IMPACT OF

INCOME INEQUALITY
ON
ENVIRONMENTAL
DEGRADATION

Radhika Gupta (22BA074)

Section A
1. Introduction
The selection of the topic, "Impact of Income Inequality on Environmental Degradation,"
stems from the recognition of the enduring debate surrounding environmental conservation
and economic development. Within this discourse, environmental conservation advocates
emphasize the critical need to prioritize the protection of natural resources and ecosystems
for sustained global well-being. Conversely, proponents of economic development assert its
indispensable role in fostering growth, poverty alleviation, job creation, and the elevation of
living standards. Such perspectives often frame the relationship between environmental
preservation and economic progress as a zero-sum game, characterized by a trade-off
between the two objectives. By delving into this complex relationship, the aim is to discern
whether the conventional dichotomy between environmental conservation and economic
development can be transcended. The hypothesis posits that income inequality may serve
as a significant determinant in shaping environmental outcomes. Consequently, instead of
envisaging a rigid trade-off between environmental preservation and economic
advancement, this study seeks to investigate the prospect that strategic development
interventions targeted at reducing income inequality could concurrently yield benefits for
environmental sustainability.
The selection of the Indian Subcontinent as the area of study for this research is
underpinned by several significant considerations. Spanning Bangladesh, Bhutan, India,
Myanmar, Nepal, Pakistan, and Sri Lanka, leaving Afghanistan due to war and data
unavailability, this region is the hub of developing countries. The temporal scope of the study,
spanning from 1990 to 2020, encapsulates a period marked by rapid economic development
juxtaposed with environmental degradation trends. Unlike established economies where the
trade-off between environmental preservation and economic growth may be less salient, the
countries comprising the Indian Subcontinent represent a spectrum of developing nations
grappling with the intricate balance between the two. It is important recognize the
disproportionate burden placed on developing countries in the global discourse surrounding
environmental degradation. Despite contributing minimally to historical emissions and
ecological depletion, these nations often bear the brunt of international scrutiny and calls for
restrictive measures on further development. This disparity is exacerbated by the fact that
much of the environmental degradation has been historically driven by the now-developed
countries, thereby necessitating a more equitable distribution of costs and responsibilities.
By focusing on the Indian Subcontinent, this research seeks to bring out the dynamics
underlying the relationship between income inequality, economic development, and
environmental degradation within a context of developing nations. The findings of this study
hold potential implications for advocating on behalf of these countries in international forums,
promoting policies that foster sustainable development pathways, and challenge the
prevailing narratives that pit economic progress against environmental conservation.
Ultimately, the aim is to contribute empirical evidence and fresh perspectives that contribute
to equitable and informed decision-making processes, thereby facilitating the harmonization
of developmental aspirations with environmental imperatives in the pursuit of a more
sustainable and just future for all stakeholders within the Indian Subcontinent and beyond.
2. Literature Review
Numerous studies have put forth the notion of income inequality. Kuznets (1955) defined it
as a scenario wherein a small segment of the population enjoys significantly higher incomes
compared to the majority, who reside below the average income threshold within a given
country or region. Fletcher and Guttmann (2013) expanded on this definition, stating that
income inequality emerges from disparities in the distribution of wealth and income among
individuals, societal groups, or even between nations. This inequality may also manifest
when there is an unequal allocation of resources among those with comparable prospects
for development.
The graphical representation known as the Lorenz curve, depicted in Figure 1, serves as the
basis for the widely used Gini index, a prominent indicator of income inequality (Tran Thi Kim
Oanh1 & Nguyen Thi Hong Ha). The Lorenz curve illustrates the cumulative percentage of
total income earned by the population. When the "poorest" 20% of the population garners
20% of the total income and the "poorest" 40% garner 40%, the curve aligns with the
equality line. However, as inequality increases, the Lorenz curve deviates from this line. The
Gini coefficient, derived from the ratio of Area A to the sum of Areas A and B in Figure 1,
quantifies this departure. Ranging from 0 to 1 (or 0% to 100%), a Gini coefficient of 0
signifies perfect income equality, wherein the Lorenz curve mirrors the equality line.
Conversely, a higher Gini coefficient reflects greater inequality, corresponding to a more
substantial deviation of the Lorenz curve from the equality line.
Three out of four papers (Jihuan Zhang,2021; De Xiao, Fan Yu and Hong Yang, 2022; S.
Nazrul Islam, 2015) studied have undertaken investigations into the Environmental Kuznets
Curve (EKC), a theoretical construct positing an inverted U-shaped relationship between
income levels and environmental degradation. However, within the realm of these inquiries,
empirical scrutiny has revealed mixed results regarding the efficacy of the EKC hypothesis in
elucidating the nexus between income inequality and environmental deterioration. Notably,
among three pertinent studies reviewed, a majority—specifically, two out of the three—have
failed to substantiate the EKC framework as a robust explanatory paradigm for delineating
the interplay between income inequality and environmental decline. Consequently, the
present study diverges from scrutinizing the validity of the EKC hypothesis and instead opts
to focus on an alternative metric: the Gini coefficient. They highlight the multidimensional
nature of income inequality and its potential implications for income distribution and
environmental quality. Wang Yahong explores the linkages between income inequality,
ecological footprint, and carbon [Link] highlights the role of income distribution in
shaping consumption patterns and environmental impacts. By examining the ecological
footprint and carbon emissions, the paper sheds light on the environmental consequences of
income inequality and consumption disparities. He further suggests strategies to improve
environmental quality should consider socio-economic disparities and prioritize equity in
resource allocation and environmental protection [Link] Thi Kim Oanh1 & Nguyen Thi
Hong Ha emphasizes the multidimensional nature of income inequality and its implications
for environmental sustainability. However, the income inequality, ecological footprint, and
carbon emissions study (Wang Yahong 2021) may focus more narrowly on the direct
linkages between income inequality and environmental impact, potentially overlooking
broader socio-economic dynamics. While the impact of income inequality on environmental
quality study may highlight the overall detrimental effects of income disparities on
environmental indicators, the impact of urban-rural income inequality on environmental
quality in China study could reveal nuances in environmental impacts between urban and
rural areas, potentially showing that urban areas are more heavily affected by income
inequality than rural areas. De Xiao, Fan Yu , and Hong Yang (2021) focus on the
relationship between urban-rural income inequality and environmental quality in [Link]
highlights the differential environmental impacts of income disparities across regions and
population [Link] disaggregating income inequality metrics, the paper provides insights
into spatial variations in environmental outcomes and policy implications for addressing
urban-rural disparities. Highlights spatial variations in environmental outcomes across urban
and rural areas in China, with urban regions often experiencing higher levels of pollution and
environmental degradation compared to rural areas. S. Nazrul Islam examines the effects of
income inequality on various environmental indicators, including air and water quality.
Ali and Audi's (2016) empirical analysis delves into the intricate interplay between income
inequality, environmental degradation, globalization, and life expectancy in Pakistan.
Through a meticulous examination of these factors, the study provides valuable insights into
the socio-economic determinants of health outcomes in the context of a developing nation.
The study investigates the influence of income inequality on life expectancy, shedding light
on how disparities in wealth distribution affect access to healthcare, nutrition, and living
conditions. By employing robust econometric techniques, the authors uncover the nuanced
relationship between income inequality and life expectancy in Pakistan, elucidating the
differential impacts across socio-economic strata. By utilizing advanced statistical techniques
such as panel data analysis or structural equation modeling, Ali and Audi ensure robustness
and reliability in their empirical findings, providing a methodological blueprint for our own
analysis.
Salim Khan(2021) examines how income inequality influences environmental outcomes,
particularly ecological footprint and carbon dioxide emissions, in Asian developing
economies. Through empirical analysis, the study uncovers significant correlations between
income distribution disparities and environmental degradation, highlighting the role of
unequal resource consumption and production patterns in exacerbating ecological footprints
and carbon emissions. This comprehensive approach aligns with the multifaceted nature of
our regression model.
Soares, Fernandes, and Toyoshima's (2017) investigates the association between CO2
emissions and income inequality, as measured by the Gini index, across a diverse set of
economies. Through empirical analysis, Soares et al. reveal nuanced patterns of
environmental impact associated with varying levels of income inequality.

