“…Second, to fully capture the dynamics of the reform process and credit rating and whether the impacts of pension reforms on credit ratings accelerated, stabilized, or reverted, we applied empirical strategies similar to those used by Autor (2003) and Stevenson and Wolfers (2006) to estimate the main model augmented with leads and lags of the reform adoption, where the sums allow for the m lags ( δ − 1 , 0.25em δ − 2 , … , 0.25em δ − m false) , 0.25em 0.25emor post-treatment effects and the q leads ( δ + 1 , 0.25em δ + 2 , … , 0.25em δ + q false) , 0.25em 0.25emor anticipatory effects as follows: Third, the dichotomous reform specification is based on the DID assumption of a homogeneous intervention of the event. For this study, this assumption entails that having one pension reform has the same effect on credit rating as having multiple pension reforms.…”