The human driver
As the ability to harness the power of artificial intelligence grows, so does the need to consider the difficult decisions and trade-offs humans make about privacy, bias, ethics, and safety.

"Already, some AI-enabled practices have raised serious concerns, like the ability to create deepfake videos to put words in someone’s mouth, or the growing use of facial recognition technology in public places. Automated results that turned out to reflect racial or gender bias has prompted some to say the programs themselves are racist.

"But the problem is more accidental than malicious, says Penn computer scientist Aaron Roth. An algorithm is a tool, like a hammer—but while it would make no sense to talk about an “ethical” hammer, it’s possible to make an algorithm better through more thoughtful design.

“It wouldn’t be a moral failure of the hammer if I used it to hit someone. The ethical lapse would be my own,” he says. “But the harms that algorithms ultimately do are several degrees removed from the human beings, the engineers, who are designing them.”

"Roth and other experts acknowledge it’s a huge challenge to push humans to train the machines to emphasize fairness, privacy, and safety. Already, experts across disciplines, from engineering and computer science to philosophy and sociology, are working to translate vague social norms about fairness, privacy, and more into practical instructions for the computer programs. That means asking some hard questions, Roth says.

“Of course, regulation and legal approaches have an important role to play, but I think that by themselves they are woefully insufficient,” says Roth, whose book, “The Ethical Algorithm,” with Penn colleague Michael Kearns will be published in November.

The sheer size of the data sets can make transparency difficult, he adds, while at the same time revealing errors more easily."
*************

Listen also to
The programming ethos
In a podcast conversation, Penn professors Michael Kearns, Aaron Roth, and Lisa Miracchi discuss the ethics of artificial intelligence.

Monday, September 23, 2019

Private equity races for young talent even earlier this year

Eric Budish sends me this pointer to the continued unraveling of the recruiting of young investment bankers into private equity firms:

Private-equity firms are already interviewing 22-year-old bankers who will start in 2 years. Their earliest-ever hiring kickoff shows how crazy the battle for talent has gotten.  

"Private-equity firms are already interviewing first-year investment-banking analysts to fill 2021 associate positions, marking the earliest-ever kickoff to recruiting for those roles, sources told Business Insider.
...
"Last year, firms started interviewing in late October, recruiters said. This year, the PE firms are already moving in after analysts — typically 22-year-olds who just graduated from college the previous spring — who have only a few weeks of work experience under their belts.  

"Sources including academic advisers, recruiters, and insiders at PE firms told Business Insider that the activity was widespread, including at firms such as Thoma Bravo and TA Associates that were early movers last year but also at some of the largest firms including Warburg Pincus, TPG, and KKR. 
...
"Private-equity firms have pushed up the recruiting timeline over the past several years, despite how difficult it is to assess bankers so early in their careers. Still, they feel the need to remain competitive and get first dibs on the best talent.  "

Sunday, September 22, 2019

Neil Dorosin and school choice in Brooklyn, in the WSJ

Neil Dorosin was the director of high school operations for the NY Department of Education, back when Atila Abdulkadiroglu, Parag Pathak and I helped them design a school choice system for high school admissions. He later became the Pied Piper of school choice when he founded the Institute for Innovation in Public School Choice, which led the effort to redesign choice in a number of American cities. He still lives in NYC, and he's also a dad. The WSJ has made him the poster boy (poster dad?) for this year's middle school choice in Brooklyn (which apparently has some recently introduced random elements):

School-Choice Expert Has Unique Take on Brooklyn District’s New Admissions System
Neil Dorosin’s daughter went through the middle-school process and landed at a charter; ‘It was a complicated decision’
By Leslie Brody, Sept. 21, 2019

"His family’s choice gives a glimpse into how families grapple with decisions as Mayor Bill de Blasio ’s administration experiments with ways to better integrate one of the nation’s most segregated school systems. The School Diversity Advisory Group, appointed by the mayor, cautioned last month that if the city public schools lose students to private schools or other options, “it will become even more difficult to create high-quality integrated schools that serve the interests of all students.”

Here's the WSJ's accompanying photo of Neil:

And here's a picture of the four of us modeling casual wear in Stockholm in 2012.
Parag Pathak, Al Roth, Atila Abdulkadiroglu, and Neil Dorosin

Saturday, September 21, 2019

Reproductive technology and ethical dilemmas: are artificial wombs on the horizon? Will they change the meaning of abortion?

