Tuesday, September 18, 2012

Market design course at Stanford (Econ 285, Autumn 2012)

Muriel Niederle and I will be teaching an introduction to market design this quarter (the first quarter of a three quarter graduate sequence whose other quarters will be taught by Paul Milgrom and Fuhito Kojima).

The first class session is on Monday September 24.

This will be the first course I've taught at Stanford since 1978 (when I taught a course on Axiomatic Models of Bargaining, while on leave from the University of Illinois), and it will likely resemble the market design course I taught last Fall at Harvard. You can find the web page for that course, which includes the slides I lectured from here. (Since this will be followed by a quarter taught by Paul Milgrom, we plan to spend less time on auctions than when I taught at Harvard, and more time on matching markets: see the course description below.)

We think the class might be interesting not only to economists but also to operations researchers and computer scientists...

ECON  285 - 01   Market Design
Stanford University | 2012-2013 Autumn | Lecture
Class Details
Status
Open
Open
Class Number
34287
Session
Regular Academic Session
Units
2 - 5 units
Lecture
Required
Class Components
Career
Graduate
Dates
9/24/2012 - 12/7/2012
Grading
Letter or Credit/No Credit
Location
Stanford Main Campus
Campus
Stanford Main Campus
Meeting Information
Days & TimesRoomInstructorMeeting Dates
MoWe 11:00AM - 12:50PM
Econ 106
Alvin Roth,
Muriel Niederle
09/24/2012 - 12/07/2012
Class Availability
Class Capacity
Wait List Capacity
17
0
Enrollment Total
Wait List Total
10
0
Available Seats
7
Description
This is an introduction to market design, intended mainly for second year PhD students in economics (but also open to other graduates students from around the university and to undergrads who have taken undergrad market design). It will emphasize the combined use of economic theory, experiments and empirical analysis to analyze and engineer market rules and institutions. In this first quarter we will pay particular attention to matching markets, which are those in which price doesn¿t do all of the work, and which include some kind of application or selection process. In recent years market designers have participated in the design and implementation of a number of marketplaces, and the course will emphasize the relation between theory and practice, for example in the design of labor market clearinghouses for American doctors, and school choice programs in a growing number of American cities (including New York and Boston), and the allocation of organs for transplantation.  Various forms of market failure will also be discussed.
Assignment:  One final paper. The objective of the final paper is to study an existing market or an environment with a potential role for a market, describe the relevant market design questions, and evaluate how the current market design works and/or propose improvements on the current design.

Monday, September 17, 2012

Summer law associates: offer rates

The blog Above the Law is tracking law firms which give permanent offers to 100% of their summer associates, before they head back to their third year of law school.

They compiled a similar list for 2011, and updated it here.

That update ends as follows:


One caveat: note that these 100 percent offer rates might include so-called cold offers, in which a firm makes an offer to a candidate, but suggests that perhaps the candidate should not accept it. E.g., “We’re making you an offer [so we can boast about our 100 percent offer rate], but we think you might be happier elsewhere [wink wink], so you might want to look into the 3L recruiting process [don't come here unless you want to work out of a utility closet].”
Cold offers are frowned upon in many quarters. Here is what NALP has to say about them:
11. Cold or Fake Offers
Q. It is reported that some employers give offers to summer associates with the understanding that the offer will not be accepted. What is NALP’s view of this practice?
A. NALP does not condone this unethical practice. Whether initiated by students to appear more attractive to future employers or by employers to enhance their offer ratios, the practice is fraudulent and unprofessional. NALP suggests that employers adopt a standard policy of extending or confirming offers in writing, signed by a representative of the organization, so that only legitimate offers are made.

Sunday, September 16, 2012

Miscellaneous School choice in the news

Here are some stories that I marked but never got around to blogging about...

Denver:DPS busts parents lying for seats

"DPS officials confirmed this week they’re analyzing address anomalies to root out people misleading the district about their home addresses as a way to ensure placement of their children in coveted schools, such as East High School, the Denver Schools of Science and Technology, and Bromwell, Steck, Stapleton and Cory elementary schools."
***********

New Orleans:
Recovery School District school assignments appealed by about 5 percent of students

"About 5 percent of the 32,000 students assigned to public schools this year through the Recovery School District's new central enrollment system appealed their assignment, district officials said Tuesday. Of those, about 69 percent got into one of their top three choices of school after the district reconsidered their case.

