Research " AN ASSESSMENT OF SPATIAL DIFFERENTIATION IN VIETNAM’S SOCIAL-ECONOMIC DEVELOPMENT 1990-2000 " doc

Trang 2

® UMI

UMI Microform 3083400

Copyright 2003 by ProQuest Information and Learning Company All rights reserved This microform edition is protected against

unauthorized copying under Title 17, United States Code

ProQuest Information and Learning Company 300 North Zeeb Road

P.O Box 1346

Trang 4

Significant changes have been taking place in the process of national development in Vietnam since the late 1980s Adjustments made in the country’s international policy

and economic policy have the effect of returning it into the world’s economic

development mainstream and, eventually, bringing about a steady economic performance

not observed before However, efficiency-based growth in general and the export- oriented development model in particular may also cause a distinctive pattern of uneven

development in the form of disparities across space and population strata This trend is

known as the polarization reversal hypothesis, an argument advanced by Western neo-

classical scholars in order to deter an active public policy intervention aimed at alleviating inequalities

This study is aimed at detecting whether regional disparity, one of the key

expressions of uneven development, has increased in Vietnam under the new pattern of growth in the 1990s Data analysis shows that there is an increase in disparities between

provinces in Vietnam with respect to important indicators such as the gross output of

industrial, agricultural production, and services sector The country appears to follow the first half of the polarization reversal hypotheses in terms of both widened gap in crucial economic indicators and spatial concentration of economic activity Results also indicate that some aspects of regional disparities such as human development indicators had already existed even under centralized planning

Trang 5

I wish to express profound thanks to my dissertation committee members, Dr Ashok Dutt (advisor), Dr Gasper Garofalo, Dr Ralph Hummel, Dr Peter Leahy, and Dr Larry Lebedur (Cleveland State University) for their invaluable and knowledgeable advising, and patience A special thank goes to Dr P Leahy, whose advice and support

was crucial not only on this final stage but also throughout my study at the University of Akron

Trang 6

Page

LIST OF TABLES .-QQQQQQQ HH HH HS HT HH KT Bàn kh vu vil

II 6):8:11€)0):6 zaiẦẳẳẳiẳẳẳiẳlẳlddiidiiiẳẳẳảỶÝ 1X CHAPTER

I NTRODUCTTIƠN HQ HH HE Km nh Hinh nà ni khu bee 1

1 Theoretical Framework in Regional Development Studies | ¬—— 4

2 Research Signiicance, Purposes, and Hypotheses 10

3 Quantitative Methods of Analysis and Data SOUFC€S 23

Il LITERATURE REVIEW OF NATIONAL AND REGIONAL

DEVELOPMENT THEORIES QQQQ nQn nn HH nà nh he nh ng 33

1 National Development TheOTI€S - Ăn nen 33

2 Regional Development TheOrIes -.- cớ 44 3 Overall Review of National and Regional Development Theories : 79 1H NATIONAL AND REGIONAL DEVELOPMENT IN VIETNAM 87

1 National Developmeit - - cọ nọ HH nh He nh như, 87

2 Regional Developm€nt ch nh nh 108

3 Overall Review of National and Regional Development in Vietnam 133 1V.DATA ANALYSIS AND INTERPRETATION nen He, 139

Trang 7

3 Regional/Provincial Disparitles in Development - - 152 4 Rural-Urban Gap in Developmert - «chen he nhe nh 169

5 Urban Centers and the Process of Concentration/Dispersion of

EconomiC ÀCVIẨV HH HH TY ni KT Ki th nh th kề kế 173 6 Some Caveats on Data Any§1S - nh nh nh nh nh nh nh 183

V SUMMARY AND CONCLUSION nh nh nh nh kh nhe 186 REFERENCES - cọ HH nọ HH nh KH Ki KH HT th nh 192 APPENDICES -. con HH nh ĐH HE Ki Km TK nh nh th tin nh hà 208

