FOREIGN TRADE UNIVERSITYFALCULTY OF INTERNATIONAL ECONOMICS---ECONOMETRICS 1 REPORTASSIGNMENT FOR MIDTERM EXAMTopic: ANALYZING THE DEPENDENCY OF ECONOMIC GROWTH ON INDICATORS FOR A CIRCU
OVERVIEW OF THE TOPIC
Theoretical framework (Related economic theories)
Theoretical framework (Related economic theories)
2.1 Definition and indicators of CE
The circular economy is a method of producing and consuming goods that emphasize on 3Rs: reducing, reusing, and recycling existing materials and products as long as possible When a product reaches the end of its life, its materials are kept within the economy wherever possible Unlike the traditional model where profits are the priority, the circular economy is a framework for systems-level solutions to problems including climate change, biodiversity loss, waste generation, and pollution The phrase CE was first introduced in a Pearce and Turner (1990) study that examined the connections between environmental issues and economic activities However, Boulding's work (1966), which developed the concept of a closed system to highlight the finite natural resources accessible for human activity, is where the CE principles originate.
Many nations are focusing on the development of a circular economy because it is crucial to establish a sustainable way of life without impeding global economic progress The benefits of a circular economy are the ability to benefit the local economy, drive employment growth, and promote resource independence
According to McDonough and Braungart (2002), there are two cycles in which resource loops flow make up the design of a closed loop system like the circular economy: the “technical cycle” and the “biological cycle” The “technical cycle” refers to the process in which inorganic materials can stay in the loop without losing any value The “biological cycle” refers to organic materials that can decompose without any detrimental effects on the environment
Circular economy is an umbrella concept (CIRAIG, 2015; Blomsma and Brennan, 2017; Homrich et al., 2018), which means a single definition or indicator to classify this concept is unachievable The classification framework of CE divides the indicators into 2 groups: the direct indicators and the indirect indicators The direct
CE indicators with Specific Strategies (Graedel et al.,2011) focus on measuring material and waste production The indirect indicators mainly measure the relevant areas of CE but not certainly including circularity This framework provides the explanation of the proposed indicators such as: Labour productivity, Trade in recyclable raw materials, use of renewable energy, recycling rate of municipal waste, and resource productivity.
2.2 Impact of circular economy’s indicators on economic growth
According to a paper by Mihail Busu and Carmen Lenuta Trica, the authors created a regression model using multiple variables of CE indicators Six independent factors of CE indicators being used in this model include: labor productivity, recycling rate of municipal rate, resource productivity, trade in recycling materials, circular material use rate, and environmental tax revenue The result of this model is the significant impact of the selected CE indicators on the economic growth, which is the real GDP per capita In the other paper by Mihail Busu, the author also creates a regression model based on the impact of CE indicators on economic growth Other variables included in this model are use of renewable energy and the share of innovative enterprises The result of this model is that all the independent factors of
CE indicators have significant impacts on the growth in the real GDP per capita in the country
Based on the previous proven theories of the circular economy, all the variables such as resource productivity, trade in recyclable raw materials, real labour productivity, and recycling rate of municipal waste are believed to have an impact on economic growth To be more precise, every variable in this model is expected to have a positive impact on economic growth (real GDP per capita), which is the dependent variable in this model
H1: Resource productivity have a positive effect on real GDP per capita
H2: Trade in recyclable raw materials have a positive effect on real GDP per capita H3: Recycling rate of municipal rate have a positive effect on real GDP per capita H4: Labor productivity have a positive effect on real GDP per capita
Research hypotheses
1.1 Method used to collect data
Using Excel and Stata software to briefly process data and compare matrix matches between variables.
1.2 Method used to analyze data
We run Stata software to model regression using Ordinary Least Squares (OLS) method to estimate the parameters of multivariable regression models From Stata software, we are easily:
Use Correlation matrix in Stata software to figure out correlation matrix among variables.
Use F-test to test the overall significance of variables in the model.
Use t-test to test the fitness between average value and expected value.
Use the test that examines the equality of two regression coefficients.
Based on theories of economic growth through GDP as well as ways of calculating GDP, at the same time inheriting previous studies and combining with the ability to collect research data for research solving the variables, we estimate the determinants of economic growth with the linear regression model.