"On the other hand, about 31 percent of the roughly 1,400 students who appealed got matched to a school that they hadn't ranked on their application at all, meaning about 430 families are likely still feeling frustrated.

"We need more good and great schools to give parents more options," Recovery District Superintendent Patrick Dobard said. "That's what we look at these numbers as speaking to."

"This spring marked the first time that the Recovery District has offered one application, known as the OneApp, for a slot at any district school. In the past, each of the city's independent charter schools, which make up most of the district, handled enrollment on their own, presenting parents with a sometimes frustrating search for an open seat. None of the district's schools have traditional neighborhood attendance zones.

"With the OneApp in place this year, the district took in about 25,000 applications. Technically, everyone is required to turn one in, whether they plan on staying at the school they already attend or not. Overall, the district assigned 32,000 students to schools, keeping those who didn't turn one in at their current schools and manually assigning those who didn't turn in an application but were scheduled to finish their school's last available grade -- eighth graders slated to move on to high school, for instance."
**************


Orleans Parish School Board approves resolution moving toward common enrollment

"The city's local school board made a formal commitment Tuesday to work with state officials on a common enrollment process for all of the city's public schools. It's a step that could simplify life for families in a city where parents have both the burden and opportunity of choosing a school on their own, rather than letting geography decide.

"But negotiations between local and state education officials, as well as the independent charter schools that now predominate in New Orleans, may still have far to go before a common enrollment system is in place.

"In a unanimous vote, the Orleans Parish School Board approved a resolution promising to work toward joining the so called OneApp, a central enrollment system deployed for the first time this year by officials at the Recovery School District, the state agency that has governed most of the city's public schools since shortly after Hurricane Katrina.

"Bringing all, or at least the vast majority, of the city's public schools into one enrollment system would mark a significant reversal. In Katrina's wake, the state swept most schools into the Recovery School District and began handing over the reins of individual schools to independent charter operators that handle their own process for enrolling students, drawing applicants from across the city rather than a single neighborhood. The School Board, left with about 17 campuses, is now a majority charter district as well, leaving parents with a patchwork of varying applications and deadlines.

"Responding after years of complaints about the complexities of the new system, the Recovery District put all of its traditional and charter schools into the same enrollment system this year. Experts from Duke, Harvard and MIT, helped engineer a computer-driven matching scheme that assigns students based on their top choices, the school their siblings attend, their proximity to the campus and a randomly assigned lottery number.
The School Board, which governs a district encompassing about a quarter of the city's public school students and all of the city's magnet schools, sat the OneApp's first round out, concerned about ceding further authority to the state without being assured of a fair process.

"Quiet negotiations between the two districts have yielded an apparent compromise. As it's spelled out in the School Board's resolution, the board will be taking over management of the OneApp itself, although when and on what terms isn't clear, leaving room for any potential agreement to fall apart.

"Officials from either district tell slightly different versions of how they see a merger of enrollment systems playing out. In an interview on Tuesday, Recovery District Superintendent Patrick Dobard said he's hoping to incorporate School Board campuses into the OneApp as early as next year -- before the Recovery District would be ready to hand over the keys of that system for the School Board to run. Dobard said any transition toward School Board control of the OneApp would likely be settled with some type of written contract, likely years down the road and as a part of a broader discussion of bringing all of the city's schools back under some type of unified governing body.
"We don't have to be the long-term manager," Dobard said. "But we would likely have to some kind of memorandum of understanding."

"Thomas Robichaux, the School Board's president, however, said Tuesday, "It's our intention to manage it right away, or at the very least to be the joint manager."

"The School Board's 11 independent charter schools -- publicly funded but operating autonomously in a contract with the district -- may present another hurdle. Robichaux has said that the district cannot force those schools into a common enrollment process, so officials will have to negotiate with each of them.