APPENDIX 1 PROVINCES’ GROSS OUTPUT, PER CAPITA OUTPUT,

AND AVERAGE ANNUAL GROWTH RATE 1990-2000 209 APPENDIX 2 PROVINCES’ FOREIGN CAPITAL INVESTMENT AND

PERCENTAGE URBAN - con nhe hee 211 APPENDIX 3 PROVINCES’ GROUPING BY THE INDEX OF

ECONOMIC ACTIVITY AND GROSS OUTPUT 1990-2000 213

Trang 8

Table 3.3 3.4 3.5 3.7 3.8 3.9 4.1 4.3 4.4, Page Comparison of Three Main Development Schools:

Modernization, Dependency, and World-System PerspectIves 43 Friedmann’s Stages of National Development and Regional Policy Role 57

Outflow and Inflow of Foreign Direct Investment Stock and Sectoral Distribution for the Largest Home and Host Countries from 1970

TO S210 98

Foreien Direct Investment 1n Vietnam 1988-2000 nŸàằ 101 Income Indicator in Four Asian Countries1980-1999 cằ 103 Economic Structure 1980-1999 HQ HH HH nh kh nh nh nh 105 International Trade Volumes as Percentage of GDP cà cà 106 Urbanization in Four Countries 1960-2000 cành he 108

Vietnam’s Economic Development Reglons - -+- «<< 110

Selected Regional Indicators in 1999 che nhà “Ce naeeeeees 117 Shares of Urban and Rural Population in North and South Vietnam,

ÌTI D€TC€TÍ -c SH n HH» TS nh nh kh ng Ki min tà kh hà th 3g 126

Contribution of International Trade and Foreign Investment to GDP 143 Correlation Coefficients between Foreign Investment and

Main Economic Sectors at Provinces” Level «<< 50 Correlation Coefficients between Percentage Urban and Main

Economic ŠS€CfOFS -c HS nu HH nh nh nh nhà nh 151

Coefficient of Determination RỶ for Main Economic Indicators

Trang 9

Provinces” LeVeÌ - cọ n HH HH nh nh ng nh ng 155

4.6 Unweighted Coefficients of Variation for Selected Per Capita

Indicators, Provinces” LeVelÌ - cac nàn nh nh nh nh 156

4.7 Gini Coefficients for Selected Per Capita Indicators, Provinces’ Level 157

4.8 Gini Coefficients for Selected Per Capita Indicators, Regional Level 158

49 Factor Analysis Principal Component cesceceteneeeeeeeneneeeen eee eee 161 4.10 International Comparison of Regional Disparities in Income 166

4.11 Rural-Urban Differences in Selected Indicators 1990-2000 171

4.12 Regions’ and Largest Cities’ Shares in Industry and

Services 1990-2000 LH ng HH HH Hà nh nà nà nà kh nh nh 174

4.13 Regions' Shares im the Primary Sector 1990-2000 c es 175 4.14 Vietnam City Rank-Size Distribution in Comparison with

Selected Asian CoUntri€§ nàn nhe nh nhe Hy kg 181

Trang 10

Figure 1.1 1.2 2.1 2.2 3.1 3.3 3.4 3.5 3.6 4.1 4.2 4.3 4.4 Page

U-inverted Relationship between Regional Inequalities

and Economniic Developmmnen( -. sen nen 17

Vietnam Reglons and PTOVIICES - cSQ nnnnns - 32

A Classiication of Regional Development Theorles - - - 47 Regions’ Convergence in Capital/Labor Ratios under

Neo-Classical AssumptiOnS ‹ - con SH» HH nh nh hưn 53 Vietnam in a Classiication of National Development Strategles 96 Per capita GNP PPP in Four Asian COunfTi€S - cà cà se se eị 103 Sectoral Economic Structure 1980-1999 cu HH HH nhe kh nh hen 105

Interational Trade as Percentage of GDP nàn nen 107

Urbanization Trenđs in Four COunr1€§ - <2 cà S2 108 State Policy Preference and Capacity and Reglonal Disparities 129

Shift in Sectoral Shares of GDP, 1990-2000 cc ccc eee c cece eee eee e een ea nanos 146

Gini Coefficients and Coefficients of Variation for Per Capita

Gross OUtPUt ccc cece cece eee eee eee nee eee nh KH KH ng tt Bế 159

Factors 1 and 2 Scores Distribution -.- con nen nh ướn 163 Factors 3 and 4 Scores DistrIibution cuc nh He nen nhe nen 164