METHODOLOGY AND MODEL SPECIFICATION
Empirical Model
Based on theories of economic growth through GDP as well as ways of calculating GDP, at the same time inheriting previous studies and combining with the ability to collect research data for research solving the variables, we estimate the determinants of economic growth with the linear regression model.
GDPpc i = � 0 + � 1 RP i + � 2 LP i + � 3 TRRM i + � 4 RRMW i + u i
GDPpc i : represent the economic growth
RP i : represent the resource productivity
LP i : represent the labor productivity
TRRM i : represent the trade in recyclable raw materials
RRMW i : represent the recycling rate of municipal waste
: the estimator of intercept of the model
: the estimator of coefficient �j with j = (1,2,3,4)
: the residuals - estimator of disturbance ui
As opposed to the traditional economy, CE is developing economic models to better use the resources In our study, five macroeconomic factors to describe CE were used, and their impact on economic growth was analyzed In our model, the five indicators will be the independent factors of the regression model.
Table 1: Definition of variables used in the model
Variable Name Meaning Unit Expected sign Dependent variable
Measurement of the total economic output of a country divided by the number of people and adjusted for inflation, representing the economic growth
Gross domestic product divided by domestic material consumption Euro/kg (+)
The productivity per person employed in relation to EU average
Trade in recyclable raw materials
The trade in recyclable raw materials (tons) imported from extra boundaries of
Recycling rate of municipal waste
Measures the tonnage recycled from municipal waste divided by the total municipal waste arising
Real GDP per capita is referred to as a dependent variable (GDPpc):
The impacts of circular economy indicators on economic growth have been analyzed by many researchers It was proven that there is a close link between the use of circular economy and economic growth (Browne et al 2009).
Moreover, economic growth is a key factor in the economic development of a country The growth rate of a country appears in the value of the Gross Domestic
Product (GDP) per Capita (Elistia et al 2018) Because of those mentioned above, the Real GDP per capita (GDPpc) will be used to analyze the correlation between Circular Economy indicators and Economic Growth.
One of the most important indicators of the CE is the productivity of resources (resource productivity) (Blomsma and Brennan 2017) This is defined as the ratio of a country's GDP to the domestic consumption of materials and shows us the economy's efficiency in the 27 EU member states to use materials to produce well- being (Haas et al 2015) The value of resource productivity (RP) is calculated in euro/kg
Some authors concluded that human capital has a positive impact on economic growth (Grossman and Krueger 1995; Brock and Taylor 2005; Lyasnikov et al 2014) Hence, another fundamental indicator of the CE is given by the real labor productivity representing the productivity per person employed in relation to the EU27 average and calculated in percentage (%) units.
Trade in recyclable raw materials (TRRM): measured by the tonnes of recyclable raw materials imported by the European Union from non-EU countries.
Trade in recyclable raw materials is considered as an exogenous variable in the Circular Economy that affects EU economic growth since more countries are likely to prefer using recyclable raw materials to protect the environment.
In a circular economy, residual materials are recycled and re-injected into the economy as new raw materials - then called 'secondary raw materials' This may have several benefits, both reducing wastes and increasing the security of raw
Data
Recycling rate of municipal waste (RRMW):
The municipal waste recycling rate is an indicator of integrated resource use, the growth of which shows a reduction in the consumption of primary materials and a decrease in waste disposal to landfills (in favor of their recycling)
The recycling rate is calculated as the percentage of municipal waste generated that is recycled, composted, and anaerobically digested, and might also include preparing for reuse In research for establishing the link between RRMW and economic growth, Eglantina Hysa, Alba Kruja, Naqeeb Ur Rehman and Rafael Laurenti (2020) indicated that the recycling rate of municipal waste was found to be significant for EU28 and positively impacted the GDP per capita It should be mentioned that the significance level was within 1% As a result, municipality management is being considered as a direct indicator in the Circular Economy as well as a crucial factor in economic growth in the EU From these theories, RRMW is chosen to test its effect on EU economic growth.
