OVERVIEW OF THE TOPIC
Historically, many researches have been carried out about the quantity of Vietnam’s timber and wooden products being exported to foreign nations and the factors that have impact on it.
Vietnam is a prominent player in the global wood export market, ranking as the second largest exporter in Asia and fifth worldwide, according to a study by the General Statistics Office of Vietnam The country contributes to 6% of the global timber and wooden furniture market, which is valued at approximately $130 billion.
Besides, many trade agreements and documents that have been signed so far will set up legal corridors and mechanisms to encourage market expansion for the forestry sector.
The future of wood exporting in Vietnam shows significant development potential, yet there is a lack of comprehensive research addressing all factors influencing the forestry sector, particularly in the timber and wooden furniture trading market To bridge this gap, our study focuses on "The factors affecting the quantity of Vietnam’s timber and wooden products exported to foreign nations," aiming to uncover how these elements impact the export volumes of timber and wooden products.
Gross Domestic Product (GDP) is the total monetary or market value of all the finished goods and services produced within a country's borders in a specific time period.
A population is the number of organisms of the same species that live in a particular geographic area at the same time, with the capability of interbreeding; here
Timber harvesting encompasses several key processes, including the planning of harvest and reforestation, cutting down trees, and transporting them to a designated landing area This process also involves the processing, sorting, and loading of timber for transportation In countries that import timber from Vietnam, the annual timber harvested area refers to the specific acreage of forest land that is utilized for wood extraction each year.
International economic relations serve as a vital information system within today's information society, functioning as a social and market-based control mechanism that clearly illustrates its current dynamics This system significantly influences global alliances, drives globalization, and impacts the economic well-being of nations.
Economic distance encompasses two interconnected definitions: it serves as a relative indicator of household well-being, calculated as a percentage of median income, and it also represents the absolute disparity in per capita income among different social groups.
International trade, encompassing the export of goods and specifically timber, plays a crucial role in the GDP of many countries Economists have developed various models to forecast the dynamics of international trade and assess the impact of different factors on export activities.
International trade theory, a crucial area within economics, examines the patterns and origins of global trade, along with its effects on welfare Since the 18th century, international trade policy has sparked significant debate and controversy, continuing to be a relevant topic in contemporary discussions.
International trade theory and economics itself have developed as means to evaluate the effects of trade policies.
Adam Smith was an 18th-century philosopher renowned as the father of modern economics and a major proponent of laissez-faire economic policies The core of
Smith's thesis posits that humans' inherent self-interest fosters prosperity He contended that allowing individuals the freedom to produce and exchange goods through free trade, alongside encouraging competition both domestically and internationally, leads to greater economic success than imposing strict government regulations.
This free-market force became known as the invisible hand, but it needed support to bring about its magic.
Adam Smith's theory of trade highlights that countries engage in commerce when they possess an absolute advantage in producing certain goods compared to others Absolute advantage occurs when one country can produce a good using less labor than another, leading to more efficient production and trade between nations.
The Ricardian Model of Trade is developed by English political economist David Ricardo in his magnum opus On the Principles of Political Economy and Taxation
In 1817, David Ricardo introduced the first formal model of international trade, providing a theoretical framework that emphasizes the logic of comparative advantage This concept is crucial in economics and has been recognized for its significance; Nobel laureate Paul Samuelson famously identified it as a fundamental and non-trivial proposition in the social sciences.
The Ricardian Model, like other economic theories, is built on fundamental assumptions, including the existence of only two countries and the production of two goods using a single input: labor This labor is limited and perfectly immobile due to strict border controls Each country experiences a constant opportunity cost between the goods, represented graphically by a straight-line production possibility frontier Additionally, the model assumes the absence of transaction and transportation costs, which together create the conditions for comparative advantage, as long as the relative production costs differ between the two countries.
Under autarky, each country produces a mix of two goods without trade However, when trade is introduced, countries are incentivized to specialize in the production of goods where they hold a comparative advantage, optimizing the use of their limited resources As a result, both countries can consume more of both goods than they could without trade, leading to increased satisfaction and welfare The Ricardian Model ultimately demonstrates that international trade is beneficial for all participating countries.
The Heckscher-Ohlin model, also known as the H-O model or 2x2x2 model, is an economic theory that suggests countries export goods they can produce most efficiently and abundantly This model is utilized to assess trade dynamics and the equilibrium between two nations with differing specialties and natural resources.
The model advocates for countries to export goods that utilize their abundant factors of production while importing items that they cannot produce efficiently It suggests that nations should focus on exporting surplus materials and resources, while proportionately importing the resources they require, promoting a balanced trade approach.