Trang 11

INTRODUCTION

The process of social and economic development in Vietnam as one country since the re- unification in 1975 has witnessed two salient periods In the period from 1975 until the late 1980s, the country’s development path was faced with serious obstacles and

challenges such as post-war rehabilitation, unstable and hostile regional and world political-economic environment resulting in the loss of traditional trading partners and

aid donors, and an economic and trade embargo Under these circumstances, already limited resources had been diverted to unproductive defense and security spending that

otherwise could have gone to capital accumulation, a crucial process especially at the early stages of industrialization In addition, within the economy there was a severe resource misallocation problem that resulted from an ineffective developmental policy aimed at self-reliant and self-sufficient industrialization based on heavy state intervention through a cumbersome centralized planning regime Contemporaneously and

intertwiningly, these events contributed significantly to retarding national social and

economic development In fact, Vietnam’s GDP annual growth rate averaged around 4 percent in the period 1975-1990 (Fforde and Vylder 1996: 167), as opposed to that of 7 to

10 percent in a number of neighboring East and Southeast Asian economies (Papanek

Trang 12

accumulation and other development needs

Since the late 1980s, however, socioeconomic development in Vietnam has

undergone significant transformation and important changes Utilizing a more favorable international environment and the ever-intensifying process of the world’s economic

globalization and integration, the country liberalized its trade and investment policy,

introduced some elements of free enterprise, and was able to reverse the previous downward trends in the economy (Tran 1997: 13-15) Thus, there have been created favorable conditions for Vietnam’s crucial “take-off” stage of modernization, to use a W Rostow’s (1960) term

The process of economic restructuring at the national level has accelerated growth

in a number of sectors and regions and decline in some other sectors and regions A parallel process, the process of the state structures’ change in function from essentially a

redistributional system typical for a socialist economy to an accumulatory system typical

for free enterprise, has the effect of realigning national priorities in order to conform to

the needs of economic growth These processes unavoidably lead to socioeconomic stratification among regions and population groups Further, economic reforms in centrally planned economies often imply decentralization of decision making including investment and resource allocation decisions that, in turn, have a positive relationship

Trang 13

1999; World Bank 1999a; Nyland and Pol 2000; Glewwe at al 2002)

Although the problem of spatially inequitable development persists in almost all

economies and systems, it is especially acute and widening in a developing economy because of its limited resources and multiple urgent needs and priorities Lipton (1977:

173-177) posits that in developing countries, inequitable development in general and

spatially uneven development in particular have numerous ramifications and that inequity

is inefficient This proposition is consistent with that of Harvey (1973: 15) who,

emphasizing the totality of any social phenomenon, emphasizes that distribution must be analyzed in conjunction with production and that, therefore, efficiency is equity in

distribution in the economy Apart from overall societal ramifications of inequalities that

abound in the works of Harvey, Lipton, and other social science researchers, economists

have provided various formal models to incorporate equity into economic analysis to

show how inequality can hinder efficiency Persson and Tabellini (1994), for instance,

drawing on historical evidence in nine most developed countries, provided a model to prove that inequality is harmful for economic growth

In general, as Gore (1984: 25-26) has observed, regional disparities can be

Trang 14

Obviously, regional socioeconomic problems and public policy measures aimed at

solving those problems can be approached from various perspectives On the other hand, the goals of intraregional and interregional equitable development remain the cornerstone

of regional policy Further, as Slater (1989) notes, spatial development policy does not

occur in a political-economic vacuum; rather, it is one of the important components of the

national development strategy a political-economic system uses for its reproduction, or perpetuation -to use Harvey’s (1973: 200) terms To the extent that national development theories are diverse, regional development theories themselves are also very diverse Within a given national development policy framework, then, the rationale for regional policy may be multifaceted and have an ever shifting content depending on the context of development

