Table 2.

Effects of COVID on infant health outcomes.

	(1)	(2)
	birth weight	$\mathbb{1}[\text{Apgar}<8]$
Panel A: extensive margin
COVID	−1.046 (19.556)	−0.023 (0.014)
Indigenous	−59.908∗∗∗ (10.075)	−0.003 (0.009)
Indigenous ​× ​COVID	−10.552 (20.329)	0.013 (0.015)
N	55,731	50,513
R-squared	0.105	0.061
Panel B: intensive margin
COVID	0.225 (0.439)	−0.000 (0.000)
Indigenous	−60.833∗∗∗ (9.809)	−0.002 (0.009)
Indigenous ​× ​COVID	−0.090 (0.294)	0.000 (0.000)
N	55,731	50,513
R-squared	0.105	0.061

Notes: The table shows coefficients from four different regressions. The dependent variable is birth weigh in the two models displayed in column (1), and $\mathbb{1}[\text{Apgar}<8]$ in the two models of column (2). Panel A: COVID variable captures the extensive margin of potential COVID-19 exposure. Panel B: COVID variable captures the intensive margin of potential COVID-19 exposure. All regressions include mother and child covariates in addition to pre-pandemic municipality characteristics interacted with a linear trend. All regressions control for municipality and month-by-year fixed effects. Standard errors clustered at the mother's residence municipality are in parentheses. Sample includes all births in the state of Amazonas from Jan 1, 2019 to July 31, 2020 (excluding births in the capital city Manaus). ∗p ​< ​0.1, ∗∗p ​< ​0.05, ∗∗∗p ​< ​0.01.