"One other caveat, even if the Recovery District and the School Board succeed in unifying their enrollment, will be the city's so-called Type 2 charters, authorized by the state and operating outside the jurisdiction of either district. Recovery District officials have said they would like to include Type 2s in the OneApp as well, but there has been no public move in that direction.

"Still, Dobard said he's optimistic about getting a deal done, with the School Board at least. Recovery District officials say there are technically problems to overcome. They will need a new computer algorithm to incorporate the selective admissions requirements at some School Board campuses. But none of it is beyond working out. "
*************

Washington D.C.


Create one single lottery for charter and non-charter schools

"The current application process for DC's charter and non-charter public schools is a chaotic mess that confuses parents and hurts education for students. DC could fix many problems by creating a centralized lottery process for all public schools, charter and non-charter.
"Steve Glazerman called for a centralized application for charter schools in 2010. Since then, DC Public Schools (DCPS) instituted a common application for the District's specialized high schools.
This is a great step, but it could go a lot further to include charter schools and traditional neighborhood schools at all grades. It wouldn't be hard; the company that operates whose software enables the centralized application for DCPS application-only high schools is currently implementing a centralized application for charters and non-charters in Denver.
District officials generally agree. Scheherazade Salimi, Senior Advisor to the Deputy Mayor for Education, says that "a common application is something the Deputy Mayor would like to explore in partnership with DCPS and [the Public Charter School Board]."
In a centralized application, parents would select several schools, rank-ordered by preference. They would select charters and non-charters, and could conceivably select up to 12, 15 or 20 schools.
A single lottery would select applicants one by one, and assign each to the first school on his or her list with an open slot. This is similar to how many colleges assign dorm rooms, for instance.
This type of centralized application would have many benefits over the current system.
Parents are more likely to get into their top choice schools.
When parents apply to schools now, they apply for DCPS schools using a centralized application, and apply to each charter school separately. Pre-K programs have lotteries for all children, while students in 1st grade and older enter lotteries only for out-of-boundary DCPS schools.
As a result, one applicant in Capitol Hill could be waitlisted at a nearby charter that was their top choice and accepted into a Columbia Heights charter that was their 2nd choice, while a Columbia Heights family that preferred the nearby charter could be waitlisted there but accepted to the Capitol Hill charter school.
The result is that neither child can go to his or her top choice charter, and both families are making unnecessary drives to get the kids to school.
Spots at competitive schools won't be locked up by parents who don't plan to send kids there.
Schools hold their lotteries in the spring for spots in the fall. In the current system, if a child gets accepted to multiple charter schools and/or an out-of-boundary DCPS school, parents might tell each school that the child will attend in the fall.
When they decide which school to attend, they inform the schools at some point in the summer or they just don't show up for the schools they didn't select. There's no deposit or penalty, so they don't pay a cost for this, but other families lose out who might have taken the slot but had to make a decision earlier to go elsehwere.
Some parents do this to give themselves more time to research the schools; some want to wait until school starts to assess the facilities of charter schools that were still preparing their facilities in the spring.
When a student attending an out-of-boundary DCPS school gets into a different out-of-boundary DCPS school, the principal of the first school "releases" the student before they can secure their spot into the new school. Charter schools have no such process.
Squatting on multiple school slots is unfair to everybody. When children accepted through the lottery don't show up in the fall, principals have to scramble to contact any remaining applicants on their wait list. Squatting also leads to the next problem.
Principals could provide better estimates of enrollment for funding purposes.
One of the most common grievances from charter advocates is that DCPS principals overestimate their enrollment to receive extra funding.
DCPS schools project fall enrollment in the spring and these projections determine funding for the following year. If the actual enrollment is lower, DCPS' budget doesn't shrink. But charter schools receive funding quarterly based on their actual enrollment. If a charter school's enrollment declines, it loses money.
Some principals might be doing this on purpose, but it's also difficult for DCPS principals to accurately estimate enrollment for the following fall when applicants hold a spot at their school while they spend the summer deciding whether to attend charter schools.
A centralized application would eliminate much of this problem. Each school, DCPS and charter, would know that every child on its list isn't going to suddenly go elsewhere in the DC system. They could go to private school or move to another jurisdiction, but that applies to a smaller number of children.
Charter principals wouldn't be able to "skim the cream."
Charter school critics often complain that charter school principals find ways to weed out students during enrollment who may be harder to educate. The lottery initially fills all charter school slots randomly, but as parents of children who got in on the lottery tell the school that they won't be taking the slot, the charter itself contacts applicants off of their wait list.
There are opportunities for principals to intentially or unintentionally abuse this system. For example, principals can give an applicant more or less time to respond and claim the slot before they move on to the next child. They might give more "desirable" children more time than others.
A charter school in New York was put on probation last year for weeding out applicants in the enrollment process. While there hasn't been a specific accusation like this against any DC charter school, a centralized application system could remove this because students would be assigned to a single school.
We would have data on capacity needs at all grades, especially pre-K.
District officials say that DC has achieved universal pre-K, but the city's auditor of pre-K capacity disagrees. Who is right? We won't know until we have data on the actual demand for pre-K.
A centralized application for pre-K, including all of the pre-K programs, would generate this data. It would then be easy to compare the number of total children applying against the number of public pre-K slots.
The data wouldn't be perfect, as some parents apply to DCPS pre-K programs as a backup to their private pre-K applications, while other parents miss the pre-K lottery (in February) but still want to send their children to pre-K. But it would be far better than the current audit, which effectively measures nothing.
All students would start school on time together.
One of the unintended consequences of the plethora of charter school choices is that schools don't really know who will show up for school in September. This is largely due to parents holding spots through the summer for multiple schools but only sending their kids to one school.
The result is that classroom compositions are in flux throughout September and October as principals contact students off the wait list to fill suddenly vacated spots. This is challenging for teachers and ultimately hurts students' education.
District education officials and the State Board of Education can start pushing toward a single lottery right away. An education committee in the Council, as many have suggested, could also help move this forward.

