Tải bản đầy đủ

Structural transformation and economic growth of asia developing countries and vietnam

UNIVERSITY OF ECONOMICS
HO CHI MINH CITY
VIETNAM

INSTITUTE OF SOCIAL STUDIES
THE HAGUE
THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

STRUCTURAL TRANSFORMATION AND
ECONOMIC GROWTH OF ASIAN
DEVELOPING COUNTRIES AND VIETNAM

Summary version
By

TRAN THIEN TAI

Academic Supervisor:

Dr. TRAN TIEN KHAI

HO CHI MINH CITY, NOVEMBER 2012

1


ABSTRACT

This paper investigates the structural transformation and growth of some
developing Asian countries and Vietnam, using data extracted from World
Development Indicator and Global Finance Development of World Bank from 1985
to 2010. The paper uses polynomial model regression and description statistics
method. Findings from the paper includes: (1) except Korea and Malaysia, others
Asian developing countries are all in the first phase of structural transformation.
Agriculture sector trends to decrease once GDP per capita increases. Industry sector
trends to increase once GDP per capita increases. Service sector increases once
GDP per capita increases; (2) the threshold of structural transformation from the
first phase to the second phase is when GDP per capita equals US$ 6,600 per
person. At that level, sectoral share of agriculture, industry, and services reach 7%,
45% and 48% respectively; (3) Asian developing countries including Vietnam are
not all followed the same process and are not homogeneity of structural
transformation; (4) compared to Malaysia, Thailand and the Philippines, the share
of agriculture in GDP of Vietnam is still high and is the highest in the four
countries. The share of services in GDP of Vietnam is always the lowest in the four
studied countries; (5) the rate of labor distribution in the agricultural sector of
Vietnam is high compared to Malaysia, Thailand, and the Philippines and in the
opposite direction, the rate of labor in services of Vietnam is low compared to
Malaysia, Thailand, and the Philippines; (6) labor productivity in all three sectors of
Vietnam are lower than Malaysia, Thailand, and the Philippines but the most
inefficient is agricultural sector, followed by the service and industrial.

Key Words: structural transformation, GDP per capita, growth, Asian developing
countries, Vietnam.

2


TABLE OF CONTENT


CHAPTER 1: INTRODUCTION ............................................................................... 4
CHAPTER 2: LITERATURE REVIEW .................................................................... 6
2.1

Theoretical review ...................................................................................... 6

2.2

Empirical studies ........................................................................................ 7

2.3

Conceptual framework ............................................................................... 9

CHAPTER 3: RESEARCH METHOLODOGY ...................................................... 11
3.1

Data .......................................................................................................... 11

3.2

Research methodology ............................................................................. 11

CHAPTER 4: EMPIRICAL ANALYSIS OF STRUCTRUAL
TRANSFORMATION AND GROWTH.................................................................. 13
4.1

Overview of economic growth of Asian developing countries in period

1985 -2010 ............................................................................................................. 13
4.2

Experimental study result of structural transformation Asian developing

countries during 1985-2010................................................................................... 16
4.2.1 Result of statistics descriptive model ................................................................. 16
4.2.2 Result of economestric model ............................................................................ 21
4.2.3 Structural transformation and labor productivity of Vietnam and acomparision
with Malaysia, Thailand and the Philippines .............................................................. 27
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS ............................ 34
5.1 Conclusions ...................................................................................................... 34
5.2 Recommendations ............................................................................................ 34
REFERENCES .......................................................................................................... 36

