Monday, May 22, 2017

Broken Chains and Reneging: A Review of 1,748 Kidney Paired Donation Transplants

Broken chains in non-simultaneous kidney exchange chains begun by non-directed donors are not a big problem at all.
Here's the article, forthcoming in  the American Journal of Transplantation:

Broken Chains and Reneging: A Review of 1,748 Kidney Paired Donation Transplants
Nick Cowan, H. Albin Gritsch, Nima Nisseri, Joe Sinacore, Jeffrey Veale
Accepted manuscript online: 10 May 2017
DOI: 10.1111/ajt.14343

Abstract: Concerns regarding the potential for broken chains and reneges within kidney paired donation (KPD) and its effect on chain length have been previously raised. While these concerns have been tested in simulation studies, “real world” data has yet to be evaluated. The purpose of this study was to evaluate the actual rate and cause of broken chains within a large KPD program. All patients undergoing renal transplantation through the National Kidney Registry from 2008 through May 2016 were included for analysis. Broken chains and loops were identified. A total of 344 chains and 78 loops were completed during the study period yielding a total of 1,748 transplants. Twenty broken chains and one broken loop were identified. The mean chain length (# of transplants) within broken chains was 4.8 compared to 4.6 of completed chains (p=0.78). The most common causes of a broken chain were donor medical issues incurred while acting as a bridge donor (n=8), donors electing not to proceed (n=6), and kidneys being declined by the recipient surgeon (n=4). All recipients involved in a broken chain have subsequently received a transplant. Based on the results broken chains are infrequent, rarely due to lack of donor motivation, and have no significant impact on chain length.

Sunday, May 21, 2017

New York City school choice in the NY Times: Not all NYC high schools are good yet

The NY Times recently ran this story, largely critical of school choice in NYC:
The Broken Promises of Choice in New York City Schools
The city’s high school admissions process was supposed to give every student a real chance to attend
a good school. But 14 years in, it has not delivered.

Here's a paragraph that summarizes the main point:
"Ultimately, there just are not enough good schools to go around. And so it is a system in which some children win and others lose because of factors beyond their control — like where they live and how much money their families have."

The story follows several students at a  middle school in the Bronx:
"The Times spent months following the high school application process at Pelham Gardens, where families do not have the advantages that routinely open doors to the city’s best schools. Many families are new to the country, and most are poor."

Parag Pathak (who played a critical role in organizing the NYC high school match--see e.g. here and here) wrote a letter to the NY Times summarizing his reaction to the story. As it appears that the Times won't publish the letter, he gave me permission to reproduce it:

"May 5, 2017

In “The Broken Promises of Choice in New York City Schools,” Elizabeth Harris and Ford Fessenden miss a key point in describing the New York City High School choice system. The choice system does not create good schools.  It exists because there aren’t enough good schools.

I worked with NYC DOE to design the choice system described by authors.   By any objective measure, this system provided better access to schools than the one it replaced.  Without a comparison to the old system, Harris and Fessenden’s description of choice outcomes is misleading.  In the old system, half of applicants from Pelham Gardens (zip code 10469) were assigned to choices they did not rank; in 2003, that number drops to 23%. Under the new system, students from that neighborhood also travelled two miles further to schools they wanted.  Across the city, the new system allows more kids to go to schools they ranked and the benefits were largest for those most likely to be administratively assigned, like those in Pelham Gardens (see,   This is not to say the process is perfect and couldn’t be improved.  But it is foolish to expect the process to produce miracles, without changing the set of school options.

A broader premise of the article is that schools with highest test scores and graduation rates are indeed “the best.”  Our research, including on NYC’s exam schools (, strongly suggests otherwise.  Naïve discussions of school quality and the role of school choice confuse efforts to improve school quality, where our attention should be devoted. 

Parag Pathak
Carlton Professor of Economics, MIT"

For reference, here's the 2003 NY Times story that covered the school choice system when it was introduced:

Saturday, May 20, 2017

Discovering Prices, by Paul Milgrom

Just published and well worth reading (here's a link at which you can discover its price:)
Discovering Prices  -- Auction Design in Markets with Complex Constraints
Paul Milgrom
Columbia University Press

Friday, May 19, 2017

School choice in San Francisco faces ongoing problems (and not just school choice)

The SF Chronicle has the story:

Why San Francisco needs a full-time school board
By Gail Cornwall, May 17, 2017  

"Ever wonder why the pace of change in public education falls somewhere between inching and crawling in arguably the most progressive, innovative city in the world? San Francisco Unified School District’s red tape and lack of resources are to blame, but there’s also a story of unpaid workers, organizational mutiny and missed opportunity.

