Solving Problems in the Panel Regression Model for Truncated
Dependent Variables: A Constrained Optimization Method
5.3 Methodological Suggestions
I do not argue that constrained optimization has to be imposed to panel regression, but this article provides three different resolutions when empirical or theoretical boundary violation occurs. Given the same model specification, my analysis suggests that the current panel regression (Model II) and the maximum likelihood with correction of demeaning bias (Model V) can serve as the optimistic and conservative results in hypothesis testing, respectively. If a significance finding appears in both models, then the robustness of this finding is corroborated. However, if the results in both models do not agree, we need to be careful in interpreting them, since different results are associated with different methodological assumptions. Nevertheless, I do recommend adopting constrained optimization whenever panel regression suffers from empirical boundary violations, because their occurrence is illogical.
Welcome all math lovers!