Assisted reproductive technology (ART) has brought us some modern possibilities that are sometimes viewed as repugnant.  In vitro fertilization has become a standard part of treatment for some kinds of infertility.  It also makes possible gestational surrogacy, in which the surrogate may or may not be paid, and the legality of both those things (surrogacy and commercial surrogacy) varies around the world.

There's a still nascent technology of artificial wombs--probably not coming to a hospital near you anytime soon--that raise questions about abortion.  But it's not too early to ask if a new technology could help resolve an old ethical question (while perhaps creating new ones...).

The NY Times takes up the story:

The Abortion Debate Is Stuck. Are Artificial Wombs the Answer?
The technology would allow fetuses to develop outside the female womb so women would no longer have to be pregnant.  By Zoltan Istvan

"Could an emerging technology reshape the battle lines in the abortion debate? Since Roe v. Wade was decided in 1973, that fight has been defined by the interlocking, absolute values of choice and life: For some, a woman’s right to choose trumps any claim to a right to life by the fetus; for others, it’s the reverse. But what if we could separate those two — what if a woman’s choice to terminate a pregnancy no longer meant terminating the fetus itself?
...
"Artificial human wombs are still far in the future, and there are of course other ethical issues to consider. But for now, the technology is developed enough to raise new questions for the abortion debate.

"In a 2017 issue of the journal Bioethics, two philosophers, Jeremy V. Davis, a visiting professor at the United States Military Academy at West Point, and Eric Mathison, a postdoctoral associate at Baylor College of Medicine, argue that while a woman has a right to remove a fetus from her body, she does not have the right to kill it. The problem is that, for now, the latter is inherent in the former.

"Their argument builds upon that of the pro-choice philosopher Judith Jarvis Thompson, who famously argued in her 1971 paper “A Defense of Abortion” that women have a right to not carry a fetus for nine months — but that women do not have a right to be guaranteed the death of the fetus.
...
"Biobag technology could be available for humans in as little as one to three years, according to Dr. Alan Flake, a fetal surgeon in charge of the Children’s Hospital of Philadelphia artificial womb experiments. Another team performing ectogenesis research at the University of Michigan also believes they could have devices ready for humans in a similar time frame."
**************

In an article published in 2017, also in the NYT, Dr Flake thought these questions would arise only in the more distant future:
Weighing the Ethics of Artificial Wombs

"Dr. Flake agreed that what the field did not need was another intervention for premature infants that creates more problems than it solves. “This system will either work and work very well, or I won’t apply it,” he said."

Friday, September 20, 2019

Paul Romer thinking about market design via urban planning, at Burning Man

The NY Times has the story:

A Nobel-Winning Economist Goes to Burning Man
Amid the desert orgies, Paul Romer investigates a provocative question: Is this bacchanal a model of urban planning?
By Emily Badger

"White-haired and 63, he was dressed in black gear he’d bought at R.E.I., figuring black was the thing to wear at Burning Man. It was the first time that Mr. Romer, the former chief economist of the World Bank, had attended the annual bacchanal.
...
"Urbanization in the developed world has largely come to an end; nearly everyone who would move from farmland toward cities already has. This century, the same mass migration will run its course across the rest of the world. And if no one prepares for it — if we leave it to developers to claim one field at a time, or to migrants to make their way with no structure — it will be nearly impossible to superimpose some order later.

"It will take vast expense, and sweeping acts of eminent domain, to create arterial roads, bus service, trash routes, public parks, basic connectivity.

"That prospect agitates Mr. Romer, because the power of cities to lift people out of poverty dissipates when cities don’t work. To economists, cities are labor markets. And labor markets can’t function when there are no roads leading workers out of their favelas, or when would-be inventors never meet because they live in gridlock."

Thursday, September 19, 2019

History job market conference interviews are history

Inside Higher Ed has the story on the history job market (which they conflate with the AEA's recent decision to try to eliminate interviews in hotel rooms):

Killing the Conference Interview
American Historical Association ends annual meeting interviews and American Economic Association ends single hotel room interviews.
By Colleen Flaherty

"It's official: the American Historical Association will stop supporting first-round job interviews at its annual meeting.


