This paper attempts to estimate the environmental Kuznets curve (EKC) in the case of Mauritius by taking the role of carbon emissions and its implications on the Gross Domestic Product (GDP) of the country, together with variables linked directly and indirectly to GDP and carbon emission which includes inflation rates, unemployment and population size.
This paper examines the relationship between economic growth and carbon emissions in Mauritius for the period 1990-2010. The multivariate regression techniques approach as the estimation and approach to cointegration was adopted. This paper examines the data through differencing techniques in order to make the data stationary which ultimately helped us to manipulate the data effectively.
After running correlation techniques, we found that growth and carbon emissions are highly correlated and quadratic equation formulation was used in the prediction modelling. It was deduced that, Mauritius has reached its "turning point". However due to the heavy use of fuels and high carbon emissions in Mauritius it is very difficult for the country to meet Kyoto Protocol requirements.
Various conclusions are drawn out of this paper which includes national income which has the most significant impact in environmental degradation, many indicators like unemployment rate, tends to improve as countries approach middle income levels and econometric evidence shows that macroeconomic policies have certain impact on income and carbon emissions. Similarly forecasted CO2 emissions were derived for different sectors.
Keywords: Environmental Kuznets Curve, Carbon Dioxide Emissions, Gross Domestic Product, Time Series, Mauritian Economy.
Introduction
The object of this research paper is to investigate empirically the validity of the Environmental Kuznets Curve hypothesis for Mauritius with a time span of more than two decades.
According to the EKC hypothesis as the economy of one's country develops the degradation of the environment increases, but when the economy reaches a specific level of income per capita, known as turning point, pollution starts to decline. The EKC hypothesis implies that despite the fact that at first stages of development, pollution is unavoidable, in the end the economic growth will be one of the solutions to the pollution problem.
The concept that economic development will eventually lead to the improvement of the quality of the environment is very appealing. Growth, which has been accused as the main cause of the environmental degradation and now is seen as a "savior" of the environment has spurred the interest of policy designers. Instead of hampering the growth of the economy, measures that make economy to grow even faster is what needs.
Environmental Kuznets Curve
Environmental Kuznets Curve (EKC) is the relationship that is assumed to trace the pollution path followed by countries as their per capita gross domestic product (GDP) grows and describes the relationship between per capita income and of environmental degradation indicators (Unruth and Moomaw, 1998). In the infant stages of development, the levels of some pollutants climb with increases in per capita income, while at advanced levels of development, environmental degradation follows a downward trend as income per capita is moving upwards. These results give rise to a bell shaped curve relating economic growth to environmental degradation, redolent of the relationship hypothesized by Kuznets (1995) between economic and income inequality (Nahman and Antrobus, 2005). The concept of EKC came from the early with Grossman and Krueger's (1991) path-breaking study of the potential impacts of NAFTA (North American Free Trade Agreement).
Origins of the EKC
The environmental Kuznets curve consists of a hypothesized relationship between different indicators of environmental degradation and income per capita. At first stage of economic growth, degradation and pollution increase, but further than some level of income per capita, the movement shift, that is for substantial high income of the country it leads to higher levels of economic growth leading to environmental improvement. This means that the impact of environmental indicator is an inverted U shape function of real income per capita (Stern, 2004)
In other words, the distribution of income becomes more asymmetrical in early stage of income growth and then the distribution moves towards greater equality as economic growth continues (Kuznets, 1955). This liaison between income per capita and income inequality can be represented by a bell-shaped curve. This is viewed as an empirical phenomenon known as the Kuznets Curve (Dinda, 2004).
Criticism and drawbacks of the Kuznets Curve
The Kuznets Curve has helped in studying the relationship between environmental pollutants and GDP of countries, but it does have drawbacks too. Even Kuznets (1955) himself indicated that the Kuznets Curve Theory is not a perfect one and the relationship between income inequality and economic development cannot be assumed. He also declared that lot information in the paper has been speculated and thus further research work must be carried out.
