Tiểu luận the factors affecting the infusion of foreign direct investment capital into a locality vietnam

Among the factors that led to this success, foreign direct investment FDI has played a crucial role, providing Vietnam’s economy with its relatively scarce factor, capital, and represent

Table of Contents

Chapter 1: INTRODUCTION 2

Chapter 2: THEORETICAL BASIS 3

Chapter 3: DATA COLLECTION 5

Chapter 4: EMPIRICAL MODEL AND HYPOTHESIS TESTS 7

Chapter 5: CONCLUSION 14

Chapter 1: INTRODUCTION

Since the introduction of doi moi (renovation) economic reforms in 1986,

Vietnam’s economy has been among the fastest growing economies in the region Its economic structure reflected an increasing share of industry and services while the share of agriculture declined Vietnam has been successful in poverty reduction strategies and has been able to ensure rapid growth with relative equity Among the factors that led to this success, foreign direct investment (FDI) has played a crucial role, providing Vietnam’s economy with its relatively scarce factor, capital, and representing an extremely important instrument for integration in the world economy, especially at the regional level

However, FDI infusion in different localities is not identical Therefore, this paper attempts to identify the main factors that help with the attraction of foreign direct investment capital in a locality of Vietnam The survey results that there are some factors that have been evaluated as much more important; while, others are considered

to be relatively less important in the current context of Vietnam

With purpose to determine the decisive factors that influence the selection of

investment locations in Vietnam, our group chooses the topic: “THE FACTORS AFFECTING THE INFUSION OF FOREIGN DIRECT INVESTMENT CAPITAL INTO A LOCALITY IN VIETNAM”.

Chapter 2: THEORETICAL BASIS

According to the eclectic theory developed by professor Dunning and Mr Nguyen Manh Toan’s report from Da Nang Economics University, location advantages of different countries are the key factors to determining who will become host countries for the activities of transnational corporations

The specific advantages of each country in general and each locality of it in particular can be divided into three categories:

 The infrastructure benefits consist of export processing zones ( EPZs ) as well as industrial zones

 Political advantages: common and specific government policies that affect FDI flows

 Resources advantages includes geography, number of labors After considering the factors as above, we come to decision on researching these ones:

1 Industrial zone (the number of industrial and export processing zones):

this belongs to technical infrastructure and seems to be one of the most important factors to attract FDI to any area in Viet Nam Therefore, the

expectation of Industrial zone is positive (+)

2 School (the number of all types of schools in one area): This kind of

social infrastructure is one criterion that encourages foreign companies investing in a locality More schools will reflect higher level of labor

force The expectation School is positive (+)

3 Policy (measured based on level): open and flexible policies of the local

authority will motivate foreign firms to establish their new business or invest money in this locality There are 3 increasing levels of good policy :

 0: normal

 1: good

 2: very good The coefficient is expected to have positive (+) sign

4 Density (people/km2): Foreign companies when decide to invest in one area also want to exploit a great number of young labor forces with low

salary Therefore, the expectation density is positive (+).

5 Finally, Region: Different kinds of topography may create advantages or

disadvantages for economy development in each locality and this affect the amount of FDI as well This Dummy variable includes:

 0: mountain area and midland

 1: coast

 2: Delta Its expectation sign is positive (+)

Therefore, the model proposed is:

FDI = INDUSTRIAL ZONE + SCHOOL + POLICY

+ DENSITY + REGION

Chapter 3: DATA COLLECTION

3.1 Source of survey:

The data is collected from some websites of General Statistic Office as well as Industrial Zones in Vietnam

3.2 Scope of survey:

My group collected the data from 45 provinces in Vietnam randomly, after that we

classified them into 5 categories: population density, the number of industrial

zones, school, policy, and region

3.3 Data table:

Number Province FDI

(million $) Region Populaiton density School Industrial zone Policy

2 Bà Rịa - Vũng Tàu

19 Hà Tĩnh 8371.4 2 604 540 1 1

Chapter 4: EMPIRICAL MODEL AND HYPOTHESIS TESTS

1) Model 1 FDI = INDUSTRIAL ZONE + SCHOOL + POLICY +

DENSITY

+ REGION

We have the result from Gretl software:

Model 1: OLS, using observations 1-45

Dependent variable: FDI

const -3023.01 862.137 -3.5064 0.00116 ***

INDUSTRIAL ZONE 757.328 127.255 5.9513 <0.00001 ***

SCHOOL 4.47475 1.46112 3.0626 0.00397 ***

POLICY 2778.14 914.435 3.0381 0.00423 ***

DENSITY 2.64933 0.810737 3.2678 0.00227 ***

REGION -879.838 734.027 -1.1986 0.23790

Mean dependent var 4171.496 S.D dependent var 6966.936 Sum squared resid 1.88e+08 S.E of regression 2197.309 R-squared 0.911832 Adjusted R-squared 0.900528 F(5, 39) 80.66754 P-value(F) 1.69e-19 Log-likelihood -406.9070 Akaike criterion 825.8139 Schwarz criterion 836.6539 Hannan-Quinn 829.8550

The estimated model 1 is:

FDI = -3023.01 + 757.328 INDUSTRIAL ZONE + 4.47475 SCHOOL

+ 2778.14 POLICY + 2.64933 DENSITY- 879.838 REGION

COMMENT :

According to the R-squared of the model, 91.18 % of the total variation of FDI

amount in an area in Viet Nam is explained by the joined association between FDI

and the number of industrial zones, schools, level of policy, density and type of

region

It can be seen from the result that coefficients going with all variables have

positive signs as our initial expectation, except for only variable REGION, which is negative So, types of topography may not have influence on the amount of FDI

infusion into this region We will do T-test to show you more exact conclusion.

 Firstly, we start testing with variable

- Null hypothesis : H0 : = 0

- Alternative hypothesis : H1 : 0 From the table, (t = 5.9513) > (tcrit 2.02) (at 5% significance level)

Therefore, we reject H0, which means variable INDUSTRIAL ZONE has

significant explanation to the model

We do the same with variables: SCHOOL, POLICY, DENSITY and we come to

conclude that all of these independent variables are significantly explaining

the amount of FDI to Vietnamese locality at 5% level

 Similarity, with variable REGION:

( = 1.1986) < ( tcrit 2.04) (at 5% significance

level)

Therefore, we do not have enough evidence to reject H0: = 0, we can remove variable REGION

This can be explained as following facts: When the foreign organizations intend

to invest their capital in one locality, they initially carried out surveys on some important aspects of this area and they may be deeply impressed by local infrastructure, the labor force and natural resources If they find out one locality providing them with all of these things, they will almost invest money in this area without caring much about what kind of topography this is They believe that whether the land is coast, midland or delta, they can cope with it and take advantages of it Thus, in this model, independent variable REGION is statistically insignificant in explaining the variation of FDI

Further modifications must be done to the original model in order fix the problem.

2) Model 2 FDI = INDUSTRIAL ZONE + SCHOOL + POLICY +

DENSITY

We have the result from Gretl software:

Model 2: OLS, using observations 1-45

Dependent variable: FDI

const -3669.4 676.321 -5.4255 <0.00001 ***

INDUSTRIAL ZONE

785.685 125.718 6.2496 <0.00001 ***

SCHOOL 4.49338 1.46899 3.0588 0.00395 ***

POLICY 2043.06 681.999 2.9957 0.00468 ***

DENSITY 2.53454 0.809444 3.1312 0.00325 ***

Mean dependent var 4171.496 S.D dependent var 6966.936 Sum squared resid 1.95e+08 S.E of regression 2209.273 R-squared 0.908584 Adjusted R-squared 0.899442 F(4, 40) 99.39003 P-value(F) 3.18e-20 Log-likelihood -407.7210 Akaike criterion 825.4419 Schwarz criterion 834.4752 Hannan-Quinn 828.8094

The estimated model 2 is:

FDI = -3669.4 + 785.685 INDUSTRIAL ZONE + 4.49338 SCHOOL

+ 2043.06 POLICY + 2.53454 DENSITY

COMMENT

According to the R-squared of the model, 90.85 % of the total variation of FDI

amount in an area in Viet Nam is explained by the joined association between FDI

and the number of industrial zones, schools, level of policy and density

The four independent variables all have positive signs as expected Moreover,

their P-value ( P-value < 0.00001; P-value = 0.00395; P-value = 0.00468;

P-value = 0.00325) < 0.05 (at 5% significance level )

The P-value indicates that the four independent variables are statistically significant

at the moment of explaining FDI amount

Now, we use F-test to examine whether the equation has a significant overall

fit or not

- Null hypothesis : H0 : = = = =0

- Alternative hypothesis : H1 : 2 + 2 + 2 + 2 0

We have Fob= 99.39003 > Fcrit 2.69

Reject H0, or the equation does indeed have a significant overall fit

2.1- Multicollinearity testing:

In model 2, we can calculate VIF = = = 10.94 > 10

multicollinearity may happen

However, t-ratio is quite high and there is no correlation among independent

Variables, because:

Correlation coefficients, using the observations 1 - 45

5% critical value (two-tailed) = 0.2940 for n = 45

INDUSTRIAL ZONE

SCHOOL POLICY DENSIT

Y 1.0000 0.2306 0.6828 0.6979 INDUSTRIAL

ZONE 1.0000 0.4689 0.3808 SCHOOL

1.0000 0.4638 POLICY

1.0000 DENSITY

Or

Minimum possible value = 1.0

Values > 10.0 may indicate a collinearity problem

INDUSTRIAL ZONE 3.221

SCHOOL 1.505

POLICY 2.406

DENSITY 2.246

The correlation matrix indicates that there is no collinearity between variables

because pair correlation coefficient between independent variables is quite low

( smaller than 10.0)

(The specific result is shown in the Appendix, page 16 )

2.2- Normality of residual testing

After taking test statistic for normality, we have the result :

Test for null hypothesis of normal distribution:

Chi-square(2) = 1.285 with p-value 0.52593

P-value is equal to 0.5259 which is greater than 0.05

Alternatively, we can test as following:

- Null hypothesis : H0 : U has normal distribution

- Alternative hypothesis : H1 : H0 is wrong Based on the result above, JB= = 1.128, whereas = 9.49 ( = 5%)

< we do not have enough evidence to reject H0

This indicates that the disturbance term has normal distribution and the model 2 is right with the assumptions of the Ordinary Least Squares estimator

(The specific result and graph is shown in the Appendix, page 17 )

2.3- Heteroscedasticity testing

We detect heteroscedasticity by using White test, and the result is:

Unadjusted R-squared = 0.489011 Test statistic: TR^2 = 22.005491,

with p-value = P(Chi-square(14) > 22.005491) = 0.078502

P-value is equal to 0.078502 and greater than 0.05

On the other hand, we can test as following:

- Null hypothesis : H0 : Var(Ui) = for all i

- Alternative hypothesis : H1 : H0 is wrong Based on the result above, nR2 = = 22.00549, whereas, = (17)= 27.59 ( = 5%)

< we do not have enough evidence to reject H0

This indicates that disturbance term is homoscedastic and the model 2 is right with the assumptions of the Ordinary Least Squares estimator

(The specific result is shown in the Appendix, page 18 )

2.4- Autocorrelation testing

Although the structure of our economic data is cross-sectional data, we try testing our model with autocorrelation as Time-series data Then we get the result from Durbin-Watson test:

Durbin-Watson statistic = 1.73645 p-value = 0.171644

The P-value here is more than 0.05, this means our model does not have Autocorrelation

Besides, we also have another table to prove that our model does not have Autocorrelation

(The specific result is shown in the Appendix, page 19 )

Chapter 5: CONCLUSION

After conducting some tests, our group could find the most optimal model It is

FDI = -3669.4 + 785.685 INDUSTRIAL ZONE + 4.49338 SCHOOL

+ 2043.06 POLICY + 2.53454 DENSITY

5.2 Practical significance:

One of the striking developments in Vietnam in recent years has been the large external capital inflow mostly in the form of foreign direct investment (FDI) The effect of FDI has been largely positive on Vietnam’s domestic economy, providing an engine of economic growth by increasing productive capacity and enhancing productivity Foreign – invested operations now contribute to nearly 10% of Vietnam’s GDP, more than 30% of gross capital formation Therefore perception of factors which influence FDI infusion in Vietnam, nowadays, is getting more and more important

5.2 Model application:

Using the OLS on this model has been very useful to determine that FDI in Vietnam

is greatly affected by the number of industrial zones, schools, government policies and population density The overall outcome of the regression shows the robust results For the most of regressions run for this model the results show that the independent variables are important factors that influence the FDI attraction in Vietnam This regression indicates

Therefore our group suggests some ideas to increase FDI inflows into Vietnam:

- Firstly, Vietnam must complete infrastructures by building more factories, industrial zones, improving transportation in order to make distribution faster and more convenient

- Secondly, Vietnam government also needs to encourage foreign investor factors, issuing the open – policies, tries to create an equal business environment

- Last but not least, we should concern more about quantity as well as quality of work force To make sure the good work source, whole society must focus and spend money

on building schools, universities, replacing old facilities by modern ones, advance teacher’s capacity…

Multicollinearity

Normality of residual testing