Here’s an example: In 2009 the school board set out to redesign its method for assigning students to schools. Though the topic sounds dry, matching thousands of children to seats at more than a hundred programs — while taking into consideration parental preference, geography, diversity and more — involves the sexiest corner of economics: game theory.

Luckily, the board had the assistance of a group representing Harvard, Stanford, Duke and MIT. Nobel Prize winner Alvin Roth, Muriel Niederle, Clayton Featherstone and others proposed helping to create, monitor and adjust a cutting-edge algorithm for free. In March 2010, the board voted unanimously to take the offer.

But that September, district staff sent Roth’s team an email amounting to “Thank you, goodbye.” District officials had decided to instead “develop software to implement the new design on their own,” Roth reported.

Today, the state of the district’s homegrown assignment algorithm, known to parents as “the lottery,” is described by board member Mark Sanchez as “broken” and “untenable,” and by board member Rachel Norton as “probably the biggest policy issue that our community engages with us on.”

Neil Dorosin directs the Institute for Innovation in Public School Choice, a nonprofit Roth and his team formed. When I recently asked Dorosin what kind of personnel would be needed to create an effective school assignment algorithm, he said, “Either a mathematician or an economist who knows about algorithms, and … a software engineer who could operationalize it. I would be stunned if they have that.”

Those who shared these concerns back in 2010 called on district staff to explain themselves. Despite making a pledge to the board to disclose the algorithm developed in-house, by March 2012 staff still hadn’t issued “a complete enough description to [know] … if they in fact implemented the plan … the board adopted,” said Roth.

Lack of compliance with board directives sounds crazy, but Sanchez, who served on the board from 2001 to 2009 and won re-election in November, said it happens all the time. “There are so many examples,” he said.

How could that be possible? Because board members each receive “about $6K a year — and everyone has a full-time job doing something else — they’re just too busy to check in and cajole, Sanchez said. The only thing the board really can do, he said, is fire the superintendent when “a lot of that piles up.”

That’s why Sanchez and San Francisco Supervisor Jane Kim have discussed putting forward a ballot measure to increase school board member compensation. Sanchez said it would give board members “the average beginning pay for a teacher in the Bay Area ... probably ending up at around $45,000” (drawn from the city’s budget, rather than the school district’s). Following the model adopted by Los Angeles in 2006, the full amount would be available only to those who forsake other employment, he said.

Meanwhile, parents fret over the lottery. The long, complex application process — where paperwork is submitted in person in January and decisions are issued in March, then again in May, through three more rounds of supplication and the first two weeks of school — fails low-income families who lack the time or bureaucratic savvy to effectively engage. Those who do manage to navigate the process, one emotional parent told the school board committee Monday night, often find the experience “time-consuming, frustrating and stressful.” Raman Khanna, a member of the Ulloa Elementary School PTA, referred to another outcome: professional “flight.” Because of the lottery, he said, “a lot of the colleagues that I talk to … leave the city or they go to private school.”

Roth’s and Dorosin’s organization has worked with cities across the country to use data and technology to improve school assignment. Dorosin said the nonprofit’s modest fees are often covered by outside grants and other funding. This March they invited SFUSD board members and district officials to reach out again.

The response? The board committee announced Monday it would “not be taking action,” and district staff proposed two timelines for reform: one would give the board two years to articulate a new direction for the assignment system; the other, labeled “if policy development moves quickly,” would still give them a full year to do so and then another 18 months for district staff to implement it. Tommy Williams, a parent who works for the U.S. Army Corps of Engineers, said: “The fact that we’re talking about the 202[0] or 2021 school year is very frustrating.”

Board members could start meeting with national experts with just days’ notice to hash out a broad-strokes plan, but that won’t happen until the fall. “It’s clearly an urgent issue,” Sanchez said, “but it’s one of many, many things that we have to deal with. ... We had to hire a superintendent, and now we’re involved in negotiations for the contract so we wanted to focus on that.”