The reason behind the development of the Environmental Kuznets Curve
Since the last decades, the increasing threat of global warming and climate change has been of major continuing concern. Organisations such as the United Nations have been trying to diminish the unfavourable impacts of global warming through intergovernmental and binding accords. After immense negotiations, the agreement namely the Kyoto protocol was signed in 1997. This protocol has the objective of reducing greenhouse gases (GHG) that cause climate change. The Kyoto protocol recognises limitations to environmental pollutants and necessitates a timetable for realisation of the emission reductions for the developed countries. During 2008 - 2012 periods the demands reduction of the GHG emissions to 5.2 % lower than the 1990 level. In 2005 it came into force: 178 states have signed and approved the protocol since April 2008 (Halicoglu, 2008). Greenhouse gas emissions particularly carbon dioxide (CO2) emissions, are considered to be the core causes of global warming. Consequently, to prevent global warming a number of countries have signed the Kyoto Protocol and agreed to diminish their emission levels. Galeotti and Lanza (1999) indicated that some developing states refused to sign the Kyoto Protocol based on the argument that the industrialisation and development process should be subject to no constraints, particularly for energy production and consumption.
One probable foundation for this position is the belief that while pollution increases with growth in GDP, it happens a point where pollution goes down. This view calls for a careful analysis of the relationship between economic growth and pollution. This relationship is obviously very complex as it depends on numerous different factors such as:
The country's size,
The sectoral structure, including the composition of the demand for energy,
The vintage of the technology,
The demand for environmental quality,
The level and quality of environmental protection expenditures.
Shafik (1994) reported that the relationship between economic growth and environmental quality has been a source of great disagreement for a lengthy period of time. On one side it has been observed that greater economic activity unavoidably leads to environmental degradation and finally to possible economic and ecological collapse. At the other side is the view that those environmental nuisances worth solving will be tackled more or less automatically as a consequence of economic growth.
Previous to 1970, there was a conviction that the raw materials consumptions, energy and natural resources were growing at the same pace as economy grew. In the early 1970s, the Club of Rome's Limits of Growth view (Meadows et al. 1972) was brazen about the concern for the accessibility of natural resource of the Earth. They argued that the finiteness of ecological resources would prevent economic growth and advocated for a solid state economy with zero growth to avoid striking ecological circumstances in the future. This view has been criticised on both hypothetical and empirical grounds. Experimental works showed that the ratio of consumption of some metals to income was falling in developed countries during the 1970s, which brought divergence with the predictions set out in the Limits to Growth view (Maleness, 1978). Natural environment not only provide natural resources important for economic development but also execute the vital function of supporting life, if man persist to exploit environment recklessly, then it would not be able to sustain life any longer.
Environmental Kuznets Curve definition and graphical illustration
The EKC follows the name of Nobel Laureate Simon Kuznets who had remarkably hypothesized an inverted 'U' income-inequality relationship (Kuznets, 1955). In the 1990s economists detected this relationship between economic growth and environmental degradation. Since then this relationship is known as Environmental Kuznets Curve.
According to the EKC theory as a country develops, the pollution increases, but after reaching a specific level of economic progress pollution begin to decrease. The EKC hypothesis suggests that environmental degradation is something unavoidable at the first stage of economic growth, so a developing country is forced to tolerate this degradation in order to develop. In a graphical representation the x-axis symbolize the economic growth which is measured by GDP per capita and the y-axis represents the environmental degradation which is measured by many different pollution indicators like carbon dioxide, nitrogen oxide, sulphur dioxide, deforestation etc.
The shapes of the Environmental Kuznets Curves
The relation between income and environmental pressure can be sketched in several ways; firstly, one can distinguish monotonic and non-monotonic curves. Monotonic curves may show either mounting pollution with rising incomes, as in the case of municipal waste per capita or decreasing. But, non-monotonic patterns may be more probable in other cases and two types have been recommended, namely inverted-U and N-shaped curves. The pattern discovered in experiential research depends on the types of pollutants scrutinised and the models that have been used for inference. Four speculative opinions are presented in favour of an inverted-U curve for (local) air pollutants, which are listed as:
Positive income elasticity's for environmental quality,
Structural changes in production and consumption,
Rising information on environmental consequences of economic activities as income rises and
More international trade and more open political systems with increasing levels of income (Selden and Song 1994).