The Heckscher-Ohlin model analyzes trade equilibrium between countries with different specialties and natural resources, illustrating how nations should engage in trade when resource distribution is uneven globally This model extends beyond commodities, incorporating various production factors, including labor By providing a mathematical framework, the Heckscher-Ohlin model guides countries on optimal operations and trade practices amid resource imbalances.
It pinpoints a preferred balance between two countries, each with its resources.
MODEL SPECIFICATION
Methodology
In this analysis, we will investigate the impact of independent variables on the dependent variable through a multiple regression model To facilitate this examination, we have chosen to utilize STATA software due to its user-friendly interface and ease of use for data analysis.
Panel data is advantageous for several reasons, primarily because it enhances the number of observations compared to traditional time-series and cross-sectional data Additionally, it enables more detailed research and provides better control over unobserved factors that may differ across subjects This capability is essential for minimizing errors and variances in estimations, leading to more accurate and reliable results.
This data choice is also going to lessen the chance of multicollinearity in variables and make the parameters’ estimators more precise and accurate.
To estimate parameters in a panel data model, we employ three distinct methods: the Ordinary Least Squares (OLS) method, the Fixed-Effect (FE) model, and the Random-Effect (RE) model Our team utilized STATA 14 to establish and test the regression model effectively.
Theoretical model specification
Utilizing the theoretical framework and insights from previous research outlined in Section I, we developed econometric models to analyze the impact of various factors on the volume of timber exports to select trading partners of Vietnam.
Q = f (GDPi, POPi, Si, Dis) (PRF) Q = β0 + β1 ∗ GDPi + β2 ∗ POPi + β3 ∗ Dis + β4 ∗ Si + ui
The equation for the Supply Response Function (SRF) is represented as Q = ^ β 0 + ^ β 1∗ GDPi + ^ β 2 ∗ POPi + ^ β 3 ∗ Dis + ^ β 4∗ Si + ui, where Q denotes the quantity of wood exported to importing countries in thousands of USD The model includes several key variables: GDPi, which reflects the Gross Domestic Product of importing countries measured in trillion USD; POPi, representing the population of those countries; Dis, indicating the geographical distance between Vietnam and the importing countries in kilometers; and Si, which signifies the area covered by timber-producing forests in square kilometers The coefficients β0, β1, β2, β3, and β4 are constants and slope coefficients corresponding to each independent variable, while ui accounts for the disturbance term in the equation.
To explain the variables we have the following table:
Table 2.1: Variables description, proxies to measure and their units
No Variable Meaning Variable type Unit Expected sign
1 Q Quantity/The amount of wood exported to importing countries
2 GDPi Gross Domestic Product (GDP) of importing countries
3 POPi Population Independent variable People +
Geographical distance between Vietnam and the partner countries importing wood (kilometre)
5 Si Land covered with timber- producing forests
Square kilometre - c Dependent variable? Independent variables? Their theoretical relationships? o Dependent variable: Q o Independent variables: GDPi, POPi, Si, Dis
An independent variable (GDPi, POPi, Dis, Si) is the variable that is changed or controlled in a scientific experiment to test the effects on the dependent variable.
In a scientific experiment, the dependent variable (Q) is the one being tested and measured, while independent variables, such as GDP, population, geographical distance from Vietnam to importing countries, and land covered with timber-producing forests, influence the results Changes in these independent variables can affect the amount of wood exported, illustrating a cause-and-effect relationship It is essential to note that while both variables can change during an experiment, the independent variable is controlled by the experimenter, whereas the dependent variable responds to these changes.
Describe the data
a Specify the source(s) of data
The research utilizes secondary and panel data to analyze factors influencing wood production exports to seven key markets Key variables include national income, population, geographical distance from Vietnam to importing countries, and the area of timber-producing forests The data spans from 2001 to 2016, providing a comprehensive overview of these dynamics.
Table 3.1: Source of data Name of variable Source of data
Q https://wits.worldbank.org/Default.aspx?lang=en
GDPi https://www.worldbank.org/
POPi https://www.worldbank.org/
Dis https://www.worldbank.org/
The World Bank provides comprehensive descriptive statistics and interpretations for each variable, detailing their meanings and measurement methods Additionally, it offers statistical indicators that summarize the key aspects of these variables, facilitating a better understanding of the data presented.
Command Sum indicates the number of observations (Obs), mean, standard deviation (Std Dev), min and max of variables. o Tab : distribution of values of variables
- The amount of wood exported to importing countries:
The amount of wood exported to importing countries ranged from 2752.781 to809388.7 thousand USD with the same frequency (1) and percent (89%).
- Gross Domestic Product (GDP) of importing countries
GDP ranged from 378.3761 to 6203 trillion USD with the same frequency (1) and percent (89%)
Population ranged from 1.94e+07 to 1.28e+08 people with the same frequency (1) and percent (89%).
