Public debt shocks and the interest-growth differential: a two-step CCE procedure with Monte Carlo evidence

This study proposes a two-step heterogeneous CCE procedure to estimate non-systematic variation in public debt and its effects on the interest-growth differential. We consider two dynamic DGPs that satisfy the rank condition and assume that the set of unobserved common factors is characterized by mixed orders of integration. The empirical results indicate that unanticipated variations in public debt exert positive average effects on the interest-growth differential. However, the distribution of the unit-specific estimated coefficients suggests that the effects are markedly heterogeneous across countries. The Monte Carlo experiments suggest that the CCE estimators employed in this paper exhibit good finite-sample properties in the proposed dynamic heterogeneous framework, even under unusual and complex time-series properties of the latent stochastic factors.