A pitfall of expectational stability (E-stability) analysis can arise in models with multiperiod expectations: if an auxiliary variable is introduced as substitute for an expectational endogenous variable in such a model, this shrinks the region of the model parameters that guarantee E-stability of a fundamental rational expectations equilibrium. Moreover, in the model representation with no auxiliary variable, the same E-stability region as in that with the auxiliary variable is obtained if economic agents are assumed to make multiple forecasts in an inconsistent manner. Therefore, we argue that the introduction of an auxiliary variable as substitute for an expectational endogenous variable in models with multi-period expectations can induce misleading implications that are biased toward E-instability.