- Geographical distance between Vietnam and the partner countries importing wood
Distance ranged from 3109 to 11565 kilometre with the same frequency (1) and percent (89%).
- Land covered with timber-producing forests:
Timberland ranged from 29674 to 3477568 square kilometre with the same frequency
The correlation matrix reveals significant relationships between various variables and quantity (Q) The correlation coefficient between GDP and quantity is 0.2884, indicating a positive effect of GDP on Q In contrast, the correlation coefficient between population and quantity is 0.5700, suggesting a strong positive effect of population on Q Conversely, the correlation coefficient between distance and quantity is -0.5055, demonstrating a negative impact of distance on Q Additionally, the correlation coefficient between land size and quantity is -0.2276, further indicating a negative effect of land size on Q.
Population is the most significant independent variable affecting the total quantity of wood exported to importing countries Both GDP and population exhibit a positive correlation with the amount exported, while distance and size have a negative impact on the quantity Overall, the correlation signs of all variables align with the expected outcomes.
Furthermore, all correlation coefficients are smaller than 0.8, which proves that there is no perfect multicollinearity in this model.
ESTIMATED MODEL AND STATISTICAL INFERENCES
This research examines the relationship between five key factors and the quantity of wood exported from Vietnam, highlighting that each factor influences export levels differently Utilizing data from the World Bank, the study analyzes various countries to establish these associations and gain insights into Vietnam's wood export dynamics.
First of all, by using command reg: reg QUSThousand GDPiUSTrillion POPiMillion Diskm Sisqkm, we have these results:
Following the results by using regression model by OLS in STATA, we have below Sample Regression Funcion:
The meaning of Coefficient of each independent variables:
94431 In the case that all other variables equal 0, the quantities of wood that Vietnam exported is
In terms of other variables are constant, If GDP of wood importing countries increase 1 Trillion USD more, total amount of Vietnam wood exporting would increase 25.6 USD β3
In terms of other variables are constant, if the population of wood importing country increased
1 thousand people, total amount of Vietnam wood exporting would increase 0.00177613 USD β4 -27576.15
In terms of other variables are constant, if the distant between wood importing country and Vietnam increased 1 km, total amount of Vietnam wood exporting would increase 27576.15 USD Β5 18.2681
In terms of other variables are constant, if the timberland of wood importing decreased 1 km 2 ,total amount of Vietnam wood exporting would decrease 18.2681 USD
RSS (Residual Sum of Squares) unexplainted factors: 1.3356e+12
Total Sum of Squares: TSS= ESS+RSS 2.8264e+12
Total degrees of freedom: 111, includes: The statistic has 2 numerator and
71 denominator degrees of freedom o 2: consumed by the model o 71: consumed for the residual
R 2 or coefficient of determination is the measurement of how well the sample regression line fit the data.
The regression analysis yields an R-squared (R²) value of 0.5272, indicating that approximately 52.72% of the variance in the dependent variable can be explained by the independent variables The adjusted R-squared is 0.5095, which accounts for the degrees of freedom in the model Additionally, the root mean squared error (Root MSE) is reported as 1.1e+05, representing the square root of the mean squared error for the residuals in the analysis of variance table.
The R² value of 0.5272 indicates that 52.72% of the variation in the quantity of Vietnam wood exports can be accounted for by the five independent variables in the model, while the remaining percentage is influenced by other factors not included in the analysis.
2.1 Ramsey RESET Testing The model may omit some vital variables that we could not figure out.
Therefore, we are about to testing whether out model omitted any important variables by using Ramsey RESET Testing
Hypothesis { H 0 : Model has no omitted variables
Using command ovtest in STATA, we have these results:
As we can see clearly that P-value=0.0658>5% cannot reject H 0
Conclusion: the model has no omitted variables
2.2 Multicollinearity testing a Theoretical Consequences of Multicollinearity
The term multicollinearity is due to Ragnar Frisch Originally it meant the existence of a
“perfect” or exact, linear relationship among some or all explanatory variables of a regression model.
When the assumptions of the classical model are met, Ordinary Least Squares (OLS) estimators are considered BLUE, meaning they are the Best Linear Unbiased Estimators Although multicollinearity does not violate any regression assumptions, it complicates the estimation of coefficients with small standard errors As a result, unbiased and consistent estimates are still achievable, and their standard errors can be accurately assessed Therefore, testing for multicollinearity is essential to understand its impact on the reliability of coefficient estimates.
To assess the presence of multicollinearity in our model, we will utilize the Variance Inflation Factor (VIF) in this report The VIF quantifies the extent to which the variance of an estimator is increased due to multicollinearity among the predictors.
