Decentralized Neuro Phillips-type Policy Models for Fiscal and Monetary Decision-makers

M. Hosoi and Y. Ito (Japan)


Decentralized, neural networks, Phillips-type stabilization policy, fiscal and monetary decision-makers, econometric model , Japan.


The purpose of this paper is to consider applicability and theoretical framework of the Neuro Phillips-type stabilization policy for decentralized fiscal and monetary decision makers with the upper level policy-coordinator. Two decision-makers have the mutual interaction controlled by this coordinator. These three decision-makers have hierarchical policy structure aiming their desired policy targets such as constant growth rate of GDP, the rate of price change and unemployment rate. This policy is formulated by the Phillips-type stabilization policy problem consisting of feedback rule policies with varying the feedback gain parameters leading to the desired targets subject to the econometric models characterized by the respective fiscal and monetary decision-makers. In order to compute the feedback gain coefficients, we use neural networks algorithm. This gaming policy solutions are considered by simulating a small econometric model of Japan including simplified fiscal and monetary sectors.

Important Links:

Go Back