In this study, an analytical framework proposed by C Gore shall be followed In addition, as J Friedmann (1973: 42), and F.C Lo and K Salih (1981) suggested for developing economies, in order to articulate regions and nation, regional studies must also concern overall national development issues In a more recent study, W Y Li (1990)

proposed a‘perspective on spatial development problems in China —a former centrally

planned economy in transition that is also somewhat similar to Vietnam in many terms: cultural, political, economic structure, etc., albeit not by scopes and scales Some major conclusions in Li’s study are also reviewed here to complement the theoretical

Trang 15

Specifically, Gore (1984: 25) has succinctly pointed out that developing countries most commonly face the following three regional policy problems:

i Interregional disparities in development;

‘ii The problem of the excessive size of the national metropolises; iii Rural-urban inequality

In consequence, Gore suggests that the objectives of regional policy are to

alleviate interregional and rural-urban inequalities, and constrain the growth of primate

cities

Based on their extensive observations and analysis of regional problems in a number of countries in Southeast Asia, two regional planning scholars from the

developing world, Lo and Salih (1981) called for a new approach to regional development

in the developing world They synthesized six issues faced by planners in elaborating

regional development policy Overall, Lo and Salih’s approach to regional problems is similar to Gore’s three policy problems The only difference between the.two approaches is that Lo and Salih emphasize, in the first two issues, more on the articulation between

regions and nation Clearly, the issue of nation-regions articulation is important, as the practice has shown in the last 50 years in the developing world’s context These six issues

are as follows

i Regional problems must be analyzed within the considerations of the national development policy and alternatives

Trang 16

political processes that influence it Consequently, the formulating of regional policy ina particular developing country must hinge on the objective conditions of that county’s political and economic forces Fashionable regional development strategies like the growth pole, the agropolitan approach, etc that have been much embraced somewhere

else, must be critically analyzed within those conditions

ii Regional policy is formulated based on the particular regional economic structure in relation to national economic development as a whole

Broad national processes then exert different influence bringing about different economic

outcomes even among various regions within a country Regional policy toward nation’s

regional categories must contain differentiated contents At the same time, regional-

specific policies must be properly integrated with each other within the national

framework of development In other words, regional policy must maximize each region’s priorities within properly elaborated national priorities Consequently, Lo and Salih

proposed four different types of regional spatial policy for roughly four types of sub-

natiorial regions: regions with metropolitan dominance, regions close to a major growth

center, regions with a relatively developed agricultural economy, and regions with rural

dominance

iii Regional development process must be based on rural-urban relations

By its nature, spatial development must foster mutually reinforcing rural-urban relations for local regional development In the developing world’s context, however, it often

Trang 17

urban-effects on rural areas In fact, as Lo and Salih note, strategies focused on rural

development per se may ignore the implications of broader urban processes and their

positive impact for the surrounding region in terms of seasonal employment, rural industrialization, etc and thus may aggravate existing contradictions in rural areas And

certainly, urban-bias policies adopted in most developing countries have had anti-rural impacts in the form of lack of investment in rural areas, which increased rural

unemployment and rural poverty

iv Regional policy must view the role of the urban system in the region and fully incorporate different functions of the cities at different levels of the urban hierarchy Allegedly, one of the direct implications of the urban-bias development policy in the developing world is the problem of very big city or cities, also known as the primacy

problem In the developing world, primate cities often represent themselves as the

primary expression of over-urbanization —an urbanization process that is not supported by an adequate urban-industrial development that can provide enough jobs, housing, and other minimally necessary urban public services to the urban population as well as in-

migrants from rural areas On the other hand, it is generally agreed that urbanization is a

concomitant of the industrialization and modernization process Therefore, in Lo and

Salih’s view, rather than a controversial primacy question that needs to be tackled, the

region’s existing urban system must be taken as given, as a base for regional

Trang 18

spatial terms Rural settlement types at the lower end of the regional settlement hierarchy must then be viewed from this rural-urban relational perspective

v Regional policy must consider the territorial organization and development of rural commuhities and work toward enhancing rural-urban linkages