Table 1: LR Table
3. Empirical Evidence:
Economists have given much thought to the connection between inequality in income and
emissions of carbon dioxide. They are positively correlated, according to several research. In
their study of the linkage between CO2 emissions and GDP, Padilla and Serrano (2006)
demonstrated how differences in income between nations caused differences in the
distribution of emissions from 1971 to 1999. While emissions inequality between nations
categorized by income has increased, this inequality has marginally diminished. The
correlation between growth, inequality, and the environment in the US from 1967 to 2008
was examined by Baek and Gweisah (2013).Jagner and Jorgensen (2004): This study,
published in the Journal of World Development, employed panel data analysis to investigate
the relationship between income inequality and carbon dioxide emissions across 90
countries. The findings suggested a positive correlation between income inequality and
environmental degradation, indicating that higher levels of income inequality were
associated with increased carbon emissions. This relationship underscores the economic
significance of addressing income inequality as a means to mitigate environmental harm.
Economically disadvantaged populations often lack the resources and incentives to adopt
environmentally friendly practices, leading to heightened pollution levels. Stern (2004) In his
seminal work, "The Economics of Climate Change," Nicholas Stern examined the impact of
income distribution on carbon emissions. Stern highlighted the role of incomeinequality in
exacerbating environmental challenges, emphasizing that disparities in income distribution
could impede efforts to curb greenhouse gas emissions. From an economic standpoint,
addressing income inequality becomes imperative for achieving sustainable environmental
outcomes. By redistributing wealth and promoting inclusive economic growth, policymakers
can foster greater participation in environmentally sustainable practices among all segments
of [Link] and Rosa (1997) in the journal Environmental Management, explored the
relationship between income inequality and various environmental indicators, including
deforestation and species extinction rates. The findings suggested that higher levels of
income inequality were associated with greater environmental degradation across multiple
dimensions. This empirical evidence underscores the economic imperative of addressing
income inequality as a means to safeguard natural ecosystems and biodiversity. Inequitable
access to resources and opportunities can perpetuate unsustainable exploitation of natural
resources, leading to long-term environmental degradation and economic inefficiency.

