Urban expansion the effect of economic growth on urbannization in viet nam

SOUK SOPHAL URBAN EXPANSION THE EFFECT OF ECONOMIC GROWTH ON URBANIZATION IN VIETNAM LUẬN VĂN THẠC SĨ KINH TẾ TP... SOUK SOPHAL URBAN EXPANSION THE EFFECT OF ECONOMIC GROWTH ON URBANIZ

BỘ GIÁO DỤC VÀ ĐÀO TẠO TRƯỜNG ĐẠI HỌC KINH TẾ Tp.HCM

………

SOUK SOPHAL

URBAN EXPANSION THE EFFECT OF ECONOMIC GROWTH ON URBANIZATION IN

VIETNAM

LUẬN VĂN THẠC SĨ KINH TẾ

TP Hồ Chí Minh-Năm 2013

BỘ GIÁO DỤC VÀ ĐÀO TẠO TRƯỜNG ĐẠI HỌC KINH TẾ Tp.HCM

………

SOUK SOPHAL

URBAN EXPANSION THE EFFECT OF ECONOMIC GROWTH ON URBANIZATION IN

VIETNAM

Chuyên ngành: Kinh tế phát triển

Mã số: 60310105

LUẬN VĂN THẠC SĨ KINH TẾ

NGƯỜI HƯỚNG DẪN KHOA HỌC: Dr Nguyễn Lưu Bảo Đoan

TP Hồ Chí Minh-Năm 2013

I would also like to express my profound thanks to Dr.Pham Khanh Nam for his help in collecting Vietnam data

Special thanks to Dr.Nguyen Huu Dung, Deputy Dean at the University of Economics HoChiMinh City

I would like to thanks my family and my good friends for always encouraging me to think and supporting me spiritually throughout my life

Finally, I would like to dedicate this work to my lost relatives and I hope that this work makes them proud

HoChiMinh City, December 15, 2013

SOUK SOPHAL

ii

Abstract

Urbanization is a major change taking place globally, especially in Southeast Asia which is one of the world’s fast growing regions in terms of population and urban growth The total population in Southeast Asia accounts for 14.64 per cents of Asia’s population Meanwhile, urbanization is also increasingly taking place in many cities in Southeast Asia Phnom Penh is the capital of Cambodia with an estimated population of 2.3 million by 2013, with the area of 678.46 square kilometers Recently, the development in the economy has impacts on the growth of Phnom Penh Capital The demand for more living spaces has increased so it lead to unplanned urbanization and industrialization in the outer periphery of Phnom Penh Capital, which, has created pressure on the change in land use pattern Therefore, it

is important to study impacts of different factors on urban expansion However, the absence of data for Phnom Penh does not make a regression analysis possible Thus, the study employs available data of cities and provinces in Vietnam, a country which has a similar economic structure and level of development In this paper, the overall goals are to examine the change in population, GDP, agricultural investment and other that lead to the urban spatial growth and, the expansion of urbanized areas Based on the Monocentric model and panel analysis, the result shows that population, agricultural investment, industrialization and GDP are major pushing factors in the processing of urban expansion The analysis demonstrates the overwhelming importance of economic growth in the determination of urban land use This study examines the effect of economic growth on urbanization in Vietnam Based on this information, a case study analysis of Cambodia and Phnom Penh is also conducted

Keywords: economic growth, monocentric model, urban expansion, land use

Table of Contents

ACKNOWLEDGEMENTS i

ABSTRACT ii

LIST OF ABBREVIATION LIST OF TABLES LIST OF FIGURES Chapter 1 : Introduction……… 1

1.1 Research problem……… 1

1.2 Research Objective………2

1.3 Research Question……….3

1.4 Scope and Methodology of Research………3

1.5 Research Structure ………3

Chapter 2: Urbanization in Vietnam……… 5

2.1 The overview of the study area……… 5

2.2 Urbanization in Vietnam……….7

Chapter 3: Literature Review……… 12

3.1 Theoretical Literature Review of Monocentric………12

3.2 The Monocentric model……… 13

3.3 Empirical Study… ……….16

3.4 Analytic Framework………19

Chapter 4: Data and Methodology……… 21

4.1 Data Source……… 21

4.2 Data Description……… 22

4.3 Research Method ……… 23

4.4 Analysis Method……….……… 24

Chapter 5: Result and Discussion………26

5.1 Descriptive Analysis………26

5.2 Empirical Result….……… 33

Chapter 6: A Case Study about Cambodia and Phnom Penh ……… 37

6.1 Overview of urbanization in Cambodia……… 37

6.2 Overview of urbanization in Phnom Penh………40

Chapter 7: Discussion 7.1 Explaination of differences between theory and my analysis… 46