In fact, the regional settlement system that is based on a harmonic rural-urban relationship will eventually determine territorial organization and rural development as one of the core

issues of spatial development under the name of the spatial articulation at the intra-

regional level Intra-regional articulation, as opposed to interregional articulation —or

intra-national articulation of subnational territorial units proposed in issues 1 and 2, above~ more emphasizes policy measures that promote development from below and within That is, aside from centralized investment, regional policy should emphasize a greater internalization of development processes and enhanced self-reliance through

utilization of local resources: human and natural, organizational arrangement, and

political participation

vi Special attention in regional policy must be paid to the development of the poorest rural communities in the study region

Regional policy always implies equity and equitable development as one of its main contents In fact, the success —or failure— of regional development policy should be

Trang 19

In the contemporary China, another socialist developing economy in transition, Li

(1990) identifies four major problems of spatial development emerging as a result of economic reforms as follows

i The relationship between national economic growth and unbalanced regional

development

Overall high national economic growth, especially in the last two decades, in China and increasing imbalances between its regions imply that growth concentrates in certain regions and that there are regions that do not contribute to national growth Since its

foundation, the People’s Republic of China has attempted several national development

strategies None of them worked particularly well in terms of boosting development in the

lagging areas Reflecting current official development policy in China, Li asserts that

although recent policy emphasizes investment in the more advanced core areas tends to

increase the regional gap in the short run, but in the long run it may create the possibility of bringing around enough resources to narrow regional differences

ii Uneven development between the coastal areas and the interior

Unbalanced regional development is the most serious among spatial issues in China The growth areas are coastal regions and cities that have advantages in commerce,

governmental institutions, and scientific, educational and cultural institutions The

lagging areas are interior regions and rural areas

iii Regional comprehensive development as opposed to specialization

This is also a part of the problem in the centralized management system in China where

Trang 20

specific regions based on their potential, although the specialized sectors do not always serve local basic needs or promote regional comprehensive development

iv Rural-urban relationship

In a way similar to that of Gore, and Lo and Salih, Li also focuses on the relationship

between rural and urban areas as one of the four basic contemporary spatial issues that

have emerged in China Li observes that, although China’s rural industrialization had

achieved certain scales (see Chapter 3), the imbalance in living standards between urban and rural areas tended to increase Part of the problem is that while cities and coastal regions obtained certain discretion in economic decisions, rural enterprises were still

controlled directly by the authorities (Linde and Forbes 1990) There, therefore, needs to be a coordinating mechanism for inter- and intra-regional economic development

2 Research Significance, Purposes, and Hypotheses

2.1 The Significance of Equity and Spatial Equity in National Development The conflict, or trade-off, between equity and efficiency is a long-standing one in

economic development strategy, for economic growth and social-economic equality have

been the most essential aspects ofthe processes of modernization and development in any developed and developing countries In the socialist economies, these values are even more clearly evident where the official ideology emphasizes an egalitarian pattern in the distribution of benefits (Bielasiak 1980) In a market economy, however, a large degree of

inequality is accepted for the sake of efficiency (Okun 1975: 88) There are also

Trang 21

(Richardson 1979) But possibly because these are very much system- and time-specific,

they have not been well studied

In the socialist economies in transition, the shift from an ideological worldview entirely committed to equal distribution of benefits and wealth across population strata and regions to the one that tries to maximize equity within a given level of efficiency,

would lead first to a situation when inequality becomes acceptable, but not rising

inequality (Wang and Hu 1999: 203-204) In any perspective, equality including interregional equality is important for a variety of reasons For one, the problem of regional disparity is an issue of the overall process of economic growth and activity; it

persists in most economies and is therefore always at the center of scholarly research and public policy debates The most serious issue is, coupled with cultural differences,

regional economic inequalities are often one of the key factors leading to interregional

tensions and even national disintegration with prominent cases being the Baltic states in the former USSR, Slovenia in the former Yugoslavia, the Czech Republic and Slovakia,

etc (Kincaid 1995; Wang and Hu 1999: 202) In this light, rising regional inequalities could later become significant obstacles to the overall national development process and

therefore special attention needs to be devoted to

From an economic point of view, the regional aspect of investment in the

developing world has been a subject of intensive study by scholars since the early stages

of the newly independent countries Back in the 1950s, pioneers in economic

development theory such as G Myrdal (1957) and A Hirschman (1958) identified the

Trang 22

economic development in less developed countries In fact, after more than a decade of independence, most developing economies still grappled with basic problems such as

mass poverty and underdevelopment Analyzing this situation, Hirschman argues that, due to capital shortage, developing countries must concentrate investment in regions most favorable for growth In his view, regional inequality is concomitant of development, but it could be erased in the long run On the other hand, not every scholar agrees with