Others, for example Pezzey (1989) and Opschoor (1990), have argued that such inverted-U relationships may not hold in the long run. They anticipated a so-called N-shaped curve which demonstrated the same pattern as the inverted-U curve initially, but beyond a certain income level the relationship between environmental pressure and income is positive again. Delinking is thus considered a temporary phenomenon. Opschoor (1990), for example, argued that, once technological efficiency enhancements in resource use or abatement opportunities have been exhausted or have become too expensive, further income growth will result in net environmental degradation. Despite these considerations empirical evidence so far has been largely in favour of the inverted-U instead of the N shaped relationship (de Bruyn et al., 1998).
The shortcomings of EKC analysis
A number of critical studies of the EKC literature have been published (e.g. Coodoon and Dinda, 2002; Ekins, 1997; Fare et al., 2001; Permanand Stern, 2003; Stern, 2004).
Theoretical critiques
This section discusses the criticisms that were raised against the EKC on theoretical (rather than methodological) grounds.
One of the main criticisms of the EKC models is the assumption that environment and growth are not interrelated. In simple words the EKC hypothesis assumes no feedback between income and the pollution of environment.
The key criticism of Arrow et al. (1995) and other criticism was that the EKC model as presented in the 1992 World Development Report and somewhere else assumes that there is no respond from environmental damage to economic production as income is assumed to be an exogenous variable. The declaration is that environmental damage does not lessen economic activity sufficiently to stop the growth process and that any irreversibility is not too severe to reduce the level of income in the future. In other words, there is a theory that the economy is sustainable. But, if elevated levels of economic activity are not sustainable, attempting to grow fast in the early stages of development when environmental degradation is rising may prove counterproductive.
According to Ekins (1997), consideration in assessing the strength of the estimation is the reliability of the data used. However, there is little sign that the data problems are serious enough to shed doubt on the basic environment-income link for any particular environmental indicator, but the results in fact imply that this might be the case.
Stern (2004) drew his attention to the mean - median problem. He underlined that early EKC studies showed that a number of indicators: 2 SO emissions, x NO, and deforestation, peak at income levels around the current world mean per capita income.
A hasty glimpse at the available econometric estimates might have lead one to believe that, given likely future levels of mean income per capita, environmental degradation should turn down from the present onward. Income is not yet, normally distributed but very skewed, with much larger numbers of people below mean income per capita than above it. Hence, this shows a median rather than mean income that is the relevant variable.
Another problem related with the EKC studies is the little attention that has been paid to the statistical components of time series analysis. Very few studies in the past investigated the presence of unit root in time series of variables used to investigate the validity of the EKC.
Econometric critiques
Stern (2004) in a survey argued that, the econometric criticisms of the EKC falls into four main categories: heteroscedasticity, simultaneity, omitted variables bias, and cointegration issues.
Perman and Stern (2003) investigated the data and models for unit roots and cointegration respectively. Panel unit root tests designated that all three series - log sulfur emissions per capita, log GDP capita, and its square - have stochastic trends. Results for cointegration are less definite. About half the individual country EKC regressions cointegrate but many of these have limitations with "incorrect signs". Some panel cointegration tests point out cointegration in all countries and some accepted the non-cointegration hypothesis. However, even when cointegration is found, the form of the EKC relationship varies radically across countries with many countries having U-shaped EKCs. In case there's a common cointegrating vector in all countries it will be strongly rejected.
Coondoo and Dinda (2002) carried out an analysis for Granger Causality between CO2 emissions and income in various individual countries and regions. In general, model that emerged was that causality runs from income to emissions or that there is no significant relationship in developing countries, while in developed countries causality runs from emissions to income. Still, in every case the relationship is positive so that there is no EKC type effect.
Waheed et al. (2006) research began testing the presence of a unit root in each of the macroeconomic series using the Augmented Dicky-Fuller (1979). The ADF test constructs a parametric correction for higher-order correlation by assuming that the series follows an AR(k) process and adding lagged difference terms of the dependent variable. There had been a propose simple modification of the ADF approach to construct DF-GLS test, in which the time series are detrended so that explanatory variables were "taken out" of the data prior to running the test regression. Also Lopez, (2009) analysis included a median-unbiased estimation based on Augmented-Dickey-Fuller (ADF) regressions with an extension of median-unbiased estimation to the DF-GLS regression of Elliott, Rothenberg, and Stock (1996).
