Market Design and Computer-Assisted Markets: An Economist's Perspective
summary
Shows that running a deferred acceptance algorithm by phone with a fixed 7‑hour selection day creates congestion: offers take five minutes and rejections one minute, simulation gives mean termination ~18 hours and ~900 calls needed so the market stops early and resorts to exploding offers. A computerized centralized clearinghouse removes that bottleneck, makes truthful preference reporting dominant for applicants, vastly reduces blocking pairs, and handles couples (the couples case can be NP‑complete in worst cases though practical algorithms work on large sparse markets).
AI-written summary · about
chapters (27)
- 0:00Introduction
- 0:45ComputerAssisted Markets
- 2:23Matching Markets
- 4:16Algorithms
- 6:18Simple Model
- 7:41Matchings
- 9:27Deferred Acceptance Algorithm
- 12:01Stable Matching
- 14:05Incentives
- 14:50Stability
- 17:32Clinical Psychology
- 19:26Market Rules
- 21:43Deferred Acceptance Algorithms
- 22:28Acceptance Rules
- 26:04Behavioral Observations
- 33:36Conclusions
- 37:27Why didnt this work
- 39:44A Hard Problem
- 40:09A Puzzle
- 42:03A corollary
- 42:19Open problems
- 42:54What do markets do
- 44:53Couples
- 45:42Preferences
- 46:52Why not a deferred acceptance algorithm
- 49:13Natural extensions
- 53:26Threats
related talks
29mClearing Matching Markets Efficiently: Informative Signals and Match Recommendations
44mClearing Matching Markets Efficiently: Informative signals and Match Recommendations
50mDiversity in Matching Markets
44mFacilitating the Search for Partners on Matching Platforms
01h 38mRidesharing
44mThe Design of Financial Exchanges: Some Open Questions at the Intersection of Econ and CS
36mThe Intractability of Hylland-Zeckhauser and its Aftermath
44mOnline Advertisements and Online Algorithms