7.2 Limitations of this analysis……… 46

Chapter 8: Conclusion……… 48

REFERENCES

APPENDIX

LIST OF ABBREVIATION

CBD Central Business District

FEM Fixed Effect Model

GDP Gross Domestic Product

GSO General Statistics Office IND Industrialization

SVS Service Sector

REM Random Effect Model

LIST OF TABLES

Table 3.3.1 The proxies and list studies of independent variables……….17

Table 5.1.1 Descriptive Statistics……… 25

Table 5.1.2 Correlation test………26

Table 5.1.3 Multicolinearity test………27

Table 5.1.4 Normality test……….27

Table 5.2.1 Estimation result……….31

Table 5.2.1.1 Heteroskedasticity test………33

LIST OF FIGURES

Figure 2.1.1 Expansion of city boundaries between 1999 and 2009 … 6

Figure 2.1.2 Urbanization rate 1960-2010………8

Figure 3.1 PLUREL concept of peri-urban areas and rural- region……… 12

Figure 3.2 Dimension of regional urban form ………13

Figure 5.1.5 Graph of changes in urbanized land across the country………… ….28

Figure 5.1.6 Graph of explanation for changes of province &cities ……… 30

Figure 6.1 Phnom Penh map by districts……… 39

Figures 6.2 Planning of city development……….……….41

of the development process Rapid urbanization is closely linked to economic development that is contributed with the rising demand of land in manufacturing and service sector in urban and the decline of agricultural land in rural Urbanization occurs as countries shift from rural agricultural activities into urban industrial activities Successful urbanization requires the expansion of the industrial and service sector (Parker, 1996) Similarly, urban expansion mainly depends upon the urban requirement, facilities available and industrialization The development of cities plays

a very important role in economic development in developing countries The major dynamic driving forces of cities in developing countries are economic and social

1

http://2012books.lardbucket.org/books/a-primer-on-social-problems/s17-01-a-brief-history-of-urbanizatio.html

forces (population growth and urbanization) that change the urban system Of course, the problems are different and unique from one to another urban according to the urban growth patterns, economic and social characteristics With high urban population density and potential for economic growth, there are demands more and more land to ,support, yet, there is no unlimited amount of land for growth The increase of demand for goods and services of all kinds is pushing up the price especially the price of land Urban growth is the growth and decline of urban areas when an economic phenomenon is inextricably linked with the process of urbanization Urban growth or city growth is the expansion of towns and cities with respect to the increase of the size of a built-up area Cities offer important opportunities for economic and social development and are they are the focal points for economic growth, innovation, and employment Nevertheless, as cities grow, managing them becomes increasingly complex As a result, it creates pressure on infrastructure, reduction in farmland, and damage to natural environment and changes

in diversity use of land in particular area Particular concerns are the risks to the immediate and surrounding environment, to natural resources, to health conditions, to social cohesion, and to individual rights Since an urban land is very often being developed at the cost of a lot of arable lands that would raise many people, if there is

no enough arable land to supply enough food, it would threat to human being’s surviving It is also known that demography and economy are two driving factors for urban expansion Demands for space by humans are the original force for urban expansion However, the urban expansion may not happen without economic support High economic growth may also promote the demands Therefore, the link between

urban expansion and economic growth is equivalent and very closely

(i) Does GDP growth rate have effect on urban expansion?

(ii) Does population growth have effect on urban expansion?

(iii) Does agricultural investment have effect on urban expansion?

(iv) Does industrialization have effect on urban expansion?

(v) Does service sector have effect on urban expansion?

1.4 Scope and Methodology of Research

The thesis will be examine on 63 cities and provinces in Vietnam except some provinces such as Hoa Binh, Hau Giang and Lai Chau because of missing data, during 2000-2011 period In order to get the goal of this research that can explore the effect of economic growth on urbanization (urban expansion), the STATA 12 software will be apply to analyze the panel dataset Particularly, the specific econometrics consists of Fixed Effects Model (FEM) and Random Effect Model (REM) but in my analysis I only use Fixed Effects Model Several statistics tests are also taken into account in order to find the most suitable model that can explain the best effect of economic growth on urbanization

1.5 Research Structure

This thesis is organized into eight chapters and reference and appendix as follow Chapter 1 (Introduction) draws a general scope of this research Chapter 2 (Urbanization in Vietnam) covers the urbanization underway and some government policies on urbanization Chapter 3 (Literature review) summarizes the theory of Monocentric model and empirical studies of economic growth and urban expansion in cities and provinces in Vietnam Chapter 4 (Data and Research methodology)

illustrates the data source, data description, research method, method for analysis and descriptive analysis Chapter 5 (Empirical Results) shows the estimated result on this research study Chapter 6 (A case study of Cambodia and Phnom Penh urbanization) shortly reviews urban planning, government policy on housing and some policies on urbanization Chapter 7 (Discussion) explains the differences between theory and my analysis and finally the limitation of this analysis Chapter 8 (Conclusion) evaluates the empirical study and recommends some policies