As a rule of thumb, if the VIF of a variable exceeds 10, which will happen if R i 2 exceeds 0.90, that variable is said to be highly collinear
Therefore, by observing the STATA result, the mean VIF = 3.03 which is smaller than
10, we can conclude that multicollinearity doesn’t exist in our model.
Conclusion: multicollinearity doesn’t exist in our model.
A fundamental assumption of the classical linear regression model is that the disturbances, denoted as u i, possess constant variance, σ 2 When this assumption is violated, it leads to a condition known as heteroskedasticity While heteroskedasticity does not compromise the unbiasedness and consistency of Ordinary Least Squares (OLS) estimators, it does affect their efficiency, meaning they are no longer the Best Linear Unbiased Estimators (BLUE) Therefore, testing for heteroskedasticity is essential to ensure the reliability of regression analysis.
In this report, our group choose White’s General Heteroskedasticity Test to detect whether our model suffer from Heteroskedasticity or not.
By using command imtest, white in STATA software, we obtain the following result:
Prob > chi2 = 0.0000 < 0.05 we reject null hypothesis and accept alternative hypothesis H 1 Heterokedasticity is detected.
This regression model has heteroskedasticity problem It calls for the correcting solution.
White’s Heteroskedasticity corrected standard errors are also known as Robust standard errors Applying robust command in STATA we could alleviate heteroskedasticity
Fix the problem of heteroskedasticity
There are two popular method to fix the problem of heteroskedasticity:
One effective approach to address the issue is to utilize alternative functional forms, such as semi-logarithmic or double logarithmic models However, as previously discussed in this report, the author has chosen not to implement this method to resolve the problem.
To create an optimal model with minimal variables that have a negligible impact, while avoiding multicollinearity and heteroskedasticity issues, it is advisable to exclude Siskm variables and apply robust regression techniques to address the heteroskedasticity problem.
Using “reg Q USDi GDPi Dis, robust”” command in STATA we have the final result:
And the newly established model is:
The R-squared index shows a decrease of over 1%, indicating that the new regression model does not significantly differ in its ability to explain the percentage of total campus crime However, the notable changes in the coefficients and standard errors suggest that this updated model should be regarded as a recommendation by the authors.
A statistical error, also known as a disturbance, represents the difference between an observed value and its expected value derived from the entire population For instance, if the average height of 21-year-old men is 1.75 meters, and a randomly selected man measures 1.80 meters, the statistical error is 0.05 meters; conversely, if he is 1.70 meters tall, the error is −0.05 meters Since the expected value reflects the mean of the whole population and is generally unobservable, the statistical error itself cannot be directly observed.
The error of an observed value represents the deviation from the true, unobservable value of a quantity, such as a population mean Conversely, the residual refers to the difference between the observed value and the predicted value in a statistical model.
To examize whether the disturbance follow the distribution or not, we use the Skewneww/Kurtosis Testing
Using command predict resid to predict the residual After that we run “sktest resid”, we have the following results:
The model analysis indicates a significant disturbance distribution, as evidenced by a Prob>Chi2 value of 0.0000 Throughout the modeling period, we identified several potential factors contributing to this error, including climate changes, shifts in consumer preferences, the overall economic climate, and specific trading conditions affecting Vietnam and other wood-importing nations.
3.1 Testing the Result consistency with the theories.
A 5% increase in GDP positively correlates with wood exports, supporting the theory that as the GDP of a wood-importing country rises, the demand for wood also increases.
In this analysis, we reject the null hypothesis (H1) based on a t-value of 2.58, which exceeds the critical value of 1.65922 at a 5% significance level The positive coefficient of GDP supports the initial theory that an increase in the population of wood-importing countries correlates with an increase in wood exports.
Hypothesis { H H 0 1 : : β β 4 4 ≤ > 0 0 T= -6.38 < t0.05(107)= 1.65922 There is not enough evidence to reject H 0
At a significant level of 5%, the negative coefficient of distance indicates that as the distance between wood-exporting countries and Vietnam increases, the volume of wood exports to Vietnam tends to decrease.
Hypothesis { H H 0 1 : : β β 4 4 ≤ > 0 0 T= 1.37 < t0.05(107)= 1.65922 There is not enough evidence to reject H 0
DO-FILE
1 Statistical description sum Q GDPi POPi Dis Si
2 Correlation matrix corr Q GDPi POPi Dis Si
3 Regression model reg Q GDPi POPi Dis Si
6 Test for Heterokadasticity imtest, white
7 Robust test reg Q USDi GDPi Dis Si, robust
8 Test for normal distribution of the disturbance predict res, residual