3


CHAPTER 1: INTRODUCTION
Some empirical studies show structural transformation process is
accompanied with economic growth of developed countries. By history record,
Kuznets (1971) in Economic of Nations emphasizes that there are six characteristics
that every developed country manifested in the process of economic growth. One of
them is the high rate of structural transformation of the economy. Chenery (1979) in
Structural Change and Development Policy examines the pattern of development of
some developing countries after World War II period. The empirical study identifies
several characteristic features of development process. One of them is the shift
away from agricultural to industrial production.
Asian developing countries, including Vietnam, are under developing process.
Therefore, they maintain sustainable growth in the last two decades and played a
key role in economic growth of the world. It is useful to analyze what is the
structural transformation process of Asian developing countries including Vietnam?
My paper tries to achieve three main objectives: (1) to analyze structural
transformation process1 of some Asian developing countries, including China,
India, Indonesia, Korea, Malaysia, Nepal, the Philippines, Sri Lanka, Thailand and
Vietnam, during 1985-2010; (2) to analyze labor productivity of between Vietnam
and Malaysia, Thailand and the Philippines; (3) to implicate ways to improve
structural transformation process of Vietnam. Therefore, the main questions of this
paper research are: (1) how is the structural transformation process of Asian
developing countries? (2) is the structural transformation process of Asian
developing countries homogenous? (3) what are the differences of structural
1

Structural transformation process is transformation process between sectors in an economy such
as the transformation between agriculture, industrial and service sector through time or through
development (GDP or GDP per capita). Agriculture sector covers forestry, fishing, hunting and
agriculture as a whole; Industrial sector comprise mining, quarrying, manufacturing, construction,
electricity, gas, water; Service sector includes all service activities, such as transportation, logistics,
communication, whole sale, retail, banking, insurance, real estate, public administration, defense
and others services

4


transformation process and labour productivity between Vietnam and Malaysia,
Thailand and the Philippines? These questions will be answered upon the analysis
in chapter four.
The paper is continued with following chapters. Chapter two recalls the
literature review including the theories and empirical studies of structural
transformation in the world and Vietnam. Chapter three describes the dataset and
research methodology. Chapter four analyzes the structural transformation process
of Asian developing countries and the comparison of structural transformation and
labor productivity of Vietnam versus Malaysia, Thailand and the Philippines. Base
on the main findings identified in chapter four, chapter five will come out with the
main conclusions, policy implications and limitations of this research.

5


CHAPTER 2: LITERATURE REVIEW

2.1

Theoretical review
According to Begg et al (1995), Gross Domestic Product (GDP) can be

measured by the formulation:
i

gdpi   vaij

(1)

j 1

Where:
gdpi is GDP of a country in year i

vaij is value added of sector j in year i

j includes three sectors of an economy: agriculture, industry and service.
Solow (1962) uses the Cobb-Douglas production function to form up Solow
growth model
q=Akα

(2)

A is multifactor of productivity or technology progress of an economy.
k is capital per capita of an economy.
q is output per capita of an economy.
Equation (2) explains output per capita will be increased significantly once
productivity, efficiency or technology change happens to the economy. We all
know that a market economy tends to allocate resources from less efficient areas to
more efficient areas. Therefore, this model will support this research of structural
transformation in the following sections of this chapter.
Lewis (1955) develops the two-sector labour surplus model in 1955. In this
model, the underdeveloped economy consists of two sectors which are traditional
and modern sectors. Traditional sector has a surplus of labour while a limited
resource of land. Its marginal product of labour (MPL) tends to diminish until MPL
equal to zero (MPL=0). The proportion of surplus labour in traditional sector will be
transferred to the modern sector and makes the modern sector’s output grown. The

6


labour transfer process and employment expansion in modern sector continue
happening until all of surplus labour in traditional sector is absorbed. The twosector labour surplus model provides a basic theory of structural transformation.
The structural transformation of the economy can take place with the growth of the
modern sector and modern industry (industrial and service sector) without reducing
agricultural output.
According to Perkins et al (2006), the Engel’s law was developed by Ernst
Engel in the nineteenth century. The law states that when household income
increases, the proportion of income spent on food decreases. This is one reason to
explain the decline of agriculture’s share in total production when the GDP per
capita increases. Another reason comes from the productivity gains in agriculture
due to technological change which promotes the process of liberalization of the
labour force and allow them joining in non-agricultural sector such as industry and
services.
Kuznets (1971) finds out that developed countries are following up the same
process of structural transformation. He distinguishes structural transformation into
two different phases. The first phase is in the beginning of development process, in
which an economy allocates most of its resources to agriculture sector. As the
economy continues to develop, resources are then re-allocated from agriculture to
industrial and service sector. In the second phase, resources are re-allocated from
both agriculture and industrial to service sector2.
2.2