5

CHATER 2: URBANIZATION IN VIETNAM

2.1 The overview of the study area

Vietnam is a developing country located in the centre of the Southeast Asian region It borders the Gulf of Thailand, Gulf of Tonkin, and South China Sea, alongside China, Laos, and Cambodia This S-shaped country has a north-to-south distance of 1,650kilometers and about 50 kilometers wide at the narrowest point2.Vietnam is a development success story Political and economic reforms (Doi Moi) launched in

1986 have transformed Vietnam with per capita income below $ 100 to a lower middle income country within a quarter of a century with per capita income of $1,130 by the end of 2010 Vietnam has already attained five of its ten original Millennium Development Goal targets and it is well on the way to attaining two more by 20153 According to a report of Vietnam, recently, on account of Doi Moi reforms and export-oriented industrialization policies, there is a strong co-movement of urbanization and economic growth Vietnam is in the incipient (initial) stage of urbanization and it is transitioning to an intermediate stage (the second (middle) stage in the process of successive) with rapid urbanization (currently 30% urban with 3.4% urban population growth per year) and an increasing economic transition toward industrial manufacturing as a share of both employment and economic output (World Bank(2011)*)

This Figure 2.1.1 below shows the Expansion of city between 1999 and 2009,

In 1999, the most expansion boundary was HoChiMinh city, class 1 city, Da Nang, class 2 City and Hanoi, class 1city with the land expansion on boundary with 100km circle radius And the rest of provinces and cities are in class 2 city, class 3city and 4 city

2 www.vietnamemb.se/index.php?option=com_content&view=article

&id=68&itemid=62

3 http://www.worldbank.org/en/country/vietnam/overview

Chapter 1: Vietnam’s Evolving Urban System (Figure 1.6)

Figure 2.1.1 shows that:

10 years after, HoChiMinh city and Hanoi have become special cities and Da Nang has become class 1 city And some of provinces and cities become class 1 city and the rest become class 2 city and class 3 city and about 15 provinces are in class 4 city In Vietnam, there are 63 cities and provinces And five centrally controlled municipalities such as Can tho, Da Nang, Hai Phong , 1st- class and HaNoi , HoChiMinh city are special class The rest is provincial municipalities

7

2.2 Urbanization in Vietnam

Vietnam is one of the potential countries in Asia, from 1954 to the present, its administrative boundary has been changed under the approval from the central government As Vietnam enters a crucial period of urbanization corresponding to its present stage of economic development, the developing system of cities has a strong emphasis by the government of Vietnam The history has been viewed as an effective tool in controlling the city’s size and the encroachment of urban areas onto agricultural lands With the progress of urbanization since Doi Moi policies in the late 1980s, the loss of agricultural land to urban use is increasing conflicts at the urban fringe (It is a zone along the edges of the built-up area, which comprises a scattered pattern of lower density settlement areas, urban concentrations around transport hubs, together with large green open spaces, such as urban woodlands, farmland, golf courses and nature reserves.) of many cities Meanwhile, Vietnam’s urbanization was managed by extensive consultation with various stakeholders, including officials from national and local government, private sector groups and active international and bilateral organizations in development assistance in the urban sector in Vietnam (World Bank, 2011*) Vietnam has seen its urban areas developing since 1990, when there were only

500 urban areas with an urbanization rate from 17-18% The urban areas jumped to

649 in 2000 and 656 in 2003 And the nation will reach an urbanization rate around 38% in 2015 given the national master plan on urban development and there will be

870 urban areas, according to, a report of the Ministry of Construction Urbanization play an important role in the development process in Vietnam According to Mr Nguyen Sinh Hung, Deputy Prime Minister, “Vietnam will have one chance to get urbanization right If we fail at urbanization, we will fail at industrialization and modernization.” Even though, urbanization does not guarantee economic growth and modernization, an integral part of Vietnam’s transition from a low to middle income country may depend on a well underway of transition- from a large rural to an urban economy Therefore, for the first urbanizing nearly all countries achieved high income and strong economic growth and at least 50% urbanized countries were middle income countries before urbanization Vietnam expects to reach that point by 20254

4

www.worldbank.org/en/news/press-release/2012/04/05/vietnam-needs-to-seize-its-urbanization-opportunities-to-become-a-middle-income-country

