Facilitating Heterogeneous Effect Estimation via Statistically Efficient Categorical Modifiers | II-Commons