Page 207 - AJWEP-v22i3

Effects of urbanization on air and health

remains limited compared to studies in China or India. have developed various approaches to address these
Early Vietnamese studies were primarily descriptive, analytical challenges.
documenting pollution levels and basic health Panel data methods have proven valuable for
correlations. controlling time-invariant unobservable factors. Zhou
Phung et al. examined air pollution and hospital et al. employed fixed-effect models with city-level
22
28
admissions in Ho Chi Minh City from 2010 to 2013, panel data in China to estimate urbanization impacts
finding significant correlations between daily PM10 on PM2.5 levels while controlling for economic factors
levels and respiratory hospital admissions. Hien et al. and policy interventions. Difference-in-differences
23
analyzed air quality in Hanoi from 2016 to 2018, approaches have been used to evaluate specific
identifying traffic emissions and coal combustion policy impacts; for example, Greenstone and Hanna
29
as primary sources of particulate pollution. Nhung applied this method to assess air pollution regulation
et al. found that annual PM2.5 concentrations in all effectiveness in India.
24
studied provinces exceeded the limits set by the WHO IV techniques have emerged as particularly
Air Quality Guidelines and the Vietnamese Emission valuable for establishing causality, though finding
Standards (QCVN). These studies confirmed that suitable instruments remains challenging. Appropriate
Vietnam’s urban pollution levels have frequently instruments must influence the endogenous variable
exceeded national standards and generated conspicuous (e.g., air pollution) while affecting the outcome
short-term health impacts. (e.g., health) only through the endogenous variable.
Several studies have linked Vietnam’s rapid Deryugina et al. used wind direction as an instrument
30
development to environmental degradation. By using for air pollution to estimate causal effects on mortality in
satellite imagery, Nguyen et al. documented the rapid the United States, while Liu et al. employed rainfall as
31
25
expansion of built-up urban areas in major Vietnamese an instrument for air pollution in studying recreational
cities, which were accompanied by rising air pollutant behavior impacts in Paris.
concentrations. Tran and Nguyen found that economic Our study addresses the methodological gap in
26
growth and industrialization were driving both Vietnam-specific research by applying panel data
urbanization and environmental decline, including econometric techniques with city fixed effects and IVs to
higher emissions and water pollution, across Vietnam’s establish more robust causal links between urbanization,
cities. pollution, and health outcomes. By analyzing multiple
More recent research has begun employing more cities over a decade, we can control for unobservable
sophisticated analytical methods. Vu et al. investigated city-specific factors and common temporal shocks while
27
ambient air pollution impacts on hospital admissions for leveraging industrial composition as an instrument for
respiratory diseases among children in Hanoi, finding pollution levels.
that days with higher PM2.5 and NO2 levels corresponded
to increased pediatric hospitalizations for respiratory 3. Data and methodology
issues, even after controlling for weather and seasonal
factors. While informative, such time-series analyses 3.1. Data sources and description
still face challenges in establishing causality due to In this study, we constructed a comprehensive panel
potential confounding factors. dataset encompassing ten major Vietnamese cities
However, most Vietnam-specific studies to date observed annually from 2013 to 2022. The selected
have been correlational or descriptive, with few cities – Hanoi, Ho Chi Minh City, Haiphong, Da Nang,
employing advanced causal inference methods. Can Tho, Bien Hoa, Hue, Nha Trang, Buon Ma Thuot,
This represents a significant gap given the policy and Vinh – represent Vietnam’s key urban centers across
importance of understanding causal relationships northern, central, and southern regions. This selection
between urbanization, pollution, and health outcomes provides geographic diversity despite constraints in
in Vietnam’s rapidly changing urban landscape. obtaining selected data, particularly for consistent air
quality monitoring and health outcome reporting.
2.4. Methodological approaches These cities collectively house over 30% of Vietnam’s
Identifying causal relationships between urbanization, urban population and represent diverse economic
air pollution, and health outcomes presents substantial profiles, from major metropolitan centers to secondary
methodological challenges due to potential endogeneity, industrial cities. This diversity proves crucial for
reverse causality, and omitted variable bias. Researchers identifying causal relationships, as it provides variation

Volume 22 Issue 3 (2025) 201 doi: 10.36922/AJWEP025130088