February 2021

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
Blog powered by Typepad

« Claudia Goldin's Presidential Address to AEA January 2014 | Main | On the Reading and Teaching of The Grapes of Wrath »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I've always seen market design as "improving rules to cultivate better economic performance."

Clearly, today's market design economics is a descendent of the older mechanism design theory (i.e. Hurwicz etc.) that was in some sense intended to "solve" Mises/Hayek calculation problems. I'm not sure I know just what Roth would say, but I see his ambitions as more applied/micro/practical in scope. So, at least as I see it, market design today is oriented at rules that help markets work better given that relevant knowledge is distributed, often tacit, and often only revealed or created in the context of its use.

In the presentation Roth said he liked the conceptualization of "marketplace design," and I think the metaphor is instructive. Think of how a retailer arranges his goods, cash registers, entry/exit and traffic patterns, and then rules on returns, or use or credit cards, and such things. In this case there is a designer and a unified goal to maximize profits for the business.

A stock exchange faces similar design issues even though the market is virtual/distributed and the buyers and sellers, exchange managers, and exchange board members have a wide variety of goals. The stock exchange market designer isn't trying to design allocations, but rather to design a more effective allocator.

1. The distinction between engineer versus gardener is not helpful once a Monsanto is created.

2. Roth is mostly designing matching markets. Exchanges between 2 or more parties which depend on more than price signals.

The comments to this entry are closed.

Our Books