Scientific Realism-Based Probability for Probabilistic Economic Analysis

概要

論文の詳細を見る
In realist economics formed by Cairnes, Robbins and Keynes, statistical economic analysis is regarded as useless, although statistics looks intimate with realism. What made economics conflict with statistics? This paper focuses on this problem. We start from presenting traditional realist economists' views on uselessness of statistical economic analysis in the section 1. The section 2 is focused both on lack of homogeneousness condition of data and on changing from economic data into random variable, in order to investigate the validity of application of standard sampling statistics to the economic phenomenon composed of data having specific character different from data available in natural sciences. The latter half (i.e., the section 3, 4) of this paper discuss a new interpretation of probability based on scientific realism and the adaptation of the interpretation to the problem of confirming economic hypothesis regarded as a central economic analysis, from the point of view of realist economics, not of modern econometrics. This paper insists the methodological substitutability of probabilistic reasoning based on scientific realism for statistical inference in economics.
2005-10-15