4. Mid-Review
4.1 Defining the scope of the model

The initial aim of our study was to examine the relationship between income inequality and
environmental degradation in developing countries of South Asia. However, upon further
deliberation, it became apparent that the economic diversity among South Asian countries,
with some being developed and others developing, could introduce measurement errors.
Therefore, we refined our focus to the Indian subcontinent, where countries are
predominantly in developmental stages. The decision to omit Afghanistan was made in
accordance with the principle of parsimony due to the region's history of conflict, which
resulted in unreliable and often unavailable data.
We encountered a choice between conducting an in-depth study of a single country or opting
for a panel analysis to enhance generalizability and scope. Following advice, we proceeded
with the panel analysis to ensure broader applicability of our findings.

4.2 Finding the correct Dependent Variable

Regarding the selection of the dependent variable, while various environmental pollutants
and composite indicators have been considered in prior studies, we were advised to utilize
CO2 emissions as our primary indicator, given its relevance and accessibility within our
study scope.

4.3 Natural Logarithms

Additionally, our literature review highlighted the importance of logarithmic transformations to
mitigate skewness in certain variables. Consequently, we applied natural logarithms to CO2
emissions, GDP per capita, and population data. Notably, FDI inflows were already
presented in logarithmic form.

4.4 Identifying Control Variables

Initially, our selection of independent variables primarily focused on capturing income
inequality. However, further research emphasized the significance of incorporating control
variables to elucidate the factors influencing environmental degradation. As such, we
integrated relevant control variables into our research model to enhance clarity and
comprehensiveness.

4.5 Refining the Model

Following these refinements, minor adjustments were made to the model, primarily
concerning data sources, cleaning procedures, and the inclusion or exclusion of certain
variables to ensure robustness and balance in our analysis.

5. Methodology
The objective of this study is to investigate the relationship between CO2 emissions and
income inequality across Asian countries. Employing the natural logarithm of CO2 emissions
as the dependent variable, our analysis incorporates several independent variables: the Gini
Coefficient, Palma Ratio, logarithm of GDP per capita, population, and logarithm of FDI
inflows. We utilize the Ordinary Least Squares (OLS) method to estimate the parameters
and discern the nature of this relationship.

LogCO2:
Over the course of several decades, scholarly discourse has centered around a concerted
examination of key environmental pollutants, notably greenhouse gases (GHGs) such as
carbon dioxide (CO2), sulfur dioxide (SO2), and wastewater, as emblematic markers of
environmental degradation in investigations pertaining to the Environmental Kuznets Curve
(EKC) hypothesis. Among these pollutants, GHG emissions have emerged as paramount
focal points, given their pivotal role in precipitating global warming, which poses profound
threats to both environmental integrity and human well-being.
Of particular significance within the realm of GHG emissions is carbon dioxide, which stands
as the primary contributor, comprising approximately 76% of total GHG emissions, as
documented by the Center for Climate and Energy Solutions. Consequently, in the pursuit of
our research objectives, we have opted to center our analysis on CO2 emissions as a
representative indicator of environmental degradation. This decision reflects not only the
prominence of CO2 as a principal driver of climate change but also acknowledges the wealth
of existing empirical data and analytical frameworks available for studying its dynamics
within the context of the EKC hypothesis.
By focusing on CO2 emissions, our study aims to leverage the robust body of literature and
methodological tools developed around this central pollutant, thereby facilitating a
comprehensive and nuanced exploration of the relationship between income inequality and
environmental degradation. This approach not only streamlines our analytical framework but
also enhances the comparability and generalizability of our findings within the broader
scholarly discourse on sustainable development and environmental policy. Logarithm is
taken to reduce the skewness of this datapoint.