Empirical studies
Bah (2008) analyzes structural transformation of developed countries

including nine countries, such as Australia, Canada, France, Germany, Italy, Japan,
Sweden, United of Kingdom, and the United States, during the period 1870 -2000.
Bah finds that: (1) developed countries follow a homogeneity process of structural
transformation; (2) the structural transformation of developed countries is well

2

This empirical study will be referred in chapter four and chapter five of this research.

7


suited to the one Simon Kuznets mentioned in theoretical review. Agriculture
declines in both first and second phases of development. Industry increases in the
first phase of development and decreases in second phase of development. Service
sector always increases in the first phase and second phase of development; (3) the
threshold between first phase and second phase of development is when GDP per
capita reach at 8,100 US$ per person; (4) all developed countries are in the second
phase of development.
Bah (2009) explores that beside the thing that structural transformation play a
positive role in economic growth, Total Factor Productivity (TFP) of each sector
also play an important role in economic growth. He uses panel data on sector
employment share and GDP per capita of the US, represent for developed country,
and Korea, Cameroon, Brazil, represent for developing countries, from 1950 to
2000, to analyze sectoral productivity of developed and developing countries. He
finds out that relative to the US, developing countries are least productive in
agriculture, then followed by services and manufacturing.
Hoang Kieu Trang (1998) analyzes structural change of Vietnam during 19801997. The paper reveals that (1) the growth rate of non-agricultural sector of
Vietnam increases higher than GDP growth rate; (2) structural change, including the
declining of agriculture, increasing of industry and services, provides positive
impact to economic growth.
Dekle & Vandenbroucke (2006) investigate how structural transformation
impact to economic growth of China from 1978 to 2003. They explore three sectors
in China’s economy: agriculture, private non-agriculture, and public (government)
non-agriculture sectors by using employment by sector and GDP per sector data.
The paper discloses that there are three main sources of China’s growth from 19782003: (1) high productivity in private non-agriculture sector; (2) reallocation of
labor from agriculture sector to non-agriculture sector, (3) reallocation of labor from
public non-agriculture sector to private non-agriculture sector.

8


Duarte & Restuccia (2010) examine the role of sectoral labor productivity and
the reallocation of labor across sectors to explain the process of structural
transformation. The authors find that (1) sectoral labor productivity differences
across countries are large, both at a point in time and over time. In particular, labor
productivity differences between developed and developing countries are large in
agriculture and services and smaller in industry; (2) over time, productivity gaps
between developed and developing countries have been substantially narrowed
down in agriculture and industry but not really as much in service sector.
2.3

Conceptual framework
From theoretical review and empirical studies section in this chapter,

structural transformation between sectors happens through out three main factors as
figure 1 below. The first factor includes technological change, investment and
capital (both of physical and human capital) accumulation. These three components
will affect significantly the productivity of each sector of an economy. The second
factor is the consequence of the first one. By absorbing technological change,
investment and capital accumulation, the sectoral productivity will increase and
grow continuously. The differences in productivity growth of each sector and the
differences in productivity level of each sector, such as labor surplus and low
productivity in agricultural sector, make the structural transformation of a country
happen differently. The structural transformation tends to occur from low
productivity sectors to high productivity sectors. The third factor mentions about the
structural transformation happens in the same time of resources reallocation (labor
resource and other resources) process. The resources will be allocated from lower
efficient to higher efficient areas. In the study of this thesis, I just mention three
factors as the sources and causes which impact on the structural transformation, not
a deeply analysis (qualitative and quantitative) the impact of these factors to the
structural transformation and economic growth. Consequently, the structural
transformation of a country will make a country’s development and growth. This
process will be continuously happened to push an economy continuously develop

9


and grow. The below figure 1 describes the interaction between structural
transformation and economic growth. This paper will analyze structural
transformation process and GDP per capita growth of Asian developing countries
basing on what Kuznets (1971) has realized for developed countries.