Data and Time Series Properties
To study the relationship between the GDP of Mauritius and the CO2 emission in Mauritius the annual data that are being used are; total CO2 emission from 1976 to 2008, the real GDP from 1976 to 2008, the population of Mauritius from 1976 to 2008, inflation rate of Mauritius and the unemployment rate of Mauritius.
Per Capita CO2 Emission Estimates for Mauritius graph
Figure : Per Capital CO2 Emission Estimate for Mauritius data from Energy data book (2010)
From these sets of data it can be clearly seen that while population was increasing (Figure 2), during these years the real GDP (Figure 3) has been fluctuating a bit. In mid 1970's after the independence there has been a lot of development and transformation in the country. The economy was diversified and more jobs were created.
Furthermore, the country received more foreign aids. By the late 1970's the economy deteriorated a bit mainly due to the increase in petroleum price in the world market and this lead to less government subsidies and devaluation of Mauritian Rupees. Then by late 1980's, the economy experienced steady growth and also a high level of employment, declining inflation and more domestic savings.
This period was also marked by the boom in the sugar industry. Though the development slowed down in the 1990's there was a gradual development of the local financial institutions and at the same time the domestic information and telecommunication industry boomed. By the start of the 21st century there the financial services sector became a very important pillar of the economy with an increasing number of offshore enterprises. Finally the economy developed a lot due to the seafood processing and export during the last 10 years.
Figure : Population Estimates for Mauritius from World Bank
Figure : GDP for Mauritius (without inflation) from World Bank
In the short term, real GDP is affected by inflation (Figure 4) because the latter causes a rise in general price of goods and services and consequently this causes a change in investments, savings, consumption and import and export of a country and thus the GDP of a country are affected too. The equation used to calculate real GDP is;
GDP =consumption by private individuals+ total investment + government spending excluding inflation + [total exports - total imports] (Wikipedia, 2010)
Figure : Inflation in Mauritius from the Federal Bank of Cleveland
GDP also depends on unemployment rate (Figure 5) because according to this equation GDP = compensation of employees + gross missed income + gross operating surplus + taxes less subsidies on production and imports (Wikipedia 2010). Thus if unemployment rate increases in a country, the GDP will decrease.
Figure : Unemployment Rate in Mauritius from the index mundi (2010)
While trying to prove the relationship between GDP and the CO2 emission of Mauritius, it can also be observed how the GDP affects the CO2 emission in each specific sector of Mauritius. Figure 6 shows how the Energy sector and the Transport sector are the main contributors to CO2 emission from 2000 to 2006.
Figure : CO2 emissions per sector from CSO Mauritius (2010)
Methodology and estimation techniques:
According to the EKC hypothesis, there is a nonlinear quadratic relation that exists between income of a country and carbon emissions. However, since other variables than income can also exist as the determinants of CO2 emissions, we added other variables which could have an influence on CO2 as well as the income of the country.
Econometrics methods were applied as well as time series processes to understand the data behaviour and stationarity. Unit root testing would be used to test the proposition that in autoregressive time series the parameter is always one.
Also unit root analysis would be used to ensure that the data for the different estimates that will be used in the model follow the same mean and variance, if not, differencing and redifferencing techniques would be applied to make the data stationary. The step for differencing and redifferencing includes the following general formulae which are used again to test the stationarity of the processes.
\Delta y_t = \alpha + \beta t + \gamma y_{t-1} + \delta_1 \Delta y_{t-1} + \cdots + \delta_p \Delta y_{t-p} + \varepsilon_t,
The Augmented Dicker Fuller Test (ADF) would be applied because the set of data is a larger one and a complicated set of time series model. The ADF test will be used to test the null hypothesis of stationarity and to investigate the possibility that a time series is seasonal and fractionally integrated.
The DF-GLS test will also be performed before running the regression analysis for both univariate modeling and multivariate modelling.
Likewise for the specifications of the ADF test, we will be using the Bayesian Information criteria (BIC), the Akaike Information Criteria (AIC) and the Ng-Perron Modified AIC (MAIC) ,were used to test the results for different lag lengths respectively.