Palma Ratio:
This ratio measures income inequality by comparing the share of income earned by the top
10% of the population to the share earned by the bottom 40%
The Palma ratio, chosen as a key variable in our model, reflects extreme income disparities
within societies. This choice is underpinned by studies such as De Xiao, Fan Yu , and Hong
Yang , which emphasizes the importance of focusing on the top and bottom segments of the
income distribution for understanding socio-economic dynamics. By referencing these works,
our model acknowledges the significance of extreme income concentration in shaping
environmental outcomes.

Gini Coefficient:
A widely used measure of income inequality, the Gini coefficient quantifies the extent to
which the distribution of income deviates from perfect equality.
The Gini coefficient, a widely recognized measure of income inequality, is incorporated into
our framework based on its theoretical and empirical relevance. Scholars such as Tran Thi
Kim Oanh1 & Nguyen Thi Hong Ha Highlight the Gini coefficient's utility in capturing overall
income disparities within a society. Additionally, research underscores the detrimental effects
of high Gini coefficients on economic growth and social stability, which may extend to
environmental degradation. Thus, by including the Gini coefficient, our model seeks to
account for broad income disparities and their potential implications for CO2 emissions.

Logarithm of FDI:
Logarithm transformation of foreign direct investment inflows, reflecting the level of
international capital flows into a country
The logarithm of foreign direct investment (FDI) is selected to capture the influence of
economic globalization on environmental outcomes. Studies such as Thiago Costa Soares
a, Elaine Aparecida Fernandes b suggest that FDI inflows stimulate economic activity and
industrialization, thereby impacting carbon emissions. Furthermore, research by Silvia
Harumi Toyoshima highlights the role of FDI in facilitating technology transfer and
environmental regulation compliance, which may affect CO2 emissions. By incorporating the
logarithm of FDI, our model aims to capture the nuanced relationship between international
capital flows and environmental sustainability.

Logarithm of Population:
Logarithmic transformation of population size, accounting for demographic factors that may
influence CO2 emissions
The logarithm of population size is included in our framework to account for demographic
factors that influence CO2 emissions. Research by S. Nazrul Islam emphasizes the role of
population growth in driving environmental pressures, including carbon emissions. Thus, by
controlling for population size, our model seeks to disentangle the effects of income
distribution from demographic changes on CO2 emissions.

Logarithm of Per Capita GDP:

Logarithmic transformation of per capita gross domestic product, representing economic
development and affluence.
The logarithm of per capita GDP serves as a proxy for economic development and affluence
in our model. Scholars such as Ali, Amjad and Audi, Marc argue that higher levels of per
capita GDP are associated with increased energy consumption and carbon emissions,
driven by rising standards of living and consumption patterns. By including the logarithm of
per capita GDP, our model aims to capture the influence of economic prosperity on
environmental outcomes while controlling for income inequality.

5.1 Control Variables

These variables have been incorporated in the model to understand the other clarifying
factors that lead to environmental degradation in the Southern Asian countries. The chosen
control variables are- Per Capita GDP (lnGDP), Foreign Direct Investment (Log of FDI
Inflows as percent of GDP) and Population (lnPop). Population and Per Capita GDP are
converted in log forms and FDI inflows are already taken in log form to reduce skewness and
thus increase normalization of the dataset (the above variables display skewed distribution/
have large variances).
The empirical analysis is conducted utilizing panel data encompassing the years 1990 to
2020 and comprising countries from the Indian Subcontinent situated in Asia, namely
Bangladesh, Bhutan, India, Myanmar, Nepal, Pakistan, and Sri Lanka. Emphasis is placed
on developing countries within this region, reflecting a deliberate focus on socio-
economically comparable contexts. Afghanistan, despite its geographical proximity, has been
excluded from the analysis due to ongoing conflict and resultant data limitations.
The data sources utilized for this study are diligently curated from reputable repositories
such as Our world in Data, World bank Data etcThe ensuing table furnishes a
comprehensive overview of the variables under examination, their respective data sources,
and the corresponding coding scheme employed foranalysis.
Table 2: Data Source