Saturday, September 15, 2012

Auction design within online games

This looks worth following up on...Michael Giberson reports Poor market design causing high prices in Diablo III auction house?

http://arstechnica.com/gaming/2012/06/real-money-auctions-launch-in-diablo-iii/


Why Diablo 3's Real Money Auction House Should Not Be Your Summer Job
"For years, illegal black market sellers have been hawking virtual gear for real life money to great success. There was little point attempting to combat such practices, as little or nothing could be done to stop them, so Blizzard had an alternative idea they wanted to try with Diablo 3. A legal auction house using real world cash, where they got a piece of the action.
"There’s been an in-game gold Auction House in operation since the launch of Diablo 3, but yesterday marked the first time the Real Money Auction House went live, after a few weeks of delays."

Further discussion of markets for organs: some new books

Some books that address markets for organs, including an ethnography of transplants in Israel...

A Market in Organs Pp.128-143
Miran Epstein
[View Abstract] 
This chapter presents a critical overview of the relationship which transplant medicine has had with the market as a source of organs for transplantation. It has three parts. The first two parts discuss the increasing appeal of the market option in practice and theory against the backdrop of the worsening organ crisis and the intensification of pro-transplant interests. The emerging trend suggests that the recent achievements in the struggle against international organ trafficking do not herald the abolition of the organ market but rather presage its reconfiguration in deglobalized, more or less regulated, forms. The third part rephrases the market question. It concludes that the struggle against a market in organs could make sense, let alone stand a chance, only as part of a general struggle against the conditions that have made it so appealing in the first place.

From the eBook Kidney Transplantation: Challenging the Future

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


Matching Organs with Donors: Legality and Kinship in Transplants (Contemporary Ethnography)By Marie-Andree Jacob

The blurb: "While the traffic in human organs stirs outrage and condemnation, donations of such material are perceived as highly ethical. In reality, the line between illicit trafficking and admirable donation is not so sharply drawn. Those entangled in the legal, social, and commercial dimensions of transplanting organs must reconcile motives, bureaucracy, and medical desperation. Matching Organs with Donors: Legality and Kinship in Transplants examines the tensions between law and practice in the world of organ transplants—and the inventive routes patients may take around the law while going through legal processes.