Multivariate regression techniques were applied so as to be able to build the regression curve as well as the regression formulae. The multivariate model regression is a single parameter estimation model that attempts to estimate the regression outcome with more than 2 variables.
Similarly, quadratic equation models were developed so as to smooth the Kuznets curve as well to be able to develop forecasting procedures and this was developed through the use of matrices. The quadratic equation will be used to find the turning point of the Kuznets curve and predict how the GDP and Carbon Dioxide level may vary the future.
After running a series of multivariate regression analysis we will then be able to devise a formula that would be help in predicting and forecasting of the future level of CO2 at different levels of GDP forecasts and for different specific sectors such as the Tourism sector, financial services, manufacturing, construction and others.
Hence we will be able to predict the level of carbon emission for different sectors as well as we will be able to forecast at a future level of carbon emissions what will the GDP of the country be, that is suppose we want to predict following a certain level of carbon emissions from the Tourism sector we will be able to forecast the level of GDP for that specific period of time.
Cointegration regressions would also be applied to examine the dynamic relationship between GDP per capita and CO2 emissions as well as other variables added to the multivariate regression model including inflation rate and the rate of unemployment in Mauritius. We will perform a number of tests for cointegration of the different variables to the level of carbon emissions of Mauritius.
We will also forecast through the Auto Regressive Disturbance Lag and we will test autoregression and multicollinearity, which the steps are shown in the diagram below; autoregression and multicollinearity, which the steps are shown in the diagram below;
Figure : Steps in time series modelling
Analysis and Findings
Stationarity and Unit Root Analysis of CO2 emisisions and Real GDP in Mauritius:
In this section test the time series properties of CO2 emissions and GDP were analyzed to see whether they are driven by some process and exhibit unit behaviour. We first tested the stationarity properties of the trends and we applied a series of unit root tests.
Unit Root tests:
Augmented Dicker Fuller Test (ADF)
The first unit root test that would be applied is ADF (1979) test. The ADF test constructs a parametric correction for higher-order correlation by assuming that the series follows an AR(k) process and adding lagged difference terms of the dependent variable to the right-hand side of the test regression (Waheed et al., 2006). The unit root hypothesis (a = 0) can be tested according to the following models:
In the above models, the term was used to as to take into account the time lags to accommodate for a correlation.
DF-GLS Test
Elliot, Rothenberg and Stock (1996) proposed a simple modification of the ADF approach to construct DF-GLS test, in which the time series are de-trended so that explanatory variables are "taken out" of the data prior to running the test regression (Waheed et al., 2006). Ng and Perron (2001) argued that this test is more powerful.
Various hypotheses were used to analyze the data and the main unit test that was used was the ADF test. Since the data trends were not following a stationary process, time lags had to be used so as to smooth the data. The time lags were determined using both the Akaike Information Criteria (AIC) and the Schwarz's Bayesian Information Criteria (BIC). These information criterion were used so as to be able to devise time lag properties for the date trends of CO2 and GDP.
The following figure illustrates the ADF test for CO2;
Figure : ADF for CO2
Figure 8 shows the ADF test for CO2 and it can be observed that the time lad lengths taken were two. The AIC and BIC verified that a time lag of two is enough to be able to make the data follow a stationary process and hence a time lag of two was the optimal lagged difference as per the two information criterion.
The following figure shows the ADF test for GDP;
Figure : ADF for GDP
Figure 9 shows the ADF test for GDP and it can be observed at time lag two the data went on to follow a stationary process. Again, the lagged differences were obtained following the AIC and BIC approach so as to determine the optimal lags.
To increase the efficiency and reliability of the ADF test, the equivalent test which is the DF-GLS test proposed by Elliot et al. (1996) was also applied. Under the DF-GLS test, time lags were chosen as per the Ng-Perron modified AIC (MAIC), the Schwarz's Criterion (SIC) and the Ng-Perron sequential t approach.
The following figure, analyses the DF-GLS test for CO2:
Figure : DF - GLS test for CO2
Figure 10 shows the DF-GLs test for CO2 and it can be observed that the maximum time lag devised is same as the ADF test which is two and the minimum SIC and MAIC is one. The MAIC was used, SIC when lags are set to be minimized and when a trend term is not included.