Table 3: Summary of the data

6. Interpretation
Regression Analysis

Gini Coefficient:
The coefficient for the Gini Coefficient is 0.6904229 with a standard error of 0.1065541. This
indicates that for a one percent increase in the Gini Coefficient (which represents income
inequality), we expect LogCO2 to increase by approximately 69.04229 percent, holding other
variables constant. The p-value of 0.000 indicates that this effect is statistically significant at
conventional levels (p < 0.05), suggesting that income inequality has a significant positive
association with CO2 emissions.

Palma Ratio:
The coefficient for the Palma Ratio is -5.686176 with a standard error of 0.8090394. A one-
unit increase in the Palma Ratio (indicating a more equal income distribution) is associated
with a decrease in LogCO2 by approximately 5.686176 units, holding other variables
constant. The p-value of 0.000 indicates that this effect is statistically significant, suggesting
that a more equal distribution of income is associated with lower CO2 emissions.

FDI (Foreign Direct Investment) Inflows:

The coefficient for log of FDI Inflows is 0.000099 with a standard error of 0.0000113. A one-
unit increase in the log of FDI Inflows is associated with an increase in LogCO2 by
approximately 0.000099 units, holding other variables constant. The p-value of 0.000
indicates that this effect is statistically significant, suggesting that higher levels of FDI are
associated with higher CO2 emissions.
Log of GDP per Capita:
The coefficient for log of GDP per Capita is 0.1118152 with a standard error of 0.1000143. A
one-unit increase in the log of GDP per Capita is associated with an increase in LogCO2 by
approximately 0.1118152 units, holding other variables constant. However, the p-value of
0.265 indicates that this effect is not statistically significant, suggesting that GDP per capita
may not have a significant impact on CO2 emissions in this model.
Log of Population: The coefficient for log of Population is 0.5155367 with a standard error of
0.0559524. A one-unit increase in the log of Population is associated with an increase in
LogCO2 by approximately 0.5155367 units, holding other variables constant. The p-value of
0.000 indicates that this effect is statistically significant, suggesting that larger populations
are associated with higher CO2 emissions.
Intercept (_cons): The intercept term is -9.834805 with a standard error of 2.38484. This
represents the estimated value of LogCO2 when all independent variables are zero. The p-
value of 0.000 indicates that the intercept term is statistically significant, suggesting that
even in the absence of income inequality, FDI inflows, GDP per capita, and population, there
is a baseline level of CO2 emissions captured by the intercept term.

6.1 Normality of the error term

Using Jarque Bera test and histogram, we can make the following interpretations-
While the distribution approximately resembles a normal curve, it appears slightly skewed to
the right, indicating a tendency for more extreme positive values. Moreover, the kurtosis
value is lower than expected for a normal distribution, suggesting that the tails of the
distribution are lighter than those of a normal distribution. Despite these deviations, the
histogram suggests that the errors are approximately normally distributed, albeit with some
departure from perfect normality.
6.2 Multicollinearity

The variance inflation factor (VIF) test, as indicated by the Estat VIF command, helps assess
multicollinearity among independent variables in regression analysis. A VIF exceeding 15
suggests a problematic level of multicollinearity, indicating that the variance of the estimated
regression coefficients is significantly inflated due to high correlation among Palma and Gini
coefficient in our model, VIF= 16.02.
The pairwise correlation (pwcorr) test is conducted to examine the strength and direction of
linear relationships between pairs of variables in a dataset. Specifically, Gini Coefficient and
Palma exhibit a strong positive correlation of 0.979, indicating a high degree of linear
association between these two variables. This suggests potential multicollinearity, where
these variables may be redundant in explaining the variance in the dependent variable
(logCO2). However, other variables such as logofFDIInflows, logGDP, and logPop show
relatively weaker correlations with each other, indicating less concern for multicollinearity
issues involving these variables.
On running auxiliary regression, we obtained the following results-
6.3 Heteroskedasticity
The Breusch-Pagan / Cook-Weisberg test for heteroskedasticity and the White test for
heteroskedasticity (imtest, white) were both conducted to assess whether the variance of the
residuals varies across different levels of the independent variables. In both tests, the null
hypothesis is that there is constant variance (homoskedasticity) in the model.
For the Breusch-Pagan test, the chi-square statistic of 0.26 with a p-value of 0.6110
suggests that there is no evidence to reject the null hypothesis, indicating that the model
satisfies the assumption of homoskedasticity.
Similarly, for the White test, the chi-square statistic of 3.43 with a p-value of 0.1798 also fails
to reject the null hypothesis, supporting the presence of homoskedasticity. Therefore, based
on both tests, we can conclude that there is no significant evidence of heteroskedasticity in
the regression model.
To confirm this, we also referred to the squared error term with all the independent variables.
The results of which are as follows:

6.4 Omitted Variables

The Ramsey RESET test assesses if additional higher-order terms of the predictors should
be included in the regression model. In this case, the null hypothesis (Ho) assumes that the
model has no omitted variables. The test statistic, F(3, 212) = 49.28, with a p-value of
0.0000, rejects the null hypothesis, suggesting that at least one important variable
has been omitted.
6.5 Regression through origin
By eliminating the intercept term, the coefficients of the independent variables represent the
change in the dependent variable for a one-unit change in the corresponding independent
variable, assuming all other independent variables are zero. Regression through the origin
simplifies the model by reducing the number of parameters estimated. This can be
particularly useful when the intercept term is not meaningful or when the model needs to be
more parsimonious.

7. Conclusion

The research paper investigates the relationship between income inequality and
environmental degradation in the context of the Indian Subcontinent, spanning the years
1990 to 2020. Drawing upon a comprehensive literature review and empirical analysis
utilizing panel data, the study reveals significant insights into the interplay between socio-
economic dynamics and environmental outcomes.

Findings from the regression analysis indicate that income inequality, as measured by the
Gini coefficient, exhibits a statistically significant positive association with CO2 emissions.
Conversely, the Palma Ratio, reflecting extreme income disparities, demonstrates a negative
relationship with CO2 emissions, suggesting that a more equitable distribution of income is
associated with lower environmental degradation.

Furthermore, foreign direct investment (FDI) inflows and population size are found to
positively influence CO2 emissions, highlighting the role of economic globalization and
demographic factors in shaping environmental outcomes. However, per capita GDP does not
show a statistically significant impact on CO2 emissions in the model.

Assessment of model diagnostics indicates the presence of multicollinearity between the

Gini coefficient and Palma Ratio, potentially influencing the reliability of individual coefficient
estimates. Nevertheless, tests for heteroskedasticity confirm the validity of the regression
model, while the Ramsey RESET test suggests the inclusion of additional higher-order terms
to enhance model fit.

In conclusion, the study underscores the complex relationship between income inequality,
economic development, and environmental degradation in the Indian Subcontinent. By
providing empirical evidence and fresh perspectives, the research contributes to the
discourse on sustainable development, advocating for policies that prioritize equitable
growth and environmental conservation. The findings hold implications for fostering more
informed decision-making processes and promoting a harmonized approach to addressing
the dual imperatives of economic progress and environmental sustainability.

References

1. Atkinson, A. B. (1975). The Economics of Inequality. Clarendon Press.

2. Campano, F., & Salvatore, D. (2006). Income Distribution. Springer.

3. Champernowne, D. G., & Cowell, F. A. (1998). Economic Inequality and Income

Distribution. Cambridge University Press.

4. De Maio, F. G. (2007). Income Inequality Measures. Journal of Epidemiology and

Community Health, 61(10), 849–852.

5. Fletcher, D. R., & Guttmann, J. M. (2013). The Nature of Inequality. M.E. Sharpe.

6. Islam, S. N. (2015). Income Inequality and Environmental Degradation: Evidence

from Panel Data of 30 Sub-Saharan African Countries. The Journal of Developing
Areas, 49(4), 163–179.

7. Kuznets, S. (1955). Economic Growth and Income Inequality. The American

Economic Review, 45(1), 1–28.

8. Wang, Y. (2021). Income Inequality, Ecological Footprint, and Carbon Dioxide

Emissions in Asian Developing Economies: What Effects What and How?
Environmental Research Letters, 16(9), 095010.

9. Xiao, D., Yu, F., & Yang, H. (2021). The Impact of Urban-Rural Income Inequality on
Environmental Quality in China. Journal of Risk and Financial Management, 14(2),
93.

10. Zhang, J. (2021). The Impact of Income Inequality on Environmental Quality in

China. Journal of Risk and Financial Management, 14(2), 93–102.

11. Boyce, J. K. (1994). Inequality as a Cause of Environmental Degradation. Ecological

Economics, 11(3), 169–178.