To investigate the topic, our group collected time-series data of 27 countries in the European Union in the period of 2010-2020 Since the United Kingdom officially left the EU in 2020, we excluded all UK data from the dataset, resulting in a total of
293 observations The data of all five variables in the model, which are Real GDP per capita (GDPpc i ), Resource productivity (RP i ), Real labor productivity (LP i ),
Trade in recyclable raw materials (TRRM i ), Recycling rate of municipal waste (RRMW i ), originates from EUROSTAT
3.2 Descriptive statistics and variables interpretation
To estimate the model determined the dependent relationship between the indicators of the circular economy and economic growth in the EU-27, we decided to choose the research sample in the 11-year period from 2010 to 2020 with a total of 297 observations.
A statistical description of the indicators used in the regression model - i.e., minimum, maximum, mean, and standard deviation - can be seen in Table 2.
Variable Obs Mean Std Dev Min Max
With 297 observations, from 2010 to 2020, we can see that the EU now is more concentrating on sustainable development with a relative high figure of the indicators for circular economy:
GDPpc i : the average GDP per capita is 25288.33 euro per capita, the highest value is 84750 euro, the lowest is 6.41 euro The huge difference between these data illustrates a diverse distribution of different levels in the data
RP i : the average consumption of resources in the EU in this period is about 1.74 euro/kg, with the standard deviation being relatively low, varying from 0,2996 to 4.9711 euro/kg It shows us the efficient use of materials to produce well-being.
LP i : the average productivity per person employed is about 95.89%, with the standard deviation of 29.67123%, the highest and lowest value is 211.7% and 41.4% respectively.
TRRM i : the average tonnes of recyclable raw materials imported by the
European Union from non-EU countries is 1566487 tonnes, with the standard deviation of 1983702 tonnes, the highest value is 9358093 tonnes.
RRMW i : the average recycling rate of municipal waste is around 34.5%, with the highest value reaching the number of 68.3% and the standard deviation is about 15.6%.
Table 3 below clearly illustrates the correlation, which is the degree of linear association between the dependent variable and the independent variables
Table 3: Correlation matrix between variables
GDPpc i RP i LP i TRRM i RRMW i
As can be seen from the table, the correlation between the dependent variable and independent variables are all positive and relatively high
r(GDPpc i , RP i ) = 0.7867 Real GDP per capita and Resource productivity are positively and closely linearly related, as expected.
r(GDPpc i , LP i ) = 0.9140 The correlation coefficient is quite high, which indicates a strong relationship between GDP per capita and Labor productivity.
r(GDPpc i , TRRM i ) = 0.2346 Real GDP per capita and Trade of recyclable raw materials are positively and loosely related, as expected Although in the SRM, the coefficient of TRRM is negative (-0.0018) illustrating the inverse relationship between GDPpc and TRRM, the positive correlation value has been supported by the related theory Those contrasting result will be explained below (in 3.2.1).
r(GDPpc i , RRMW i ) = 0.5781 Real GDP per capita and Recycling rate of municipal waste are positively and relatively closely related, as expected.
As for the correlation among independent variables, they are higher compared to the expectation The correlation between labor productivity and resource productivity(r(LP,RP) = 0.7953) is the strongest, while the correlation between labor productivity and trade in recyclable raw materials (r(LP, TRRM) = 0.2908) is the weakest However, when testing the equality of the regression coefficients of LP and
RP, it still shows the different significance of labor productivity and resource productivity on GDP per capita.
ESTIMATED MODEL AND STATISTICAL INFERENCES
Estimation result
Run the command reg GDPpc RP LP TRRM RRMW, we have:
Source SS df MS Number of obs = 293
GDPpc Coef Std Err t P>t [95% Conf Interval]
From STATA output, we get results as followed:
Total sum of squared (TSS): 8.2343e+10
Explained sum of squared (ESS): 7.1813e+10
Residual sum of squared (RSS): 1.053e+10
Sample regression model
After using the result from STATA, we have the sample regression model as follow:
GDPpc i : Real GDP per capita (Euro per capita)
RP i : Resource productivity (Euro/kg)
TRRM i : Trade in recyclable raw materials (Tons)
RRMW i : Recycle rate of municipal waste (%)
: the residuals - estimator of disturbance ui
Hypothesis testing
2.1 Testing the fitness of results compared to expectations
= -21363.56 : When the value of all independent variables in the model equals to 0, the average value of the expected real GDP per capita of the EU-
27 is -21363.56 euro per capita This is the average effect of other variables not included in the model on real GDP per capita of the EU-27 from 2010 to 2020.
