The NYT covers the Nobel for Economics, pretty much the only one I really care about and respect. I like the abstracts about mechanism design:
Mechanism design, Professor Maskin explained, can be thought of as the “reverse engineering part of economics.” The starting point, he said, is an outcome that is being sought, like a cleaner environment, a more equitable distribution of income or more technical innovation. Then, he added, one works to design a system that aligns private incentives with public goals.
One recent subject of Professor Maskin’s wide-ranging research has been on the value of software patents. He determined that software was a market where innovations tended to be sequential, in that they were built closely on the work of predecessors, and innovators could take many different paths to the same goal. In such markets, he said, patents might serve as a wall that inhibited innovation rather than stimulating progress.
In Chicago, Professor Myerson said he was drawn to the field of mechanism design because it promised a deeper “understanding of the logic of conflict and cooperation at the most fundamental level” that could possibly help “build better social structures.”
These are essentially very cool things to me. When I get bogged down in the mechanics of BI, I often daydream of the opportunities I might have someday to work with a real economist and build some KPIs from scratch. I've only had the opportunity to do that once in my career and the software was just barely able to manage that kind of complexity. Now that the software has matured, I never see customers who are particularly interested. One of these days when I'm not so hungry to make it big, I'll take the time to find such folks.
The example I heard tell of with regard to mechanism design on the radio that also piqued my interest was the question of whether establishing a market makes sense. If you had 7 empty seats on the city council and 15 applicants, what is the best voting scheme to apply? I worked a similar problem out by using my gut when I put together my top 100 funk songs list.
In the meantime, it's very interesting to hear about the sorts of problems they address, and it fits with my theories that most folks are not really working within the contexts of free markets and more stochastic environments with regard to their application of BI and other software. Quite frankly, only WalMart has introduced an element of near-real time supply and demand pricing in their retail business and it has revolutionized that business. It has been many years since AS/400 systems dominated the mechanics of retail pricing, and yet I wonder how far store managers or buyers have actually crunched numbers to reprice moving inventory as part of their normal business planning.
I'll be looking, in my spare time, for some of the modeling tools of mechanism design to see if they might apply to the technology I'm peddling these days. Could be nice.
Recent Comments