The following figure shows the DF-GLS test for GDP;
Figure : DF - GLS test for GDP
Figure 11 shows the DF-GLS test, whereby minimum time lag entered here is 1 and the SIC and the MAIC too is one. Reapplication of these tests indicates that both the variables now follow a stationary process.
Univariate Model Equation:
The following figure shows the regressed CO2 over GDP so as to be able to draw a regression curve and to formulate and equation from it.
Figure : Regressed CO2 over GDP
Figure 12 justified the regression equation which is as follows:
The above equation is derived from figure 5 and (X) represents the figure to be inserted concerning GDP so as to be able to get the equivalent amount of CO2.
The following graph shows the actual relationship of CO2 and GDP matched over a period of twenty years ranging from 1990 to 2010.
Figure : Graph showing the relationship of CO2 and GDP
The above figure depicts the movement of GDP in regards to CO2 level of Mauritius over a period of 20 years. The figures in the above graph concerning GDP are in Millions of rupees while that of CO2 is in metric tons.
From Figure 13, it can be deduced that the more GDP rises, the higher is the level of CO2 emissions. However it can also be deduced that in some specific moments, when GDP has fallen, the level of CO2 even fell. Hence, it can be observed that there is a positive relationship between the level of GDP and carbon emissions. Likewise, it should be noted that this graph is a non smoothed one, and the smoothed one has been done through a quadratic equation so as to be able to derive the Kuznet's curve and the turning point of Mauritius, whereby though GDP is rising but the level of carbon emissions will fall. This is discussed in further sections.
Multivariate Model Equation;
Under the Univariate model only GDP and CO2 were taken as variables, but there are various variables that affect GDP and CO2 amongst which are Inflation and Unemployment. Inflation is an important variable to take into account as it erodes the value of GDP and makes it appear to be a larger value. That is why we have taken inflation in the multivariate modelling so as to be able to consider the effects of inflation on the real value of GDP.
Likewise, Unemployment as a variable was also taken into consideration because unemployment has a direct impact on the level of GDP. The more the level unemployment is high the lower is the amount of GDP and the lower the rate of Unemployment, the higher is the level of GDP.
The figures and data trends for both inflation and unemployment were tested for stationarity and it was found that after ADF test the data were stationary and as per the AIC and BIC no time lags were necessary.
The following figures show the ADF test for inflation trends in Mauritius:
Figure : ADF test for inflation trends in Mauritius
As per figure 14, it can be observed that no time lags were assigned as following the AIC and BIC approach, it was not necessary to include a time lag and hence with zero time lags the data was stationary.
The following figure shows the ADF test for Unemployment trends in Mauritius:
Figure : ADF test for Unemployment trends in Mauritius
As per figure 15, it can be deduced that again no time lags were taken into consideration as per the AIC and BIC criteria and there was no need for data smoothing procedures.
The following figure shows the multivariate model, taking into account inflation and unemployment. Likewise, we have broke GDP into two main sectors, namely the manufacturing and transport sector and the construction which are the highest contributors of CO2 and the other sector is the financial services sector, real estate and others.
Figure : Multivariate regression
Figure 16 shows the multivariate regression equation which is as follows:
Following this equation, appropriate forecasting may be made about the level of CO2 in the main sectors of Mauritius.
The following figure shows the relationship between CO2 emissions matched against the GDP of the Manufacturing, Transport and Construction Sector;
Figure : Relationship between CO2 emission and GDP of manufacturing
The above figure depicts the relationship that exists between CO2 emissions and the level of GDP in the Manufacturing, Transport and Construction Sector. It can be found that there is an increasing trend.
The following figure depicts the relationship between the level of CO2 emissions and the level of GDP in the financial services sector, real estate and others including tourism also.
Figure : Relationship between the level of CO2 emission and level of GDP in Financial services sector, real estate and others including tourism
Quadratic Equation and Prediction
To formulate a quadratic equation the GDP and CO2 data were input in Microsoft Excel and a matrix model was used.
Thus the quadric equation was derived to be;
Y= -0.0000000600872 X2 + 0.027229158 X + 295.8429621
To be able to find the maximum point, i.e. where is the turning point of the quadratic curve (where the emission of Carbon Dioxide has reached its peak) the above quadratic equation was differentiated.
