Table 1.

Summary of model efficiency and results by general and specific policies.

Policy intervention	Treatment	Matched	Refinement	Pre-refine	Post-refine	${\text{T}}_{+2}$	${\text{T}}_{+2}$
	Units	Sets	Method	Imbalance	Imbalance	Estimate	(95% CIs)
OxGRT domestic lockdowns	253	191	CBPS	0.088	0.050	$-$ 0.667	$-$ 1.106:$-$ 0.216
COBAP complete closures	215	204	PS	0.128	0.034	0.210	$-$ 0.248:0.698
Specific country	20	18	PS	0.556	0.039	0.922	$-$ 1.397:3.235
Work exception	69	59	CBPS	0.294	0.010	$-$ 0.095	$-$ 0.889:0.614
Citizen exception	106	103	PS	0.059	0.029	0.336	$-$ 0.100:0.758
Essentials only	24	23	CBPS	0.109	0.028	$-$ 0.453	$-$ 1.129:0.209
Islands (subset)	47	47	Mahal.	0.128	0.107	0.727	$-$ 0.245:1.884
COBAP partial closures	759	535	CBPS	0.246	0.009	$-$ 0.044	$-$ 0.341:0.232
Visa ban	55	48	CBPS	0.072	0.001	0.288	$-$ 0.280:0.877
Citizenship ban	106	78	PS	0.368	0.029	$-$ 0.458	$-$ 1.273:0.271
Travel history ban	131	84	PS	0.244	0.013	$-$ 0.445	$-$ 1.151:0.201
Border closures	518	381	PS	0.127	0.002	0.026	$-$ 0.304:0.372
Islands (subset)	143	87	PS	0.364	0.289	0.029	$-$ 0.634:0.558

The first two columns show the total number of treated units available for analysis and then the number remaining after matching. Columns three through five report the refinement method used for each model as well as the reduction of imbalance between a model without refinement and after refinement, with 0 representing optimal balance across all variables. Finally, column six reports the ${\text{T}}_{+2}$ estimate, and column seven provides the 95% confidence intervals.