In this figure 2.2.1, it shows the urbanization rate of 5 counties such as South Korea, Indonesia, China, India and Vietnam that have potential for economic growth Vietnam urbanization process is continuing going up around 30% from 2000 to 2010 and similarly to the urbanization rate of India

Figure 2.2.1 Urbanization rate 1960-2010

Additionally, rapid urbanization always makes two cases occur The first is the potential for increasing congestion costs, regional inequalities, increasing urban poverty, urban pollution and land and housing prices And the second is the increasing traffic congestion, increasingly polluted water ways and high urban land prices At the same time, Vietnam must be ready to employ urbanization as an instrument to sustain economic growth by ensuring the economic competitiveness of key economic regions, ensuring the social and environmental sustainability of cities making them desirable places to live and work for all segments of society and increasing economic productivity through accelerated technological advances and a better trained, educated and mobile workforce (World Bank Vietnam No 1, 2012) Therefore, “well managed urbanization can serve to support Vietnam’s economic growth and its Socio-Economic Development Strategy objective” said Dean A Cira, Lead Urban Specialist for World

CHAPTER 3: LITERATURE REVIEW

3.1 Theoretical Literature Review of Monocentric city

The traditional models of urban economics view cities around a single point, defined

as “mono”-center, or it is called central business district (CBD) The monocentric city

as conceptualized by Alonso (1964) is the first formal model of urban spatial structure

It directly showed that cities had a unique centre, often termed the Central Business District (CBD) Wheaton (1974) had a crucial role in identifying the spatial dimensions of urban and regional socio-economic activity grounded in urban economic theory and the fundamental parameters underlying spatial growth of cities were generalized by a function of population (or population density), household income, agricultural land rent, and commuting costs In these traditional models, the cities are supposed to take place in a point representing the central business district (CBD), and all workers living in the surrounding area are supposed to commute to the CBD McMillen (2006) indicated that the net effect of time costs of commuting and income on city size was ambiguous, because an increase in income enabled urban residents who prefer to live farther from the CBD to do so, as well as to increase their opportunity cost of commuting by selecting residential locations closer to the CBD, leading to a smaller city size, not a larger one The monocentric city has been the model most widely used to analyze the spatial organization of cities (Bertaud 2003) Based on the hypothesis of monocentric city, it is showed empirically that urban size is related to the given variables (population, income, agricultural rent and commuting cost) The concept of the monocentric city involves a central unit, the central business district (CBD), surrounded by a circular residential area whereby land is allocated according to its most profitable use The general idea of the monocentric city is that most economic activities are based in the urban core, whereas suburbs only fulfill a residential function The monocentric city model and its extensive analytical predictions contributed to our understanding of the spatial expansion of the city over time, that is, the spatial variation in commuting costs, income, population, employment, agricultural land rents, and a home’s price The comparative statics predictions in the simplicity of urban spatial growth shed light on the dramatic changes

11

in urban structure from the CBD to the polycentric or sprawling development occurring farther from the CBD In the mid to latter half of the 20th century, employment began to disperse, with the proportion of jobs in the centre decreasing over time and most new growth in employment being located outside of the CBD of mega cities (Meyer & Gómez-Ibáñez 1981) According to various patterns of spatial organization and employment layout, urban spatial structure can be divided into monocentric and polycentric city models (Ingram 1997; Bertaud 2003; Ding 2007) Many cities evolved from a monocentric spatial structure into a polycentric city model Cities developed expansive suburban areas where growth rates of population sometimes exceeded those of their centre (Romein 2005) Muller (1981) proposed a concept of the Multi-centred Metropolis, which showed that some suburbs had transformed into increasingly independent and self sufficient urban entities beyond the older central city These urban entities do not only hold a rising share of the population

of metropolises, but also accommodated increase major economic activities, employment, and social, educational, cultural and entertainment services Additionally, Garreau (1991) put forward an idea of ‘edge cities’, a model of polycentric spatial structure usually located at major highway interchanges, to describe new suburban cities with office buildings and huge commercial infrastructures Anas and Kim (1996) established a computable general equilibrium model to demonstrate the emergence of urban sub-centers resulted from multiple equilibriums) Hall (1999) supposed that the Polycentric City focused on the location of business and envisaged a new polycentric urban form emerging in many contemporary cities The Polycentric City included six main elements: a traditional business core; a secondary business core; a tertiary business core or inner-city edge city; an outer edge city; outermost edge cities; and specialized concentrations In a Monocentric city, since there is a highly concentrated employment centre and residents generally live in suburb, there would be high commuting flows on radial routes to the centre because the urban commuting would have many origins for work trips but a concentrated destination In reality, a polycentric city functions in a similar manner as a monocentric city does – people are attracted by jobs from all over the city But the commuting patterns are different (Bertaud 2003) There are two commuting models of polycentric city One model is that the city has some sub-centres of employment of a similar scale and in such a kind