Figure 1: Conceptual framework – structural transformation and growth
Source: author’s creation base on theoretical review and empirical studies

10


CHAPTER 3: RESEARCH METHOLODOGY
3.1

Data
The data in this research is mainly collected from World Development

Indicators and Global Development Finance of the World Bank. A time series of 26
years from 1985– 2010 of 10 Asian developing countries, including Vietnam,
China, India, Indonesia, Korea, Malaysia, the Philippines, Sri Lanka, Nepal and
Thailand are collected to form up a panel data with 260 observations.
3.2

Research methodology
In this paper, I apply both descriptive statistics and econometric methods to

explain the structural transformation process and growth of Asian developing
countries. To determine how the structural transformation process of Asia
developing countries is and whether these Asian developing countries follow up the
same process of structural transformation, I use polynomial functions to indicate the
relationship between sectoral output share such as agriculture, industry and services
and log of GDP per capita of all countries. The polynomial function is fitted from
260 observations in a panel dataset.
For each sector, I estimate by the following equation:

va   i  1 log( gdpit )   2 log( gdpit ) 2   3 log( gdpit )3  ...   (3)
it
it
Where:
va is the sectoral output share of GDP for country i (i=1~10) in period t
it

(t=1985~2010)
 i is fixed affect of country i

 is coefficients of log( gdpit )
gdpit is GDP per capita of country i in period t

 is the error term
it

11


According to Nguyen Trong Hoai (2006), polynomial function reflects the
long-run average trend between dependent and independent variables. Since
structural transformation process will need to be observed and analyzed in long
period time to reflect the long term average trend between sectoral output share and
log of GDP per capita, therefore I select polynomial function to analyze the process.
The degree of polynomial function in equation (3) is determined by the goodness of
fit. Starting from a linear polynomial, the degree of function will be increased one
by one and continuing this process until the change of R-square is less than
0.01.The reason I do it is try to simplify the model in the lowest degree within the
highest possibility of the goodness of fit.

12


CHAPTER 4: EMPIRICAL ANALYSIS OF STRUCTRUAL
TRANSFORMATION AND GROWTH

4.1
Overview of economic growth of Asian developing countries in period
1985 -2010
Figure 2 below show the economic growth in GDP of ten Asian developing
countries.

Average annual GDP growth (%)

12.0
10.0
8.0
6.0
4.0
2.0

Ph
i ll i
pp
in
es

Ne
pa
l

La
nk
a
Sr
i

In
do
ne
si a

Th
ail
an
d

M
al
ay
s ia

Ko
re
a

In
di
a

Vi
et
na
m

Ch
in
a

0.0

Figure 2: Average GDP growth of Asian developing counties, 1985-2010 (%)
Source: Author draw base on data from World Bank, 2012

Based on the economic growth rates, the countries can be divided into three
groups. The first is very rapid GDP growth group above 7% annually during 19852010 which only China achieved average annual growth 10.0%. Follow to China is
the second group achieved good economic growth from 5% to 7%. Leading this
group is Vietnam maintained average annual economic growth 6.8%. Following to
Vietnam is India, Malaysia, Thailand and Indonesia which annual growth rates
respective 6.4%, 6.1%, 5.9%, 5.6% and 5.2%. The third group maintained moderate
annual growth from 3% to less than 5% including Sri Lanka, Nepal and the
Philippines have respective annual growth 4.9%, 4.5% and 3.7%.