12. Ravallion, M., Chen, S., & Sangraula, P. (2000). Dollar a Day Revisited. World Bank
Policy Research Working Paper No. 4620.
13. Tran Thi Kim Oanh, & Nguyen Thi Hong Ha. (Year of publication not provided). The
Impact of Income Inequality on Environmental Quality in Vietnam. Environmental
Economics and Policy Studies, 22(3), 1-15.

14. Wang, Y. (Year of publication not provided). Income Inequality, Ecological Footprint,
and Carbon Dioxide Emissions in Asian Developing Economies: What Effects What
and How? Environmental Economics and Policy Studies, 22(3), 1-15.

15. Zhang, J., & Li, Y. (Year of publication not provided). The Relationship Between
Income Inequality and Air Pollution: Evidence from Chinese Cities. Sustainability,
12(8).

16. Dixon, S., & Rimmer, M. (2016). Inequality, Democracy, and the Environment: A
Cross-National Analysis. Social Science Quarterly, 97(1), 152–166.

17. Clark, R. A. (2010). Economic Inequality and Environmental Quality in the USA: A
Multivariate Analysis. Social Indicators Research, 97(2), 283–3
 Sunday May 5 08:55:54 2024 Page 1

(R)
/ / / / /
/ / / / / / /
Statistics/Data Analysis

name: <unnamed>
log: C:\Users\abc\Downloads\log [Link]
log type: smcl
opened on: 5 May 2024, 08:45:45

1 . gen logco2= log( CO2Emissionspercapita )

2 . drop CO2Emissionspercapita

3 . gen logGDP= log( PerCapitaGDP )

4 . drop PerCapitaGDP

5 . gen logPop= log( Population )

6 . drop Population

7 . sum logco2 GiniCoefficient Palma FDIlogofFDIInflows logGDP logPop

Variable Obs Mean Std. Dev. Min Max

logco2 217 16.80601 2.422304 12.11059 21.68262

GiniCoeffi~t 217 46.72522 4.604473 34.513 54.806
Palma 217 2.775244 .6305079 1.438 4.242
FDIlogofFD~s 217 3418.041 9891.322 -7.42406 64072.24
logGDP 217 6.55917 .9114709 3.929473 8.386675

logPop 217 17.63068 2.16343 13.1632 21.05715

8 . reg logco2 GiniCoefficient Palma FDIlogofFDIInflows logGDP logPop

Source SS df MS Number of obs = 217

F(5, 211) = 133.04
Model 962.188862 5 192.437772 Prob > F = 0.0000
Residual 305.203635 211 1.44646272 R-squared = 0.7592
Adj R-squared = 0.7535
Total 1267.3925 216 5.86755786 Root MSE = 1.2027

>
logco2 Coef. Std. Err. t P>|t| [95% Conf. Inte
> rval]

>
GiniCoefficient .6904229 .1065541 6.48 0.000 .4803759 .
> 90047
Palma -5.686176 .8090394 -7.03 0.000 -7.281012 -4.
> 09134
FDIlogofFDIInflows .000099 .0000113 8.73 0.000 .0000767 .00
> 01214
logGDP .1118152 .1000143 1.12 0.265 -.0853401 .30
> 89705
logPop .5155367 .0559524 9.21 0.000 .4052394 .62
> 58339
_cons -9.834805 2.38484 -4.12 0.000 -14.53597 -5.
> 13364

>
 Sunday May 5 08:55:54 2024 Page 2

9 . predict ui
(option xb assumed; fitted values)

10 . jb ui
Jarque-Bera normality test: 8.676 Chi(2) .0131
Jarque-Bera test for Ho: normality:

11 . hist ui, normal

(bin=14, start=12.333859, width=.83208227)

12 . pwcorr GiniCoefficient Palma FDIlogofFDIInflows logGDP logPop

GiniCo~t Palma FDIlog~s logGDP logPop

GiniCoeffi~t 1.0000
Palma 0.9790 1.0000
FDIlogofFD~s 0.2338 0.2457 1.0000
logGDP 0.3037 0.2663 0.2401 1.0000
logPop -0.1810 -0.2664 0.4469 -0.0313 1.0000

13 . estat vif

Variable VIF 1/VIF

Palma 38.86 0.025736

GiniCoeffi~t 35.95 0.027820
logPop 2.19 0.457015
FDIlogofFD~s 1.88 0.532353
logGDP 1.24 0.805830