"The group floated the idea this spring, citing a decline in registered departmental searches -- from 270 for the 2005 conference to 20 this year -- and a desire to take the meeting in new directions.
"After hearing overwhelming positive feedback from members, the AHA Council voted to end the 70-year-old tradition."
*************

I'm not intimately familiar with the History job market, but for economists, I think the tradition of interviewing at the January meetings has had a good effect on the job market, helping to coordinate timings, reduce costs, and provide a thick early part of the market.  I hope that we won't be starting on the road to moving interviews elsewhere and (particularly) at earlier and more diffuse times.

Wednesday, September 18, 2019

The Department of Justice opposes limits on early admissions, and other admissions agreements among colleges

Forbes has the story:

The Department Of Justice Aims To Unravel The College Admission Market
 Brennan Barnard

"Thanks to a two-year, ongoing investigation by the Antitrust Division of the United States Department of Justice (DOJ), the wheels are about to come off in college admission. As the National Association for College Admission Counseling (NACAC) prepares to meet later this month for their annual conference, the leadership reached out to members last week about proposed changes to the Association’s Code of Ethics and Professional Practice (CEPP). These potential amendments are a direct result of fruitless conversations with the DOJ, which have left NACAC with few options.
...
"Specifically the DOJ has taken issue with ethical guidelines that prevent colleges from “offering exclusive incentives for Early Decision, recruiting first-year undergraduates who have committed elsewhere, and recruiting transfer students.”
**********

Regarding early admissions, the DOJ wants colleges to be able to compete more vigorously through early admissions, e.g. by offering special access to dormitories, or other perks to students who commit early.  It will be interesting to see where this leads, but it could easily lead to more unraveling of admissions, making more admissions decisions earlier.

Here's the relevant page from NACAC, the National Association for College Admissions Counseling:

2019 Assembly Meeting Background
NACAC’s Code of Ethics and Professional Practices and Antitrust Provisions
Kentucky International Convention Center
Saturday, September 28, 2019

Tuesday, September 17, 2019

West Point: adopts two sided matching for cadets to military branches

From the Army Times (the links in the story are worth clicking on for related detail):

West Point has changed how cadets are assigned branches — ROTC will soon follow

"West Point’s class of 2020 will serve as the first users of a new branch assignment system this fall, which the Army hopes will help with the retention of junior officers and better assign talent.
The Army is rolling out a new “Market Model branching system" that takes input from the commandants of each branch, who rank cadets as “most preferred,” “preferred” or “least preferred,” according to an Army news posting.
The number of cadets allocated into each of the three ranks depends on the branch’s needs.
The program starts with West Point cadets receiving their branch assignments this November, but will eventually be used across the service’s Reserve Officer Training Corps detachments next year.
Cadets will be judged based on a range of factors, including test scores, physical fitness scores, transcripts, personal statements and, for the first time, interviews with the branches they’re interested in.
Rankings and preferences will decide branch assignments using a variant of the same algorithm that medical school graduates use to be assigned to residency programs across the country.
This is the first time commandants from the Army’s 17 branches will have a say in which cadets come into their branches, the service said in its announcement. Previously, cadets simply ranked the branches in order of preference and received their assignment “based almost entirely on their ranking in the Order of Merit List," the Army’s posting reads.
The Army release also notes that the process allows for cadets to take on a Branch of Choice Active Duty Service Obligation, or BRADSO. This allows for West Point graduates to serve an extra three years on top of five they’re already obligated to serve in exchange for increasing the odds that the cadet will receive the branch they most desire.
BRADSO doesn’t change how well the branch commandant ranks a cadet, but it does move cadets within their own ranking.
“If you’re cadet number 25 in that most preferred bucket, and I’m cadet number 45 and I’m willing to BRADSO and you’re not, I move above you,” Maj. Jared Sunsdahl, West Point’s accessions division chief, said in the Army posting. “Now, 45 is above 25 and then depending on how many branch allocations there are, you may not have received that branch because there were only so many allocations left.”
Therefore, it won’t take a cadet from being “preferred” by the branch commandant to being “most preferred,” but it will increase their odds against other “preferred” cadets.
West Point’s class of 2020 will lock in their final branch rankings between Sept. 23-29. Branch commandants have to lock in their rankings by Sept. 19."
**********

HT: Tobias Switzer