Abstract Rationality and learning
Rationality and Learning in Survey Expectations of Inflation.
The main question of this paper is how private agents form their expectations in different environments. To answer this, we approximate survey expectations of inflation by constant gain learning algorithms.
We show that in stable economies survey respondents exhibit more rationality then adaptive learning: they provide better forecasts then the adaptive learning approximation of the perfect foresight inflation.
We also find that compared to stable economies in more volatile economies private agents pay more attention to recent data when they form their inflationary expectations.
The main question of this paper is how private agents form their expectations in different environments. To answer this, we approximate survey expectations of inflation by constant gain learning algorithms.
We show that in stable economies survey respondents exhibit more rationality then adaptive learning: they provide better forecasts then the adaptive learning approximation of the perfect foresight inflation.
We also find that compared to stable economies in more volatile economies private agents pay more attention to recent data when they form their inflationary expectations.