tag:blogger.com,1999:blog-6837159629100463303.post687858114229798888..comments2023-06-18T01:25:08.748-07:00Comments on Information Transfer Economics: Should I use math or just assume convergence of a vague algorithm?Jason Smithhttp://www.blogger.com/profile/12680061127040420047noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-6837159629100463303.post-72668660376304210202016-10-14T15:06:23.193-07:002016-10-14T15:06:23.193-07:00I re-read Kling's piece and it seems to be abo...I re-read Kling's piece and it seems to be about the use of math. Mentioned in every single paragraph.<br /><br />He also isn't doubting the conclusions of the Nobelists' work -- his primary point is that he thinks he could arrive at them without using math. <br /><br /><i>A far more interesting question is whether the state spaces in question are reasonably optimized by a sort of 'managerial gradient decent'. If so, then the choice of (complex) math or heuristic is irrelevant to the final state.</i><br /><br />The first part is exactly my point (Kling's intuition can be expressed mathematically), however I disagree with the second sentence. Different algorithms will have different results. Some equilibria are "<a href="http://worthwhile.typepad.com/worthwhile_canadian_initi/2016/03/stag-hunt-and-the-money-problem.html" rel="nofollow">learnable</a>" so how you achieve them is directly linked to the final state.Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-82348017173713327472016-10-11T15:34:55.409-07:002016-10-11T15:34:55.409-07:00At least as far as headlines go, this seems a some...At least as far as headlines go, this seems a somewhat unfair read on Kling; his point is that most economic agents aren't rational omniscient optimizers, and making that the focus of economic research is perhaps unwise. Better to assume they are local optimizers, at best.<br /><br />This seems quite reasonable as a point, and in have no idea if it is fair to the Nobelists being discussed.<br /><br />A far more interesting question is whether the state spaces in question are reasonably optimized by a sort of 'managerial gradient decent'. If so, then the choice of (complex) math or heuristic is irrelevant to the final state.<br /><br />BpUnknownhttps://www.blogger.com/profile/04773368635111455556noreply@blogger.com