13


Corresponding to GDP growth, figure 3, table 1 below show GDP per capita
and GDP per capita growth rates of these ten Asian developing countries. China
maintained highest average annual GDP per capita growth at 9.0% during 19852010. Following to China are Korea and Vietnam maintained average its average
annual growth rate at 5.3% and 5.1%. Average annual GDP per capita growth rate
of India, Thailand, Sri Lanka, Indonesia, Malaysia, Nepal and The Philippines
during 1985-2010 are respective 4.5%, 4.4%, 3.8%, 3.7%, 3.4%, 2.2% and 1.3%

GDP per capita (Current $)

25,000
1985

20,000

2010

15,000
10,000
5,000
0
a
re
Ko

a
si
ay
l
a
M

nd
il a
a
Th

C

na
hi

a
si
ne
o
d
In

Sr

ka
an
L
i

i
Ph

s
ne
pi
ill p

Figure 3: GDP per capita 1985 and 2010 (current, US$)
Source: Author draw base on data from World Bank, 2012

14

a
di
In

m
na
et
i
V

N

al
ep


Table 1: Average GDP per capita growth 1985-2010 (%)
Country / Year

1985-1990

1991-2000

2001-2010

1985-2010

China

7.22

9.28

9.83

9.02

Korea

8.12

5.22

3.75

5.32

Vietnam

2.3

5.86

6.05

5.11

India

3.57

3.51

6.08

4.51

Thailand

7.47

3.57

3.42

4.41

Sri Lanka

2.32

4.39

4.09

3.8

Indonesia

4.55

2.91

4

3.71

Malaysia

2.58

4.57

2.66

3.38

Nepal

2.38

2.45

1.76

2.17

The Philippines

0.08

0.59

2.82

1.33

Source: World Bank, 2012

GDP per capita of Asian developing countries
(excluding Malaysia and Korea)
5,000
4,500

3,500
3,000
2,500
2,000
1,500
1,000
500

VNM

CHN

IND

THA

IDN

15

NPL

PAK

PHL

LKA

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

0
1985

GDP per capita (US$)

4,000


GDP per capita of Malaysia and Korea

25,000

GDP per capita (US$)

20,000

15,000

10,000

5,000

KOR

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

0

MYS

Figure 4: GDP per capita of Asian developing countries 1985-2010
Source: Author draw base on data from World Bank, 2012

4.2
Experimental study result of structural transformation Asian
developing countries during 1985-2010
4.2.1 Result of statistics descriptive model
In this section, I will analyze the structural transformation process of Asian
developing countries through time by descriptive statistics method. The structural
transformation process of ten Asian developing countries indictaed as figure 5
below

16


17

Agr

Ind

Ser

2010

2009

2008

2007

2006

2005

2004

Ind

2003

30

20

10

0

1985

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

2006
2007
2008
2009
2010

2006
2007
2008
2009
2010

2005

40

2005

50

2003

60

2004

M alaysia

2004

Ser

2003

2002

2001

2000

Ind

2002

2001

2000

Agr

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

% Sector share

Agr

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

% Sector share

% Sector share

Kore a

60
50

40

30

20

10

0

Ser

Thailand

60

50

40

30

20

10

0


Agr

18

Ind

1987

Ind

Ser

2008
2009
2010

2010

2010

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

2009

0

2009

10

2008

20

2007

30

2008

40

2006

50

2007

60

2007

Sri Lanka

2006

Ser

2005

0

2006

10

2004

20

2005

30

2005

40

2003

50

2004

60

2004

India

2003

Ser

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

Ind

2002

2001

Agr

2000

1985
1986

% Sector share

Agr

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

% Sector share

% Sector share

China

60
50
40

30

20

10

0


1985

Agr

19

Ind

2001

Agr
Ind

1987

60

50

40

30

20

10

0

Ser

2010

2009

2007
2008

2006

2010

2009

2008

2007

2006

2005

2004

Vie tnam

2005

Ser

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

Ind

2004

2002
2003

1985
1986

% Sector share

Agr

2000

1998
1999

1997

1996

1995

1993
1994

1992

1991

1989
1990

1988

1987

1986

% Sector share

1985

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

% Sector share

The Philippine s

60
50
40

30

20

10

0

Ser

Indone sia

60

50

40

30

20

10

0


Ne pal

60

% Sector share

50
40

30
20
10

Agr

2010

2008

2009

2007

2005

2006

2003

2004

2002

2000

Ind

2001

1998

1999

1997

1995

1996

1994

1992

1993

1990

1991

1989

1987

1988

1985

1986

0

Ser

Figure 5: Structural transformation of Asian developing countries 1985 and
2010 – Time series
Source: Author draw base on data from World Bank, 2012

