Convergence in Monetary Inflation Models with Heterogeneous Learning Rules

Inflation and financing of public expenditure by are analysed in an OLG model where the deficit is constrained to be less than a given fraction of intergenerational savings. Even if there may be multiplicity of steady-state equilibria, we show that, with such a constraint, the dynamics with adaptive learning are globally convergent to a set of equilibria satisfying a local stability condition. We allow for heterogeneity of agents' learning rules and look at the role of some basic behavioural assumptions, such as a certain degree of random e-precautionary savings and inertia on agents' updating of beliefs. We also provide experimental evidence on the effect of public expenditure constraints on the stability of equilibria.