= 5109.578: The relationship between GDPpc i and RP i is positive as expected It illustrates that when resource productivity rises by 1 euro/kg while others remain unchanged, the average value of real GDP per capita increases by 5109.578 euro per capita.
= 371.3648: The relationship between GDPpc i and LP i is positive as expected, which is supported by theory It illustrates that when labor productivity per person employed being higher than EU average (in 2020) rises by 1% while others are constant, the average value of real GDP per capita increases by 371.3648 euro per capita Growth in labor productivity indicates a higher level of output for every hour worked, thus, labor productivity is a key driver of economic growth and changes in living standards.
= - 0.0018: When the value of trade in recyclable raw materials (imported from non-EU countries by the European Union) rises by 1 tonneton, while others are constant, the average value of the expected GDP per capita decreases by 0.0018 euro per capita The relationship between GDPpc i and
TRRM i is negative, which is not supported by the theory and goes against the hypothesis Although the result is not expected, it can be explained by the formula below:
GDP = C + I + G + (X – M) in which M represents Imports (2)
Looking at those formulas, we figure out that the imports might have a negative relationship with GDP in the (2) formula, thus are negatively related to the GDP per capita.
However, when we test the individual correlation between GDPpc andTRRM, the result is 0.2346 (table 3), which aligns with the related theory found by the previous authors The reason may lie in the ‘suppressor effect' that has been categorized by Falk and Miller (1992) in the Partial LeastSquares (which is based on OLS algorithm, as MLR) field They stated that when the path coefficient [regression coefficient] and the correlation between latent constructs do not have the same sign, the original relationship between the two has been suppressed Therefore, TRRM has a positive correlation with GDPpc individually but this original relationship among these has been suppressed in the SRM.
= 143.237: When the recycling rate of municipal waste rises by 1%, while others remain unchanged, the average value of the expected GDP per capita increases by 143.237 euro per capita The relationship between GDPpc i and
RRMW i is positive as expected, which is in line with the theory The growth of municipal waste recycling rate means the optimization of integrated resource use It shows a decline in the consumption of primary materials and waste disposal to landfills, benefitting the production process, hence increasing the GDP per capita.
143.237: When the recycling rate of municipal waste rises by 1%, while others remain unchanged, the average value of the expected GDP per capita increases by 143.237 euro per capita The relationship between GDPpc i and
RRMW i is positive as expected, which is in line with the theory The growth of municipal waste recycling rate means the optimization of integrated resource use It shows a decline in the consumption of primary materials and waste disposal to landfills, benefitting the production process, hence increasing the GDP per capita.
2.2 Testing significance of an individual regression coefficient
H0: βj = 0 (The coefficient is not statistically significant)
H1: βj ≠ 0 (The coefficient is statistically significant)
With P-value acquired in the table and α = 5%
RP: P-value < α (0.000 < 0.05), reject H , thus the coefficient is statistically0 significant, which means that resource productivity influences real GDP per capita
LP: P-value < α (0.000 < 0.05), reject H , thus the coefficient is statistically0 significant, which means that real labor productivity has an impact on real GDP per capita
TRRM: P-value < α (0.000 < 0.05), reject H , thus the coefficient is0 statistically significant, which means that trade in recyclable rate has an impact on real GDP per capita
RRMW: P-value < a (0.000 < 0.05), reject H , thus the coefficient is0 statistically significant, which mean that recycling rate of municipal waste has an impact on real GDP per capitaRRMW: P-value < α (0.000 < 0.05), reject H ,0 thus the coefficient is statistically significant, which means that the recycling rate of municipal waste has an impact on real GDP per capita
Conclusion: At α = 5%, all the independent variables in this regression model are statistically significant.
Or it can be rewritten as:
In which R is the extent to which all the independent variables jointly 2 explain the variation in the dependent variable In this case, 87.21% of the total variation in Real GDP per capita is jointly explained by Resource productivity, Labor productivity, Trade in recyclable raw materials andRecycling rate of municipal waste
At 1%, 5% and 10% significance level, Stata yielded the result for p-value of Fs=0.0000