By summarizing the economic structural transformation of the Asian
developing countries under the perspective of descriptive statistics and time series,
we can see the following points: (1) with different levels of development, in the
studied period 1985-2010, countries are with GDP per capita uneven; (2) because
different levels of development and GDP per capita, the starting point and ending
point of the sectoral share in GDP (agriculture, industry, services) in the economy
are different, referenced to figure 5; (3) with the differences in geographical
locations, natural resources, techniques and technology, and economic operational
policies, the process of economic structural transformation of the countries are in
different directions; (4) although the starting point, ending point, and the speed are
different, but in general the share of agriculture of all countries did go in the
direction consistent with is reduced over time, over the development and over the
growth of GDP per capita; (5) the percentage share of industrial sector in the total
GDP of the country does not go in a certain direction. Most countries have the share
of industry increased over time, however there are some cases decreased over time
such as Korea in 1997-2010, India does not change over time, and the Philippines
tends to decrease over time; (6) similar to industry sector, the service sector did not

20


follow uniform trends in chronological order. The majority of countries have the
share of services increased, but Thailand has the share of services decreased in the
period 1985-2010, Vietnam decreased in the period 1998-2010 and Indonesia did
not change in the period 1985-2010.
4.2.2 Result of economestric model
Recall data and research methodology from chapter three, I use following
polynomial function to indicate structural transformation process for each sector:

va   i  1 log( gdpit )   2 log( gdpit ) 2   3 log( gdpit )3  ...   (3)
it
it
va is the sectoral output share of GDP for country i (i=1~10) in period t
it

(t=1985~2010)
 i is fixed affect of country i

 is coefficients of log( gdpit )
gdpit is GDP per capita of country i in period t



it

is the error term
Starting with agricultural sector, I fit equation (3) by a linear polynomial

regression model ( 2 ,  3 = 0) with a fixed-effect estimator. The result comes out
with R-square equals 0.76. I continue fitting equaltion (5) by a quadratic polynomial
(  3 = 0), then R-square increases to 0.83. I also experimented with higher degrees,
but the changes in R-square did not increase larger than 0.01. Therefore, the
relationship between agriculture output share and log of GDP per capita at best
fitted by quadratic polynomial (  3 = 0). The R-square is 0.83. I do regression
similarly for industry and service sectors. For industry, the best fitted is a third
degree polynomial. The goodness of fit is lower than agriculture; R-square is 0.47.
The relationship between output share of service and log of GDP per capita is also
determined by third degree of polynomial with the R-square of 0.33, lower than

21


both agriculture and industry. Table 2 below shows regression result for Asia
developing countries.
Table 2: Summary Regression Result for Asia Developing Countries
Constant

Log(gdp)

Log(gdp)^2

Log(gdp)^3

R-squared

Agriculture

Industry

Service

112.29 ****

26.78

-91.19 **

(0.000)

(0.396)

(0.023)

-19.6 ****

-7.24

49.28 ***

(0.000)

(0.587)

(0.004)

0.89 ****

2.15

-6.18 ***

(0.000)

(0.245)

(0.009)

-

-0.13

0.28 ***

(0.117)

(0.010)

0.47

0.33

0.83

* Significant at 10%; ** Significant at 5%; *** Significant at 1%; **** Significant at 0.1%

Note: This table reports the fixed effect regression of equaltion (5) of each sector. The data
consist of 260 obersvations including 10 countries with 26 observations per country. The Pvalue is in parentheses.
Source: Author’s calculation using panel data from World Bank, 2012

