IMMIGRANT FLOWS AND SARS-COV-2 INFECTIONS IN THE UNITED STATES

ABSTRACT

BACKGROUND: The relationship between immigrant inflow and SARSCOV- 2 spread in the United States has driven much political discussion and policy, including the implementation of Title 42 by the Centers for Disease Control, effectively banning asylum claims at the US Southwestern Border. The implemenation of Title 42 has led to significant outcry from public health researchers and policymakers alike. To examine the relationship between SARS- COV-2 spread and immigrant inflow, we compared 2020-2021 immigrant flows (from US Border Patrol estimates of Southwest US border crossings and a US seasonal guest worker program) to county-level SARSCOV- 2 rates. METHODS: We drew immigraiton data from US Border Patrol (USBP) 2020-2021 estimates of successful Southwest US Border crossings and from Department of Labor data on H2-A visa allotments for agricultural guest workers. The primary outcome is the one-month lagged natural logarithm of new cases per 100,000 persons at the county level. After arithmetic manipulation, the primary outcome can be interpreted as the percentage change in one-month lagged county- level SARS-COV-2 cases per 100,000 persons as a function of each additional immigrant entrant in the same county (i.e., semielasticity). We controlled for the vaccination rate within a county as well as the interaction between the presence of state-mandated facemask mandates, restaurant closures and gym closures. We analyzed the primary outcome as a linear regression with county- and month-level fixed effects. Multiple robustness checks tested various assumptions undertaken in this analysis. RESULTS:We drew data from 3,144 counties and geographic units, 2,077 of which received immigrants over the course of the study. Among immigrant receiving counties, themedian immigrant inflow was 41 (IQR 13, 139) persons over the period studied. The RioGrandeValley border counties of southeastern Texas received the most immigrants during this period, with an average monthly immigrant inflow of 19,526, peaking in 2021. Under our primary model there was a statistically insignificant 0.0000189% decrease in SARSCOV- 2 cases per 100,000 persons for each additional immigrant (p=0.231, 95% CI -0.0000499%, 0.0000120%). When restricted to H-2A guest workers only, there is a statistically insignificant 0.000197% decrease in SARS-COV-2 cases per 100,000 persons for each additional immigrant (p=0.09, 95% CI -0.00043%, 0.0000298). When restricted to Southwest border crossings alone, there was a statistically significant 0.0000254% increase in SARS-COV-2 cases per 100,000 persons for each additional immigrant (p=0.005, 95% CI 0.0000107%, 0.0000401%). However, the result became insignificant after excluding 2021 Rio Grande Valley sector crossings (0.0000368, p=0.116, 95 CI% -0.0000116, 0.0000852). CONCLUSIONS: The contriibution of immigant inflows to the SARS-COV- 2 pandemic in the United States is small, largely nonstatistically significant, and when statistically significant, is not robust to alternative specifications.