of polycentric city, each sub-centre generates trips from all over the city The characteristics of commuting present a wide dispersion of origins and destinations, appearing almost random Another model is that there are also different sub-centres of employment but one sub-centre is stronger and more concentrated than others Fujita, Krugman and Venables (2001) developed a theoretical framework to describe the evolution of polycentric structure that had resulted from centripetal forces and the agglomeration of economic relationships between firms It had become obvious over the years that the structure of many metropolitan areas diverged from the monocentric model and that many daily trips expanded in clusters over a wide area outside the original CBD (Bertaud 2003) These urban entities do not only hold a rising share of the population of metropolises, but also accommodated increasing major economic activities, employment, and social, educational, cultural and entertainment services Moreover, based on a wide literature review (e.g Bryant et al 1982; Champion 1999; Loibl and Toetzer 2003; Gallent et al 2006: Leontidou and Couch 2007) Figure 3.1 shows that:

 Urban core: It has the Central Business District and the site of many other civil and cultural functions and some public spaces associated with these

 Inner urban area: generally, in this part the higher density builds development (built-up areas) including residential, commercial and industrial types of uses and some public open and green space

 Suburban area: In general, this region has the lower density contiguous built-up areas, which are attached to inner urban areas, and where houses are typically not more than 200 m apart, with local shops and services, parks and gardens

 Urban fringe: It is a zone along the edges of the built-up area, which comprises

a scattered pattern of lower density settlement areas, urban concentrations around transport hubs, together with large green open spaces, such as urban woodlands, farmland, golf courses and nature reserves

13

Figure 3.1 PLUREL Concept of peri-urban area and rural-urban-region

Source: (Ravetz Joe, 2013)

In Figure 3.1

 Urban periphery: It is a zone surrounding the main built up areas, with a lower population density, but belonging to the Functional Urban Area, as below: this can include smaller settlements, industrial areas and other urban land-uses within a matrix of functional agriculture

 Rural hinterland: It is a rural areas surrounding the peri-urban area, but within the rural–urban-region and it is accessible within a practical commuting time so its rural character is affected by residents with urban incomes and lifestyles

The peri-urban area therefore includes both the urban fringe and urban periphery

It is a well known axiom in urban economics that cities grow to immense size when they are successful, and that as they grow, they develop scale economies and benefits that draw more people In fact, the polycentric is extension of the mono-centric model

By adding on peripheral centers, a city can keep on growing outward instead of (or as well as) upward This keeps trips and rents at a sufficiently low level, while many of the “single-large-city” attributes are preserved (Gat, 2005)

In figure 3.2, it shows the difference of monocentric and polycentric city in urban structure in two case studies: centralized and dispersed

The centralization (higher proportion of population live in center and) and – the dispersion (the population not live in centers but spread out across the territory in a non-concentrated pattern) dimension refers to the extent to which population and employment are centralized in cities or dispersed over smaller non-urban places in the area in a non-centralized pattern The monocentricity-polycentricity dimension reflects the extent to which urban population and employment are concentrated in one city or spread over multiple cities in the wider metropolitan area (see Figure 3.2)

Figure 3.2 Dimensions of regional urban form

Source: http://www.lboro.ac.uk/gawc/rb/rb330.html

The polycentric model examines forces within a metropolitan area (metro area) which

is a region consisting of a densely populated urban core and its less-populated surrounding territories, sharing industry, infrastructure and housing The polycentric model is affected land use patterns However, this study will determine the factors that affect urbanization at provincial level because of the condition and the system in the area or the region is different among component urban areas within a province Therefore, I will use the monocentric model instead of polycentric model to analyze those factors

Following the two equations in this model, we can observe that:

 At equilibrium point the total population fits inside the urban area

 The urban land rental and agricultural land rental are equal at the edge of the urban

Based on these relationships, the implications of the model can be seen from the signs

of Wheaton (1974)’s comparative statics In short, the model implies that:

(1) An increase in population clearly must increase the distance of the edge of the city since more people must be housed then it will increase urban land use and urban expansion

(2) An increase in household income will increase urban land use and urban expansion

(3) An increase in transportation costs will reduce urban land use and limit urban expansion

(4) An increase in the opportunity cost of non-urban land (agricultural rents) makes the city more compact It will reduce urban land use and limit urban expansion