To estimate fixed effect (  i ) of each country, I use Least Square Dummy
Variables estimator model (LSDV). I call  is average fixed affect of all ten
countries which is identified by regression result in table 2. For example,  =112.29
for agriculture; 26.78 and -91.19 for industry and services respectively. Similarly, I
call  i is fixed effect of each country estimated by LSDV. For each country, I
calculate  i =  i -  and I call it is the fixed effect deviation from the mean of
each country. According to Bah (2008), coefficient distribution such as fixed effect
deviation from the mean of each country will help us to determine the extent of
heterogeneity between countries. Table 3 shows fixed effect deviation from the
mean of each country. The standard deviation of the fixed effect deviation from the

22


mean in agriculture is 3.92, while these numbers are 6.83 and 5.21 in industry and
services respectively.
After regression, I use Stata to draw scatter plots of output share of each sector
with log of GDP per capita to analyze the sectoral transformation process and
contribution of each country. Figure 6 shows the scatter plots of output share of
agriculture versus log of GDP per capita. The graph shows the fitted curve with the
lower and upper bounds which are lined at two standard deviations of the forecast
values.
Table 3: Fixed Effect Deviation from the Mean
Country

Agriculture

Industry

Service

Vietnam

1.35

1.15

-2.50

China

-3.31

10.85

-7.58

India

-0.55

-6.08

6.40

Indonesia

-2.59

7.24

-4.77

Korea

-1.17

0.60

0.52

Malaysia

1.68

4.66

-5.94

Nepal

10.09

-11.22

1.17

The Philippines

-1.56

-2.15

3.61

Sri Lanka

-0.95

-7.34

8.13

Thailand

-3.05

2.31

0.98

Std. Deviation

3.92

6.83

5.21

Note: This table reports the differences of the average fixed effect of all countries with fixed
effect of each country. The average fixed effect of all country is obtained by regression
equation (3) for each sector. The fixed effect of each country is obtained by LSDV
estimation.
Source: Author’s calculation using panel data from World Bank, 2012

23


Figure 6: Scatter chart of agriculture output share and Log of GDP per capita
Source: Author’s calculation and draw from Stata.

In general, the transformation process of agriculture of countries decreased
with the increasing in GDP per capita. Figure 6 shows that when the log of GDP per
capita reached 8.8 or higher, equivalent to the GDP per capita of US$ 6,600 per
person or more, the proportion of agriculture tend not to reduce further, at which the
share of the agricultural sector remained at 7%. In summary, we can see three points
in the agricultural transformation process: (1) when GDP per capita increases, the
agricultural share decrease; (2) when the share of agriculture declined to the level of
GDP per capita US$ 6,600, the economy of Asian developing countries will move
from the first phase to the second phase of structural transformation as Kuznets
mentioned in Chapter 2; (3) at the threshold of converting from the first phase to the
second phase of structural transformation, countries maintain an average share of
7% in agriculture as the level at which countries should maintain to ensure food
security for their countries.

24


Figure 7: Scatter chart of industry output share and Log GDP per capita
Source: Author’s calculation and draw from Stata

Figure 7 shows the scatter plots of sectoral output share of industry versus log
of GDP per capita and it reveals that in the industry we find the following
highlights: (1) the share of industry increases when the log of GDP per capita rose
to 8.8 or US$ 6,600. This share peaks at 45%, and then gradually decrease as GDP
per capita continues to rise. Similar to the agricultural sector, we see that at the
threshold of GDP per capita of US$ 6.600, there is a shift from the first phase to the
second phase of structural transformation; (2) as mentioned in section 4.2.1, the
share of industry of countries does not coincide with each other. We can see that the
share of industry of countries like Nepal is always lower than the fitted curve. China
has the share of industry higher than the fitted curve, and there are many
distribution points out of the upper bound of industry. Nearly identical to China are
Vietnam and Indonesia, there are always points that are allocated above the fitted
curve. At the same time, on the opposite side, with varying degrees, Sri Lanka,
India and the Philippines have the industrial distribution points constant to GDP per
capita, in which the Philippines tend to reduce to GDP per capita.

25


Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Tải bản đầy đủ ngay

×