Estimation of pre- and posttreatment average treatment effects with binary time-varying treatment using Stata

Abstract

In this article, we describe tvdiff, a community-contributed command that implements a generalization of the difference-in-differences estimator to the case of binary time-varying treatment with pre- and postintervention periods. tvdiff is flexible and can accommodate many actual situations, enabling the user to specify the number of pre- and postintervention periods and a graphical representation of the estimated coefficients. In addition, tvdiff provides two distinct tests for the necessary condition of the identification of causal effects, namely, two tests for the so-called parallel-trend assumption. tvdiff is intended to simplify applied works on program evaluation and causal inference when longitudinal data are available.

Adolescent Suicide, Household Firearm Ownership, and the Effects of Child Access Prevention Laws

“…The parallel trends assumption necessary for the identification of causal effects was evaluated and verified using time-trend significance tests using the tvdiff package for time-varying treatment across many treated units. 24 Benjamini and Hochberg's procedure was used to account for multiple testing to constrain the false discovery rate to 10% to balance power with type I error control in these exploratory analyses. 25 Uncorrected p values are reported and evaluated relative to BenjaminiÀHochberg critical values.…”

How Do Banking Groups React to Macroprudential Policies? Cross-Border Spillover Effects of Higher Capital Buffers on Lending, Risk-Taking and Internal Markets

“…The start of the treatment corresponds to the date when the national authorities notified their decision to the European Central Bank (ECB). 29 After the notification is issued (i.e. for t τ ), the treatment status I i,t changes, where banks with a score above a predetermined country-specific threshold are qualified as OSII and may be charged with an additional capital requirement.…”

Do company-owned academic patents influence firm performance? Evidence from the Italian industry

“…We first set out our empirical analysis by showing to what extent patenting firms obtain a stronger position on the market with respect to firms that do not patent. To this aim, we employ a dynamic difference-in-differences model as proposed by Cerulli and Ventura (2019). Subsequently, we focus on testing both H1 and H2-i.e.…”

“…Following Autor (2003), Cerulli and Ventura (2019) show that the parameters in Eq. (5)-that extended to more than one lag and more than one lead-have a causal interpretation as average treatment effects over time in Fig.…”