3.3 Empirical studies

To investigate the urban expansion land, the monocentric urban model is deployed The model generates hypotheses illustrating the change of urban land area in some of the fundamental building blocks of economy such as income, population, agricultural land values and transportation costs Studies in US, Brueckner and Fansler (1983) first estimated the monocentric model with cross-section data of 40 small metropolitan regions (urban countries) and urbanized areas from 1970 In each metropolitan region, the urban area was contained within a single, relative small country was measure In their study, they found that income, population and agricultural rental were statistically significant determinants of total urban land area Each of these had the sign that was consistent with the prediction of the monocentric model The coefficients of the variables measuring transportation costs were not significant

McGrath (2005) estimated a model with panel data set from 33 large US cities for the time period 1950-1990 The study used the total urbanized land area in square mile for

33 cities in each sample area These measures were then converted to a variable that proxied for urban radius This study also found that income, population and agricultural rent were statistically significant determinants of urban expansion in the U.S with signs that were also consistent with hypotheses Unlike Brueckner and Fansler (1983), however, McGrath (2005) found that the coefficient on the transportation costs variable was also statistically and had the expected (negative ) sign McGrath (2005) used average annual CPI for private transport action for each year, rescaled to each region using private transportation cost data for 1990 McGrath

Kurt Paulsen (2012) uses all US metropolitan regions for the year 1980,1990 and 2000 and the result of his study , which was based on the usage of panel model shows that both population and income elasticity are less substantial than unity And also he hasn’t used the transportation cost The real cost of transportation by personal automobile in the U.S is not the cost of gasoline, insurance, maintenance, etc but it is rather the cost of time spent for commuting that would already be proxy for income Agricultural land value is also negative sign.

While Brueckner and Fansler (1983) find that proxy measures of transportation costs (percentage of households owning cars and percent transit) are statistically insignificant Spivey (2008) finds negative income elasticity in some specifications Spivey (2008), when restricting her sample to the 85 areas covered by the Texas Transportation Institute's “time travel index” and “congestion cost index,” does find the expected negative coefficient (increased congestion or longer travel times reduce the size of the urban area) However, the inclusion of the travel time or congestion indices results in land prices no longer being statistically significant The developing works of Muth (1969), Mills (1967, 1972), and other scholars established a

“monocentric city model,” emphasizing the importance of the CBD with respect to the degree of decentralization of population or employment Some empirical evidence, following a pioneering work on population decentralization by Clark (1951), indicated that population density and transportation costs decline farther from the existing central city while incomes increase with greater distances from the CBD (Edmonston, 1975; Mills & Tan, 1980; McDonald, 1989; McDonald & McMillen, 2007) These empirical estimations of monocentric models haves also recently been applied to cities

in China by Deng et al (2008), as Brueckner, Fansler and McGrath’s methodology suggested Income, population and transportation costs positively affect the spatial size

of the urban cores of China’s cities And agricultural rents detract from the expansion

of urban core Their study demonstrates the powerful role that income growth has played in shaping China’s current economy, including the spatial size of its cities And

in the future economists must utilize urban spatial and economic/demographic data over the time in the analysis of determinants of urban land size and decompose more precisely the changes in the spatial size of cites

In this paper, I will present an empirical study by using a panel data analysis with 519 observations for 60 cities and provinces in Vietnam from 2000-2011 The interpretation of coefficient estimated in a cities and provinces by using fixed-effects panel structure of a monocentric model would be the extent to which changes in a province’s or city's population, income or agricultural investment and other factors would be expected to result in changes in urbanized land area In my empirical study, I follow Deng et al.(2007) and Wassmer(2006), K.Paulsen(2012) I can achieve my empirical study

Table 3.3.1: The proxies and list studies of independent variables

Population The total of population in cities

and provinces

Brueckner ad Fansler (1983) and Burchfield

et al (2006) Industrialization(GDP1_

share)

GDP by industrial sector (industry and construction) divide by total GDP

Seto and Kaufman (2003) Service sector(GDP2_

share)

GDP by service sector divided

by total GDP,

Seto and Kaufman (2003) Agricultural investment The amount of investment

allocated to agriculture

Wei, 1993; Deng et al.,

2008

19

that measures the value of all goods and services produced in the cities and provinces during the year

(2005)

3.4 Analytical Framework

This framework shows the link of urban growth with the two factors: socio-economic and policy There are 5 elements that make urban expansion such as population, GDP, agricultural investment, industrialization and service sector

An increase in population will increase urban land use and urban expansion Population is the main factor that makes urban grows fast or slow Because of the demand of people in survival increases the demand of land and foods and job Moreover, when GDP increases the demanding of land increase and urban land also grows For agricultural investment, agricultural sector is the main factor in the growth

of economic in countries especially Vietnam, a developing country Government and private sector’s budget investment in agricultural fields increases every year Because

of the demand of human for survival- industrialization is defined by doing investment

in industrial fields By definition, urbanization can be occurred when industry land increase and of agricultural land use decreases Industries can absorb many of workers , especially for people earning average or less income Industry becomes the first factor contribution in economic growth in Vietnam Similarly, the service sector or

Agricultural investment

Industrialization

Service sector

service industry is one of the three economic sectors, the others are the secondary sector (approximately the same as manufacturing) and the primary sector (agriculture, fishing and extraction such as mining) The service sector is an importance part of economy

According to 5 elements, I can study on the effect of economic growth on the urban expansion land

In this study the 26 classes of land use were aggregated further into two classes of land use-residential land and non-agricultural land In my purpose residential land use means the land which has had dwelling erecting upon it for human habitation It means that when the population increases the demand of residential land will increase Whereas non-agricultural land is the land which is used for purpose not connected with agriculture It may be used in other fields such as manufacturing, business purpose, religious or other purpose Because of the missing data, I have to drop some provinces such as Hoa Binh, Hau Giang ,Lai Chau from my analysis Socioeconomic variables, like gross domestic product (GDP) for each provinces and provinces equivalent cities are from 2000-2011 And for the data of year are from annual statistical yearbook for 2009 and 2011 The demographic population data for 2000-

2011 is from the general statistics office (GSO) Following Xiangzheng Deng(2007) the agricultural rent was replaced by agricultural investment Investment in the agricultural sector accounts for one of the most important sources of the total fixed investment for a cities and provinces It is often used in the urban studies to proxy for the value of agricultural land (for example, Firman, 1997; Seto and Kaufmann, 2003) The data on investment into the agricultural sector for each provinces and cities are from the implementation management of the capital investment project under the department of agriculture and rural development6 Industrialization data (GDP1_share) and service sector data (GDP2_share) are from Vietnam’s annual statistical yearbook

of 2009 and 2011 The database includes the time frame data of 2000-2011.Several data sets were used to generate variables that measure the socioeconomic attributes of cities and provinces of Vietnam These data sets include information that will be used

6 www.mic.mard.gov.vn

to measure the variables for testing the hypotheses about the determinants of urban land use (GDP, population, agricultural investment, GDP1_share, GDP2_share)

4.2 Data description

Gross Domestic Product: GDP is a general indicator reflecting the final results of

production and business activities of the whole economy in a given period The real GDP removes the factor of price change and measuring the value of goods and services produced in the cities and provinces during the year Following an empirical study by McGrath (2005) we include the variable of GDP (Gross Domestic Product) as

a proxy for income to measure the effect of the overall level of economic development

on urban land use If GDP increase, the cities or provinces land seems to be expanded

or still the same So, the coefficient of the real GDP is expected to be positive The real GDP or GDP at constant 1994 prices by economic sector is from three economic sectors (agriculture, forestry and fishery plus industry and construction plus service sector) The real GDP measure generated from cities and province’s annual yearbook for 2009 and 2011 of Vietnam

Population: According to the empirical studies by Brueckner ad Fansler (1983) and

Burchfield et al (2006), it is the measure of the level of the population of country sampling unit Population is the total number of people inhabiting a country, city, or any district or area The demography of population shows that the demand of land use for housing and other sector will increase The population growth rate in Vietnam in the past ten years is the fastest one, especially that in Vietnam’s largest cities of Hanoi and HCMC, its surrounding regions and in some medium sized cities With few exceptions, small cities grew the least or lost their population Here I use population data at cities and provincial level in Vietnam by general statistics office (GSO) from

2000 to 2011

Industrialization: The process in which a society or country (or world) transforms

itself from a primarily agricultural society into one based on the manufacturing of goods and services Industrialization is when a community is transformed socially and economically into an industrial society Industrialization is a part of a broad modernization process Seto and Kaufman’s (2003) measure of industrialization (GDP1_share) is constructed as the value of GDP created in the industrial sector divided by total GDP GDP1_share is an important variable for pushing economic

23

growth and urban expansion land especially in developing countries The process of industrialization data is from Vietnam’s statistical data This sector is considered as a part of industry such as industry and construction

Service Sector: Similar to industrialization, service sector (GDP2_share) is equal

GDP of service sector divided by total GDP The service sector or service industry is one of the three economic sectors, the others are the secondary sector (approximately the same as manufacturing) and the primary sector (agriculture, fishing and extraction such as mining) The service sector is an importance part of economy

Agricultural investment: There is no direct measure of agricultural rent in Vietnam

As a proxy for agricultural rent, I include a measure of the amount of investment allocated to agriculture It is also well known (Wei, 1993; Deng et al., 2008) that agricultural investment has an important effect on a regions development For example, agricultural rents-is identified as any opportunity cost for converting to urban area In Vietnam, agricultural investment is measured by the sum of budget investment project each year of each province for each year of mine, agriculture, forestry and fishery investment In Vietnam, investment in agriculture of government and private sector are increasing every year Especially, for cities and provinces have a lot of people, the investment in agriculture also increases

Urban expansion land: There are many factors that it makes urban expansion land

but in this paper I focus on two main factors such as residential land and non agricultural land Following Monocentric model, an increase in population will increase urban land then the demand of residential land for housing will increase similarly to non agricultural land for demanding of industrial land and other

4.3 Research Method

Following the monocentric urban model, I get:

Urban expansion land= f (GDP, population, agricultural investment, GDP1_share, GDP2_share, Dicities or provinces)

Where

LULA= Log of urban land area (hectare)

LGDP= Log of GDP (million dong)

LPOP= Log of population (thousand)

LAGI= Log of agricultural investment (million dong)

IND=Industrialization= GDP1_share= gdp by industry sector/total gdp (In ratio) SVS=Service sector = GDP2_ share=gdp by service sector/total gdp (In ratio)

αK: coefficient (k=0,1,2,3,4,5)

εit is error term

i and t are indices for individuals and time

i: number of cities and provinces (i=1,2,…,22,24,25,…,59,61, 63)

t: time dimension (t=2000,….2011)

 Fixed effects model (FEM)

The fixed effects model is regressed by the following equation

𝐿𝑈𝐿𝐴𝑖𝑡= 𝛼0+𝛼1𝐿𝐺𝐷𝑃𝑖𝑡+𝛼2𝐿𝑃𝑂𝑃𝑖𝑡+𝛼3𝐿𝐴𝐺𝐼𝑖𝑡+𝛼4𝐼𝑁𝐷𝑖𝑡+𝛼5𝑆𝑉𝑆𝑖𝑡+𝜀𝑖𝑡 (2)

 Place fixed effect

The place fixed effects model is regressed by the following equation

𝐿𝑈𝐿𝐴𝑖𝑡= 𝛽0+ 𝛽1𝐷1+ 𝛽2𝐷2+ ⋯ + 𝛽63𝐷63+𝛼1𝐿𝐺𝐷𝑃𝑖𝑡+𝛼2𝐿𝑃𝑂𝑃𝑖𝑡+𝛼3𝐿𝐴𝐺𝐼𝑖𝑡

+𝛼4𝐼𝑁𝐷𝑖𝑡+𝛼5𝑆𝑉𝑆𝑖𝑡+𝜀𝑖𝑡 (3)

In fixed effect model, all dummy variable are omitted, our analysis does not include those dummy variables (See Appendix)

4.4 Methods for Analysis

I used a Panel data (also known as longitudinal or cross-sectional time-series data) to analyze the effect of economic growth on urbanization in Vietnam Because of missing some data, I do analyze on a micro-panel data set and unbalance panel data from 2000

to 2011 with gaps Panel data is a dataset in which the behaviors of entities are observed across time These entities could be states, companies, individuals, countries, etc In panel data, individuals (persons, firms, cities,…) are observed at several points

in time (days, years, before and after treatment,…) Panel data are most useful when

we suspect that the outcome variable depends on explanatory variables If such omitted variables are constant over time, panel data estimators allow to consistently estimating the effect of the observed explanatory variables In this method, it includes fixed effect method and correlation test, multicolinearity test, normality test, heteroskedasticity test

25

4.4.1 The fixed effect method

Fixed-effects methods have become increasingly popular in the analysis of

longitudinal data for one powerful reason: they make it possible to control all stable

characteristics of the individual, even if those characteristics cannot be measured Fixed effects regression methods are used to analyze longitudinal data with repeated measures on both independent and dependent variables

They have the attractive feature of controlling all stable characteristics of the individuals, whether being measured or not7 Similarly, in econometrics and statistics,

a fixed effects model is a statistical model that represents the observed quantities in term of explanatory variables that are treated as if the quantities are non-random In panel data analysis, the term fixed effects estimator (also known as the within estimator) is used to refer to an estimator for the coefficients in the regression model

If we assume fixed effects, we impose time independent effects for each entity that is possible correlated with the regression8 By using fixed-effect (FE) whenever you are only interested in analyzing the impact of variables that vary over time Fixed effects explore the relationship between predictor and outcome variables within an entity (country, person, company, ect.) that may or may not influence the predictor variables When using FE, we assume that something within the individual may impact or bias the predictor or outcome variables and we need to control this The advantage of fixed effect regression is the control on all possible characteristics of the individuals in the study9