Research
Publications
Truncated History Framework for Synthetic Control Approaches
Economics Letters
Abstract:
Show
We introduce the Truncated History (TH) framework for synthetic control approaches, which excludes pretreatment periods. The in-time placebo test fits within TH as right-truncation, removing final pretreatment periods. While useful, it may over-rely on early pretreatment data, potentially undermining robustness. To address this, we propose a left-truncation strategy, the left-TH robustness check, which iteratively excludes the earliest pretreatment periods and re-estimates effects. Stable results support credibility, while instability suggests that interval estimates may be more appropriate. TH provides robustness for pretreatment horizons and helps identify stable models, which we demonstrate on the California tobacco program of Abadie et al. (2010).Co-authors:
Working Papers:
Does a reduction in teacher education tuition fees reduce teacher shortages?
Co-authors:
Student Effort and Multi-Tasking under Graduation Constraints
Single authored