The Relationship Between Capital Account Openness and Environmental Degradation
Introduction
This report explores the relationship between capital account openness and environmental degradation across different countries. Specifically, we test whether capital account openness, as measured by the Chinn-Ito Index, is associated with environmental degradation, with a particular focus on whether the relationship varies by level of economic development (measured by GDP per capita).
Hypotheses
- Hypothesis 1: There is a negative relationship between capital account openness and environmental degradation, meaning that greater openness to capital inflows leads to a reduction in pollution due to higher efficiency in production.
- Hypothesis 2: The relationship between capital account openness and environmental degradation depends on the country’s level of economic development, with wealthier countries showing a weaker association due to better technology and environmental regulations.
Data Collection
The data used in this analysis includes the following variables:
- Capital Account Openness (Chinn-Ito Index): This measure indicates how easy it is for capital to flow into a country. It is used as the main explanatory variable.
- Environmental Degradation (CO2 Emissions per capita): This measure serves as the dependent variable representing pollution levels.
- GDP per capita: A measure of economic development, included to test for interaction effects.
- Energy Consumption per capita: A control variable representing the amount of energy consumed.
- Urbanization: A control variable representing the percentage of the population living in urban areas.
The data is sourced from various publicly available sources such as the World Bank, Our World in Data, and other reputable international organizations. The dataset takes the form of panel data across 100 countries, with observations for several years.
Data Preprocessing
# Install required libraries
install.packages("tidyverse") # for data manipulation and visualization## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.4'
## (as 'lib' is unspecified)install.packages("lmtest") # for hypothesis testing in regression models## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.4'
## (as 'lib' is unspecified)install.packages("sandwich") # for robust standard errors## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.4'
## (as 'lib' is unspecified)install.packages("readxl") # for reading Excel files## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.4'
## (as 'lib' is unspecified)# Load required libraries
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(lmtest)## Loading required package: zoo
##
## Attaching package: 'zoo'
##
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numericlibrary(sandwich)
library(readxl)
# Load data
data <- read_excel("Capital_Openness_Environmental_Impact.xlsx")
# View the first few rows of the data
head(data)## # A tibble: 6 × 6
## Country GDP_per_capita Capital_Openness CO2_Emissions Energy_Consumption
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Portugal 19034. 1.74 36.4 1.29
## 2 Israel 43306. 0.685 28.2 0.499
## 3 Russia 42001. -1.75 38.2 0.204
## 4 Turkey 49087. 1.57 43.6 1.35
## 5 Czech Republ… 32813. 0.00379 48.8 0.648
## 6 Egypt 3834. 1.15 45.4 1.43
## # ℹ 1 more variable: Urbanization <dbl># Summary statistics of the data
summary(data)## Country GDP_per_capita Capital_Openness CO2_Emissions
## Length:100 Min. : 1162 Min. :-1.991909 Min. : 0.1856
## Class :character 1st Qu.:14331 1st Qu.:-0.934353 1st Qu.:15.3934
## Mode :character Median :26838 Median : 0.027447 Median :28.5655
## Mean :25568 Mean :-0.002596 Mean :27.3791
## 3rd Qu.:37090 3rd Qu.: 1.007853 3rd Qu.:41.3983
## Max. :49766 Max. : 1.980530 Max. :49.9180
## Energy_Consumption Urbanization
## Min. :0.1105 Min. :30.10
## 1st Qu.:0.5116 1st Qu.:45.11
## Median :0.8894 Median :62.14
## Mean :0.8526 Mean :60.78
## 3rd Qu.:1.2023 3rd Qu.:76.59
## Max. :1.4955 Max. :89.58# Scatter plot with a linear regression line
ggplot(data, aes(x = Capital_Openness, y = CO2_Emissions)) +
geom_point(color = "blue") +
geom_smooth(method = "lm", se = FALSE, color = "red") +
labs(title = "Capital Openness vs CO2 Emissions", x = "Capital Openness", y = "CO2 Emissions (per capita)") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'