"In this sensitive ethnography, Marie-Andrée Jacob reveals the methods and mindsets of doctors, administrators, gray-sector workers, patients, donors, and sellers in Israel's living kidney transplant bureaus. Matching Organs with Donors describes how suitable matches are identified between donor and recipient using terms borrowed from definitions of kinship. Jacob presents a subtle portrait of the shifting relationships between organ donors/sellers, patients, their brokers, and hospital officials who often accept questionably obtained organs.

"Jacob's incisive look at the cultural landscapes of transplantation in Israel has wider implications. Matching Organs with Donors deepens our understanding of the law and management of informed consent, decision-making among hospital professionals, and the shadowy borders between altruism and commerce.

Friday, September 14, 2012

Paying for human waste tissue used in research: Should it be repugnant? forbidden? allowed? mandated?

Scott Kominers alerts me to the recent debate taking place in the pages of Science.

He writes
"Gary Becker and I have a brief letter in Science supporting the possibility of compensation for donors of waste tissue: http://www.sciencemag.org/content/337/6100/1292.2.full. [Paying for Tissue: Net Benefits}

It is in response to a policy piece by Truog et al. that appeared a few weeks ago, and claims that such compensation is repugnant: http://www.sciencemag.org/content/337/6090/37[Paying Patients for Their Tissue: The Legacy of Henrietta Lacks]

The authors have published a response to our note: http://www.sciencemag.org/content/337/6100/1293.1.full. [Paying for Tissue: Net Benefits—Response]

There is also a letter from Leonard Hayflick which clears up some of the history: http://www.sciencemag.org/content/337/6100/1292.1.short [Paying for Tissue: The Case of WI-38].

Science's editors ... have made all the letters [but not the original Policy Forum article on Research Ethics] open access, and also set up a poll: at http://www.sciencemag.org/content/337/6100/1293.2.full.  (Although the poll topic is not quite parallel to Gary's and my argument – they ask whether researchers should be required to pay patients for waste tissue use, rather than whether they should be allowed to do so.)

Thursday, September 13, 2012

Experimental economics at Stanford SITE, Sept 14 and 15



Stanford Institute for Theoretical Economics
Summer 2012 Workshop

Segment 5: Experimental Economics
September 14 and 15, 2012
Organized by Judd Kesler, The Wharton School, University of Pennsylvania; Lise Vesterlund, University of Pittsburgh; and Muriel Niederle and Charles Sprenger, both Stanford University.

The session will meet in Conference Room A in Landau Economics Building.


Friday, September 14
9.00 - 9.30 Breakfast
9.30 - 9.45 Welcome and Opening Remarks
9.45 - 10.45 Laboratory Results and Field Inference
  • External Validity and Partner Selection Bias presented by Hunt Alcott, New York University and co-authored with Sendhil Mullainathan, Harvard University
  • Free Riding in the Lab and in the Field presented by Florian Englmaier, University of Wurzburg and co-authored with Georg Gebhardt, Ulm University 
10.45 - 11.00 Coffee
11.00 - 12.00 Experimental Methods
  • Incentives in Experiments: A Theoretical Analysis presented by Paul J. Healy, The Ohio State University and co-authored with Yaron Azrieli, The Ohio State University and Christopher P. Chambers, University of California, San Diego
  • Mistakes and Game Form Recognition: Challenges to Theories of Revealed. Preference and Framing presented by Charles R. Plott, California Institute of Technology and co-authored with Timothy N. Cason, Krannert School of Management, Purdue University
12.00 - 1.45 Lunchtime discussion
1.45 - 2.45 Risk and Ambiguity
  • Learning to be Probabilistically Sophisticated presented by Yoram Halevy,University of British Columbia
  • Estimating the Relationship Between Economic Preferences: A Testing Ground for Unifed Theories of Behavior presented by Pietro Ortoleva, California Institute of Technology and co-authored with Mark Dean, Brown University
2.45 - 3.00 Coffee
3.00 - 4.00 Dynamic Inconsistency?
  • Hypobolic Discounting and Willingness to Wait presented by David Eil,  George Mason University
  • Working Over Time: Dynamic Inconsistency in Real Effort Tasks presented by Ned Augenblick, Haas School of Business, University of California, Berkeley and co-authored with Muriel Niederle and Charles Sprenger, both Stanford University
4.00 - 4.15 Coffee
4.15 - 5.15 Shorter Presentation Session
  • Competitive Pricing and the Search for Quality, presented by Daniel Martin, New York University
  • Probability List Elicitation for Lotteries, presented by David Freeman, University of British Columbia
  • Unintended Media E ects in a Conflict Environment: Serbian Radio and Croatian Nationalism, presented by Vera Mironova, Department of Political Science, University of Maryland and co-authored with Stefano DellaVigna, University of California, Berkeley; Ruben Enikolopov and Maria Petrova, New Economic School, Moscow and Ekaterina Zhuravskaya, Paris School of Economics
  • Local Leadership and the Provision of Public Goods: A Field-Lab Experiment in Bolivia, presented by Maria P. Recalde, University of Pittsburgh and co-authored with B. Kelsey Jack, Tufts University
5.30 Continued discussion and dissemination of technical knowledge during dinner

