Three Essays on Health and Environmental Economics: Applications of Spatial Econometrics and Spatial Analysis
Year: 2023
Type: dissertation
Abstract: Spatial interaction and the locational structure between observations play key roles in the field of econometrics for both cross-sectional and panel data analysis. Compared to a non-spatial econometric model, a spatial model relaxes the assumption of independency in observations. This research applies spatial and non-spatial econometrics in three different fields of applied economics: (1) drinking water and air quality violations impacts on lung and bronchus cancer incidence in the contiguous United States (U.S.); (2) spillover effects of non-pharmaceutical interventions (NPIs) on COVID-19 cases across the contiguous U.S. counties; and (3) urbanization impacts on carbon dioxide (CO2) emissions in selected 119 countries. In Chapter 2, ordinary least squares (OLS) and Spatial Durbin Model (SDM) are applied to data from 48 states plus Washington D.C. in the contiguous U.S. for the period of 2006-2016 to examine the impacts of population-level exposures to environmental quality standards non-compliance on the lung and bronchus cancer incidence. The SDM reveals statistically significant impacts of population-level exposures to violations of environmental pollution standards on lung and bronchus cancer incidence. While impacts are statistically significant (direct effect for water and total effect for air), they are small relative to smoking behavior. Example calculations show that a 10% reduction in population exposure rate to drinking water quality violations across the state of Oklahoma results in a decrease of five cancer cases annually with an estimated annual monetary benefit of $20.6 million. A 10% reduction in population exposure to air quality violations in the state of Utah results in six fewer cancer cases annually with a $24.7 million annual monetary benefit in Utah and three-neighboring states. Chapter 3 examines the spatial spillover effects of NPIs policies on reductions of COVID-19 cases in the contiguous U.S. Using annual cross-sectional data for the year 2020, I apply a spatial Durbin model (SDM) to find statistically significant spillover effects from stay-at-home mandatory orders and mask mandates on COVID-19 cases per 100,000 people. Nationally, on average, 4 cases per 100,000 people can be reduced within the mandate county while an additional 9 cases per 100,000 people can be reduced in 6-nearest neighboring counties as the spillover effect from implementing a mask mandate policy for a month. On the other hand, the direct effect of mandatory stay-at-home orders is 15 cases per 100,000 people while the spillover (indirect) effect is 38 cases per 100,000 people when this NPI is implemented for one month. Thus, mandatory
DOI: 10.33915/etd.12258
Author: Mohammed Syedul Islam
Institution: West Virginia University
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4389934402