Rionda Batiste, co-chair of the district’s African American Parent Advisory Council, won many approving head nods at Monday’s meeting when she said: “I don’t understand why this is something that cannot happen simultaneously.”

“If we really wanted to speed things along,” Sanchez told me, “we’d have more meetings. Intuitively, we all know [paying board members] would make things move faster.”

Nationally, school board compensation is all over the board. Connecticut pays nothing, while Florida’s lowest paying county, as of 2014 reportedly offered $24,290 a year. According to the National School Boards Association, approximately 75 percent of small-district school board members serve as volunteers while around 40 percent of large-district ones receive a “modest salary.”

This divide makes sense, because there’s much more work to be done in a district with 50,000 students than one with only a few hundred. When being an effective board member requires a full-time commitment from someone who must already work full-time elsewhere, “it’s a structural problem,” Sanchez said.

A second one resurfaced at Monday’s meeting. Orla O’Keefe, the district’s chief of policy and operations, told the board: “We need a larger number of staff with the technical skills and knowledge needed to complete assignment runs,” including to “[e]xplore leveraging district ... online registration functionality with a potential online application pilot.” In other words, while the board takes its time deciding what major changes to make, district staff propose once again building in-house. Meanwhile, a Nobel Prize-winning economist — and the tools his team has honed — wait in the wings.

Maybe instituting board salaries can buy our elected representatives the time they need to pursue public-private partnerships that bring expertise and manpower to the task of matching students with schools. Hopefully, this time it will be with the support of district staff, such as newly anointed Superintendent Vincent Matthews who, calling the meeting “democracy in action,” said Monday he’s “looking forward to moving this forward.”

Until then, Sanchez said, “It’s in a holding pattern.”

Thursday, May 18, 2017

School Choice in the District of Columbia: apparently not all according to algorithm

The Washington Post has the story: it seems that some politically connected families were able to get the schools they wanted without going through the school choice algorithm...

Secret report shows ‘special’ treatment for public officials in D.C. school lottery

"Former D.C. Public Schools chancellor Kaya Henderson routinely helped well-connected parents — including two senior aides to Mayor Muriel E. Bowser (D) — bend or break the rules of the District’s notoriously competitive school lottery to enroll their children at coveted schools, according to a confidential report obtained by The Washington Post.

The report, based on an investigation by the D.C. Inspector General’s Office, describes in remarkable detail how Henderson used her power as head of the school system to place the children of those with political clout at campuses they could not otherwise access through the random lottery, which every year leaves thousands of families on waiting lists for their desired schools.

Inspector General Daniel Lucas found that Henderson misused her authority by giving preferential treatment to seven of 10 people who requested special school placements for their children during the 2015 lottery season. The investigation did not examine the rest of Henderson’s tenure from November 2010 to September 2016.

[Read the confidential inspector general’s report obtained by The Post]

Henderson openly acknowledged in interviews with investigators that she gave special treatment to the children of government officials. Asked about the help she gave City Administrator Rashad M. Young, a top Bowser cabinet official whose salary is $295,000, Henderson said D.C. officials “do not necessarily get paid as much as we should.”

The former chancellor bestowed such favors even as she dismissed pleas for special consideration from those with less influence, such as a deaf Vietnamese immigrant whose request that her daughter be allowed to attend a school where she could practice sign language was rejected.

The findings could shake public confidence in the city’s school lottery, which has been held up as a national model. It also raises troubling questions for Bowser, whose defense of her cabinet officials’ roles in the scandal is undercut by details in the report."
"[This is how the D.C. lottery is supposed to work]".
And here's a story from 2013 as the school choice system was introduced:
D.C. is preparing a unified enrollment lottery for its traditional and charter schools

IIPSC designed the algorithm, but not the politics.

Wednesday, May 17, 2017

Dialysis: Last Week Tonight with John Oliver (HBO)

Laughing through the tears...

The video briefly mentions a new living donor information source and database, Give and Live.

Tuesday, May 16, 2017

Replication in Economics

Professor Bob Reed at U of Canterbury points me to The Replication Network, a website he co-founded with Dr Maren Duvendack of  U of E. Anglia to promote replication in Economics, and assemble information about replication studies.

See also their paper and the others in the  Papers and Proceedings (May 2017)  issue of the American Economic Review, which begins with a section on replication: