Synthetic Control Method and LASSO under Staggered Adoption with Dynamic Treatment Effect. Babii, P. A. & Niu, Y.
abstract   bibtex   
It is often hard to analyze the effect of a policy when the policy is staggered-adopted by states, or the treatment effect is dynamic. Researchers usually make strong assumptions on the staggered adoption and the dynamic treatment effect to make evaluation possible. One popular estimation method that is robust to dynamic treatment effect is Callaway and Sant’Anna’s Difference-in-Difference (Callaway and Sant’Anna (2021)). This paper identifies the limitations of CS-DID and proposes two new methods/algorithms based on SCM, LASSO, and CS-DID (namely: SCM-CS-DID and LASCMCS-DID), that are robust to abnormal units in the treatment group and relax the assumptions from CS-DID.
@article{babii_synthetic_nodate,
	title = {Synthetic {Control} {Method} and {LASSO} under {Staggered} {Adoption} with {Dynamic} {Treatment} {Effect}},
	abstract = {It is often hard to analyze the effect of a policy when the policy is staggered-adopted by states, or the treatment effect is dynamic. Researchers usually make strong assumptions on the staggered adoption and the dynamic treatment effect to make evaluation possible. One popular estimation method that is robust to dynamic treatment effect is Callaway and Sant’Anna’s Difference-in-Difference (Callaway and Sant’Anna (2021)). This paper identifies the limitations of CS-DID and proposes two new methods/algorithms based on SCM, LASSO, and CS-DID (namely: SCM-CS-DID and LASCMCS-DID), that are robust to abnormal units in the treatment group and relax the assumptions from CS-DID.},
	language = {en},
	author = {Babii, Professor Andrii and Niu, Yi},
}

Downloads: 0