Hirschman’s (1958: 183-184) argument that interregional inequalities stemming from economic growth are an inevitable condition of growth itself and will be leveled off in the long run Some influent scholars however argue precisely for the contrary M Lipton

(1977: 178-199), for instance, states that the dispersal of resources to the rural sector is

the best way to maximize growth, provided that the agricultural product is properly

valued

A number of regional development theorists have tried to approach national economic development problems not just on a purely economic basis In fact, in the _

broadest terms of not just economic growth but overall societal transformation and

development, inequality is socially and, therefore, economically inefficient (Lipton 1977: 72-74) Inequitable growth across space and population threatens national coherence, social stability, spatial integrity, and subsequently, the process of economic growth itself Considering production and distribution as indivisible sides in economic activity, Harvey

(1973: 15) correctly observes that “production is distribution and efficiency is equity in distribution” for the society as a whole After decades of promoting efficiency-based

Trang 23

like the World Bank have recently come to recognize that equity in the form of

redistribution can be good for efficiency (World Bank 2001: 57) Thus, in certain sense, equity is efficiency

A prominent approach to harmonizing efficiency and equity at the level of national policy making can be found in the development practice in newly industrialized

countries in East Asia such as South Korea and Taiwan, and in countries close to the status of industrialized economy such as Malaysia These economies show that equitable development is an important component of their industrial development success Contrary to Western economic development theories (reviewed below), all of which posit that in

developing countries inequalities in every realm widen at the early stages of development,

these newly industrialized economies had actually achieved a declining level of

inequalities (World Bank 1993) And these developmental states strove for equity not just

post-mortem —i.e., after certain levels of accumulation and efficiency-based economic growth have been achieved— but from an early stage of development In Malaysia, for instance, the 1971 New Economic Policy, which laid the foundation for successive

industrialization, set goals explicitly aimed at eradicating the imbalances in income -

distribution, employment, and control of wealth (Anand 1983: 10-17; Lubeck 1992) By its essence, as Weaver (1981) has noted, regional development is, above all, an

“ethical/political question” For countries like Vietnam, equitable development is more

than a mere economic growth issue; it is also a moral and ethical issue, for somewhere

between 70 to 75 percent of the total population live in rural areas; and this population

Trang 24

providing the revolution with unconditional support There exist therefore certain

political pressures, or even necessities, to satisfy expenditures on social maintenance and welfare support, even though this support can be rather symbolic on a per capita basis

Further, from any policy perspective, measures aimed at advancing economic growth and improving people’s living standards without decreasing the existing gap in living

conditions between rural and urban sectors, and between the core regions and peripheral

and hinterland areas are of little meaning 2.2 Purposes of the Study

a The main purpose of this research is aimed at identifying whether regional disparities have widened in Vietnam in the reform period in the 1990s and what are the major factors

contributing to recent economic growth and spatial economic restructuring at the national and regional/provincial levels and how these factors are correlated with recent changes in

interregional inequalities and with the consolidation of the core-periphery development model Specifically, special attention will be paid to examining S Kuznets (1955) and J Williamson’s (1965) polarization reversal hypothesis in interregional income trend, and

H Richardson’s (1977, 1980) polarization reversal hypothesis in interregional economic

activities in developing countries These studies are important in many perspectives and

“have been a starting point for many researches in social-economic development in various

countries in the Third World Their main focus and conclusions are as follows

In his pioneering article on growth and inequality, Economic Growth and Income

Trang 25

Kingdom, Germany, and the U.S— during their respective industrialization This was later to be called the polarization reversal hypothesis, or U-inverted relationship between

development level and the income gap In basic terms, this hypothesis argues that this

relationship is not necessarily a positive one: at the early stages of development, the income trend is one of polarization, but later on as economic growth is sustained, it will converge According to Kuznets, income gaps are usually small when the economy is ata low level of development However, inequalities tend to widen when the economy moves

up to middle levels of development Income inequalities become smaller again at high levels of development (see Figure 1.1)