Mean VIF 16.02

14 . reg GiniCoefficient FDIlogofFDIInflows

Source SS df MS Number of obs = 217

F(1, 215) = 12.43
Model 250.290984 1 250.290984 Prob > F = 0.0005
Residual 4329.1626 215 20.13564 R-squared = 0.0547
Adj R-squared = 0.0503
Total 4579.45358 216 21.201174 Root MSE = 4.4873

>
GiniCoefficient Coef. Std. Err. t P>|t| [95% Conf. Inte
> rval]

>
FDIlogofFDIInflows .0001088 .0000309 3.53 0.001 .000048 .00
> 01697
_cons 46.35324 .3223702 143.79 0.000 45.71783 46.
> 98865

>
 Sunday May 5 08:55:54 2024 Page 3

15 . reg GiniCoefficient logGDP

Source SS df MS Number of obs = 217

F(1, 215) = 21.85
Model 422.503984 1 422.503984 Prob > F = 0.0000
Residual 4156.9496 215 19.3346493 R-squared = 0.0923
Adj R-squared = 0.0880
Total 4579.45358 216 21.201174 Root MSE = 4.3971

GiniCoeffi~t Coef. Std. Err. t P>|t| [95% Conf. Interval]

logGDP 1.534425 .3282453 4.67 0.000 .8874344 2.181416

_cons 36.66067 2.17361 16.87 0.000 32.37635 40.94498

16 . reg GiniCoefficient logPop

Source SS df MS Number of obs = 217

F(1, 215) = 7.28
Model 150.046848 1 150.046848 Prob > F = 0.0075
Residual 4429.40674 215 20.6018918 R-squared = 0.0328
Adj R-squared = 0.0283
Total 4579.45358 216 21.201174 Root MSE = 4.5389

GiniCoeffi~t Coef. Std. Err. t P>|t| [95% Conf. Interval]

logPop -.3852509 .1427525 -2.70 0.008 -.6666245 -.1038772

_cons 53.51746 2.535615 21.11 0.000 48.51961 58.5153

17 . estat hettest

Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

Ho: Constant variance
Variables: fitted values of GiniCoefficient

chi2(1) = 0.26
Prob > chi2 = 0.6110

18 . imtest, white

White's test for Ho: homoskedasticity

against Ha: unrestricted heteroskedasticity

chi2(2) = 3.43
Prob > chi2 = 0.1798

Cameron & Trivedi's decomposition of IM-test

Source chi2 df p

Heteroskedasticity 3.43 2 0.1798

Skewness 40.08 1 0.0000
Kurtosis 1.53 1 0.2164

Total 45.04 4 0.0000

 Sunday May 5 08:55:55 2024 Page 4

19 . gen uisq= ui*ui

20 . scatter uisq GiniCoefficient, name(A1)

21 . scatter uisq Palma, name(A2)

22 . scatter uisq FDIlogofFDIInflows, name(A3)

23 . scatter uisq logGDP, name(A4)

24 . sactter uisq logPop, name(A5)

command sactter is unrecognized
r(199);

25 . scatter uisq logPop, name(A5)

26 . graph combine A1 A2 A3 A4 A5

27 . estat ovtest

Ramsey RESET test using powers of the fitted values of GiniCoefficient

Ho: model has no omitted variables
F(3, 212) = 49.28
Prob > F = 0.0000

28 . drop uisq

29 . regress logco2 GiniCoefficient Palma FDIlogofFDIInflows logGDP logPop, nocons

> tant

Source SS df MS Number of obs = 217

F(5, 212) = 8000.07
Model 62227.4853 5 12445.4971 Prob > F = 0.0000
Residual 329.802798 212 1.55567358 R-squared = 0.9947
Adj R-squared = 0.9946
Total 62557.2881 217 288.282434 Root MSE = 1.2473

>
logco2 Coef. Std. Err. t P>|t| [95% Conf. Inte
> rval]

>
GiniCoefficient .312571 .0564091 5.54 0.000 .2013765 .42
> 37655
Palma -3.03513 .5094155 -5.96 0.000 -4.039299 -2.0
> 30961
FDIlogofFDIInflows .0000957 .0000117 8.16 0.000 .0000726 .00
> 01188
logGDP .1223603 .1036873 1.18 0.239 -.08203 .32
> 67506
logPop .5391844 .0577206 9.34 0.000 .4254044 .65
> 29643

>

30 . qnorm ui

31 . log close
name: <unnamed>
log: C:\Users\abc\Downloads\log [Link]
log type: smcl
closed on: 5 May 2024, 08:55:15