open access

https://hdl.handle.net/2115/44985

Reconstruction of Philips curves in `macro'economics from the `micro'scopic description of arti cial labor markets

日本語

2011-03-24

2, 4, 39 p.

We attempt to reconstruct macroscopic properties of labor markets from the view point of microscopic descriptions of multi-agent systems. Especially we derive the so-called Philips curve which generally shows universal negative-correlations between the unemployment rate U and the inflation rate . We first briefly review the macroscopic study by Neugart (2004) who derived deterministic non-linear maps with respect to the U and . He concluded that the strange attractor of the chaotic maps should be identical to the Philips curve. Whereas, in our microscopic modeling, we should set-up the behaviour of both labor and company agents to construct the macroscopic quantities such as U and . For instance, some amount of labors can access the information about the successful persons in the previous business year on the number of their applications. On the other hand, each artificial company in our modeling produces a single homogenous commodity and we shall assume that the total amount of the products in a company depends on the total amount of the labor’s abilities in the company. By utilizing computer simulations, we find that our probabilistic model is able to reproduce the Philips curve efficiently. We also confirm that the best possible number of applications for each agent is ‘three’ in the sense that the employment rate is maximized if and only if all agent post their applications (what we call ‘entry seat’) just only three times to the companies. We also discuss the possibility to reproduce the Philips curve by making use of what we call ‘mesoscale model’, which is a hybrid modeling by the mixture of our microscopic model and the macroscopic Neugart model. To check the validity of the modeling, we compare our result with the empirical evidence of Philips curves in recent Japanese labor markets. We shall report the detail in this talk and thesis.

修士(情報科学)

修士論文

Version of Record