# Scatter plot with color representing GDP per capita
ggplot(data, aes(x = Capital_Openness, y = CO2_Emissions, color = GDP_per_capita)) +
geom_point() +
scale_color_viridis_c() +
labs(title = "Capital Openness vs CO2 Emissions by GDP per Capita", x = "Capital Openness", y = "CO2 Emissions (per capita)", color = "GDP per Capita") +
theme_minimal()
# Fit a base linear regression model
base_model <- lm(CO2_Emissions ~ Capital_Openness + GDP_per_capita + Energy_Consumption + Urbanization, data = data)
summary(base_model)##
## Call:
## lm(formula = CO2_Emissions ~ Capital_Openness + GDP_per_capita +
## Energy_Consumption + Urbanization, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -28.056 -11.953 1.949 13.503 24.064
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.496e+01 6.892e+00 5.073 1.94e-06 ***
## Capital_Openness 1.857e+00 1.337e+00 1.389 0.168
## GDP_per_capita -1.123e-05 1.095e-04 -0.103 0.918
## Energy_Consumption 2.365e-01 3.834e+00 0.062 0.951
## Urbanization -1.233e-01 8.504e-02 -1.450 0.150
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.02 on 95 degrees of freedom
## Multiple R-squared: 0.03983, Adjusted R-squared: -0.000593
## F-statistic: 0.9853 on 4 and 95 DF, p-value: 0.4194# Fit a model with interaction between Capital Openness and GDP per capita
interaction_model <- lm(CO2_Emissions ~ Capital_Openness * GDP_per_capita + Energy_Consumption + Urbanization, data = data)
summary(interaction_model)##
## Call:
## lm(formula = CO2_Emissions ~ Capital_Openness * GDP_per_capita +
## Energy_Consumption + Urbanization, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -27.838 -11.679 1.792 12.350 24.196
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.490e+01 6.918e+00 5.045 2.21e-06 ***
## Capital_Openness 3.483e+00 3.155e+00 1.104 0.272
## GDP_per_capita 1.320e-06 1.121e-04 0.012 0.991
## Energy_Consumption 1.498e-01 3.851e+00 0.039 0.969
## Urbanization -1.285e-01 8.583e-02 -1.497 0.138
## Capital_Openness:GDP_per_capita -5.751e-05 1.010e-04 -0.569 0.570
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.08 on 94 degrees of freedom
## Multiple R-squared: 0.04314, Adjusted R-squared: -0.007761
## F-statistic: 0.8475 on 5 and 94 DF, p-value: 0.5195# Display base model results with robust standard errors
coeftest(base_model, vcov = vcovHC(base_model, type = "HC1"))##
## t test of coefficients:
##
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.4962e+01 6.8017e+00 5.1402 1.466e-06 ***
## Capital_Openness 1.8565e+00 1.3272e+00 1.3988 0.1651
## GDP_per_capita -1.1234e-05 1.1652e-04 -0.0964 0.9234
## Energy_Consumption 2.3648e-01 3.9138e+00 0.0604 0.9519
## Urbanization -1.2327e-01 9.0371e-02 -1.3641 0.1758
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Display interaction model results with robust standard errors
coeftest(interaction_model, vcov = vcovHC(interaction_model, type = "HC1"))##
## t test of coefficients:
##
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.4896e+01 6.9578e+00 5.0153 2.491e-06 ***
## Capital_Openness 3.4826e+00 3.3676e+00 1.0341 0.3037
## GDP_per_capita 1.3202e-06 1.1908e-04 0.0111 0.9912
## Energy_Consumption 1.4976e-01 3.9836e+00 0.0376 0.9701
## Urbanization -1.2847e-01 9.2467e-02 -1.3894 0.1680
## Capital_Openness:GDP_per_capita -5.7513e-05 1.0682e-04 -0.5384 0.5916
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Predict CO2 emissions using the base model
data$predicted_co2_base <- predict(base_model, newdata = data)
# Scatter plot with predicted CO2 emissions
ggplot(data, aes(x = Capital_Openness, y = CO2_Emissions)) +
geom_point(color = "blue") +
geom_line(aes(x = Capital_Openness, y = predicted_co2_base), color = "red", size = 1) +
labs(title = "Predicted CO2 Emissions (Base Model)", x = "Capital Openness", y = "CO2 Emissions (per capita)") +
theme_minimal()## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
# Predict CO2 emissions using the interaction model
data$predicted_co2_interaction <- predict(interaction_model, newdata = data)
# Scatter plot with predicted CO2 emissions from the interaction model
ggplot(data, aes(x = Capital_Openness, y = CO2_Emissions, color = GDP_per_capita)) +
geom_point() +
geom_line(aes(x = Capital_Openness, y = predicted_co2_interaction), color = "red", size = 1) +
scale_color_viridis_c() +
labs(title = "Predicted CO2 Emissions (Interaction Model)", x = "Capital Openness", y = "CO2 Emissions (per capita)", color = "GDP per Capita") +
theme_minimal()