We extend the ongoing literature on regime change.
We allow time variation in disturbance variances of regime-switching policy rules.
We achieve superior modeling with quarterly US data.
We find that passive monetary has stronger ARCH effects than active monetary policy.