= 2(-0.0000000600872) X + 0.027229158 + 0
At turning can be equated to zero, thus
If = 0
2(-0.0000000600872) X + 0.027229158 + 0 = 0
X =
X =226,580.353
Consequently from this value of X i.e. GDP, the year at which the turning point took place can be derived by looking at the original set of data of Year, GDP and CO2.The year at which the turning point took place was between 2007 and 2008.
Furthermore to create a curve predicting the Carbon Dioxide emission the equation Y= -0.0000000600872 X2 + 0.027229158 X + 295.8429621 was input in the programming language software MATLAB. The prediction was done for till the value of GDP will reach 300,000 millions.
DiagramKey
Figure : Kuznets graph including forecasted
The above diagram shows the scatter plot of the Kuznets graph, the Quadratic curve and the prediction part of the curve is dotted. In the graph, it can be seen that in the beginning with the increase of GDP there has also been an increase in the carbon dioxide emission of Mauritius. As the economy of Mauritius has been growing the manufacturing industry, transport industry, tourism sector and other sectors too have underwent growth. With more factories setting up and vehicles on the roads, the carbon emission of the country has increased consistently. Combustion of fossil fuels is one of the main causes of carbon dioxide emission and coal, petroleum, and natural gas have been heavily used during the last decades to produce energy. At the same time, with economic growth and increase in GDP, people have started earning more and thus they also consumed more and there have been a lot of changes in their lifestyles. The latter contributed in the increase in carbon emission. Level of education and literacy rate has also been an important factor because better educated people have a tendency to have better paid job and they have a better life style and their way of living changes.
But at the same time with an increase in education level has also helped people in becoming more environment-friendly and environment conscious.
The Ministry of Environment and Ministry of Education have during the last few years carried out a lot of sensitization campaigns to make people aware about environment and pollution and its consequences. New acts and various policies were introduced and implemented in Mauritius concerning the environment and its protection (as shown in the graph below). In Environmental Protection Act 2002 provided a legal framework to protect and management of natural environment and its resources.
This act was later on amended in 2008. In 2004 the Dangerous Chemical Act which deals with all dangerous chemical substances was implemented in Mauritius. Moreover, as from June 2008 the government of Mauritius introduced the concept of 'Maurice Ile Durable'. This concept mainly relates to building a green Mauritius and it includes plans like decreasing pollution by using wind and hydro energy instead of burning fossil fuels.
The campaign is also focusing a lot on sustainable development, which is about preserving the environment and its natural resources for the generations to come while at the same time fulfilling the needs of the people.
At the same time due to the financial crisis which started in 2007 lead to a recession which affected the Mauritius directly and indirectly through the textile industry, tourism industry, construction industry among others.
Thus the financial crisis may also be a reason why the level of carbon dioxide emission has started decreasing as from 2007. Another factor responsible for the decrease in carbon dioxide emission can be the fact that from 2006 to 2007 the price of petroleum has increased from $50 to $140. This may have lead to a slight decrease in the use of petroleum and consequently the carbon dioxide emission too was affected.
Comparison of Mauritius EKC to that of UK and USA EKC
EKC of Mauritius
The above figures show the EKC curve of UK, EKC curve of USA and EKC curve of Mauritius respectively. It can be clearly seen that all three of them are following the same pattern and have a similar form, i.e. at the start there was an increase of carbon dioxide emission with the increase in GDP of the country and then it reached a point where the amount of carbon dioxide emission was constant while the GDP kept increasing and finally at a certain time the carbon dioxide emission started decreasing although the GDP kept increasing.
Limitation of Study and Recommendation
As all other research work this study too has its limitations. In the equations formulated there were no error correction models. There were also other factors which affect the GDP of a country apart from unemployment level and inflation. For example foreign exchange rate could have been taken into consideration when formulating the multivariate equation. When there is devaluation in a country's currency its exports become more expensive while its imports become cheaper.
This causes a fluctuation in the GDP of a country. Last but not the least the sustainability and "Maurice Ile Durable" concept are something too new and recent in Mauritius. It is quite difficult to quantify it so quickly and actually say that it has caused a change in our environment and the level of carbon dioxide and pollution emission.
