Semiparametric Triple Difference Estimators
Sina Akbari, Negar Kiyavash, AmirEmad Ghassami · 2025
Read Semiparametric Triple Difference Estimators by Sina Akbari, Negar Kiyavash, AmirEmad Ghassami (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-19788.txt.
Description
Sina Akbari, Negar Kiyavash, and AmirEmad Ghassami build better estimators for a causal method called triple difference in this 2025 paper. The method extends the difference-in-differences design by borrowing data from an extra group, which loosens the strict trend assumption that design needs. Despite wide use, its statistical foundations had been little studied.
The authors give regression and weighting routes to identify the treatment effect on the treated, for both panel data and repeated cross-sections. For repeated cross-sections they drop the usual assumption that the group makeup stays fixed over time. Their estimators use cross-fitting, so flexible machine learning tools can handle background parts. They state conditions under which the estimators are efficient, doubly robust, and well behaved in large samples. As a test, they find mandated maternity benefits lowered the hourly wages of women of childbearing age by two point six percent.
Catalog record
| CREATOR | Sina Akbari, Negar Kiyavash, AmirEmad Ghassami |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 156 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-19788