Premium
Self‐fulfilling regression and statistical discrimination
Author(s) -
Hayashi Takashi
Publication year - 2019
Publication title -
international journal of economic theory
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.351
H-Index - 11
eISSN - 1742-7363
pISSN - 1742-7355
DOI - 10.1111/ijet.12190
Subject(s) - productivity , statistical discrimination , economics , simple (philosophy) , competition (biology) , econometrics , selection (genetic algorithm) , point (geometry) , regression , nothing , joint probability distribution , distribution (mathematics) , mathematical economics , microeconomics , mathematics , statistics , computer science , labour economics , artificial intelligence , ecology , mathematical analysis , philosophy , geometry , epistemology , biology , macroeconomics
This paper provides a simple model which explains that statistical discrimination can arise in a purely self‐fulfilling manner. The story is as follows. (i) At the point of hiring, employers cannot observe workers’ productivities but can observe only their signals, such as test scores, and under perfect competition they pay for expected labor productivity conditional on signal observation, based on their belief about return to signal. (ii) Given the employers’ belief, workers choose an effort level, which affects the joint probability distribution over productivity–signal pairs. (iii) In equilibrium, the employers’ belief proves to be statistically consistent. We show that there may be multiple equilibria, and that equilibrium selection has nothing to do with economic fundamentals.