Saturday, September 15
9.00 - 9.30 Breakfast
9.30 - 10.30 Bargaining and Voting Behavior
  • The Effect of Negotiations on Bargaining in Legislatures presented by Marina Agranov, California Institute of Technology and co-authored with Chloe Terigman,University of British Columbia
  • Hypothetical Thinking and Information Extraction: Strategic Voting in the Laboratory presented by Emanuel Vespa, Stern School of Business, New York University and co-authred with Ignacio Esponda, New York University
10.30 - 10.45 Coffee
11.00 - 12.30 Game Theory in the Lab
  • Emergent Star Networks with Ex Ante Homogeneous Agents presented by Daniel Houser, Interdisciplinary Center for Economic Science Contact Information, George Mason University and co-authored with Rong Rong, also Interdisciplinary Center for Economic Science Contact Information, George Mason University
  • Infinitely Repeated Games with Private Monitoring: An Experimental Analysispresented by Guillaume Frechette, New York University and co-authored withMasaki Aoyagi, Osaka University and V. Bhaskar, University College London
  • Finding the Hidden Cost of Control presented by Judd Kessler, The Wharton School, University of Pennsylvania and co-authored with Stephen G Leider,Stephen M. Ross School of Business, University of Michigan, Ann Arbor
12.30 - 2.15 Lunchtime discussion
2.15 - 3.15 Field Experiments
  • Do People Choose What They Think Will Make Them Happiest? Evidence from the National Resident Matching Program. presented by Ori Heffetz, Cornell University and co-authored with Daniel J. Benjamin and Alex Rees-Jones, Cornell University and Miles Kimball, University of Michigan, Ann Arbor
  • Death Probability Shocks and Consumption: Evidence from Denmark presented by Kasper Meisner Nielsen, Hong Kong University of Science and Technology and co-authored with Steffen Andersen, Copenhagen Business School  and Charles Sprenger, Stanford University
3.15 - 3.30 Coffee
3.30 - 4.30 Gender and Economic Behavior
  • Gender, Competition and Career Choices presented by Thomas Buser, University of Amsterdam and co-authored with Muriel Niederle, Stanford University andHessel Oosterbeek also University of Amsterdam
  • Breaking the Glass Ceiling with `No': Gender Differences in Doing Favors. Presented by Lise Vesterlund, University of Pittsburgh and co-authored with Linda Babcock, Amanda Weirup, and Laurie Weingart all Carnegie Mellon University.
4.30 - 4.45 Coffee
4.45 - 5.45 Shorter Presentation Session
  • One in a Million: A Field Experiment on Belief Formation and Pivotal Votingpresented by Mitchell Hoffman, University of California, Berkeley and co-authored with John Morgan, University of California, Berkeley
  • An Experiment on Reference Points and Expectations presented by Changcheng Song, University of California, Berkeley
  • Wallflowers Doing Good: Field and Lab Evidence of Heterogeneity in Reputation Concerns presented by Daniel Jones, University of Pittsburgh and co-authored with Sera Linardi, also University of Pittsburgh
  • Costly Information Acquisition in a Speculative Attack: Theory and Experimentspresented by Isabel Trevino, New York University and co-authored with Michal Szkup, also New York University