Kuznets’s proposition was later confirmed in an empirical study by Williamson

(1965) and, later on, by other researchers Williamson’s study was considerably larger

and used observations from 24 countries and U.S states and dealt directly with regional ' inequalities in income This study’s conclusion is, “regional inequality passes through

three distinct stages as an economy moves from early development to maturity In the early stage of development, regional inequalities widen This is followed by a period of

stable, high-level inequality, and finally a marked secular trend towards regional equality

sets in as development proceeds” (Friedmann 1973: 76)

W Alonso (1980) further expanded Kuznets’s polarization hypothesis to other

societal processes He asserted that all five major societal processes —such as society’s

Trang 26

processes tend to grow at the early stages of development, but they tend to decrease at the

later stages In general, this U-inverted type of relationship (or bell-shaped, or

polarization reversal) is often used by neoclassical economists and location theorists as a

convenient and universal explanation to describe how inequalities behave during various stages of development In their view, perhaps this is because the nature of diminishing retums in production also‘applies to social processes

With respect to the space, Friedmann (1966: 34-3 7) and Richardson (1980) also

described a similar pattern of polarization reversal in national spatial development in developing economies Friedmann argued that, under overall societal forces including economic forces, the national spatial trend is one going from a dispersion pattern of

multiple self-sufficient and self-contained iucleus to concentration of a single center- periphery and returning to dispersion again, now in the form of multiple interdependent

nucleus Richardson, on the other hand, explained the phenomenon exclusively from an economic point of view In essence, according to Richardson, spatial development in an economy undergoes a polarization reversal process in which economic activities initially concentrate in certain locations, but eventually will disperse in the space economy

Precisely like Hirschman, Richardson argued that, because of scarce resources, the urban-

Trang 27

is distorted so as to further the core’s economic interests Eventually, however, the

monocentric spatial structure formed around the core’s primate city becomes

economically inefficient due to rising costs of factor inputs Economic activities begin

dispersing out of the core region into other regions of the national spatial system, making

' the space economy more and more decentralized Regional Disparities an

Spatial Concentration

Levels of National Development

Figure 1.1 U-inverted Relationship between Regional Inequalities and Economic Development

Source: Adapted from Williamson (1965)

The process of national development in Vietnam since 1975 and especially since the late 1980s seems to bear all of the characteristics of a polarizing growth pattern prescribed in many studies for an economy at its early stages of growth For one, efficiency-oriented pattern of growth is no doubt the primary cause of inequality and polarization, as even the

economic development theorists Myrdal and Hirschman acknowledged Further, export- oriented industrialization and dependency on foreign direct investment, as Haggard

Trang 28

inappropriate products, new pattems of consumption, industrial concentration,

displacement of local producers and labor forces, segmentation of labor market, and so forth All this generates inequality not just across population strata, but also across space

b Another purpose in this study is to examine, in qualitative terms, if Vietnam’s current

development patterns and practices conform to, or can be predicted by, any of national

development theories advanced by Western scholars and some developing world scholars The national development theories, chosen for review in Chapter 2, are relevant in this

study because they have direct applications in spatial theories and practices (Hansen

1981; Soja 1989)

Within the theoretical framework proposed by Gore and Lo and Salih, and in accordance with the propositions by Kuznets, Williamson, Alonso, Friedmann, and Richardson, the following research hypotheses are then formulated and are to be tested " and statistical inference made with respect to regional development in Vietnam

2.3 Research Hypotheses

a Admittedly, the regional analysis framework theorized by Gore, and Lo and Salih, and empirical studies conducted by Williamson, Richardson, and other scholars had been

synthesized for economies with more or less stable and normal conditions for

development for a relatively long duration of time More specifically, these studies had

been intended for, and actually based on research and data from, developing countries

with a market economy orientation, which also did not have to endure any major

domestic political and economic calamities In the Southeast Asian region, for instance,

Trang 29

the other hand, the development process in Vietnam had been, as noted previously,

distorted for a long time Since the late 1980s and early 1990s, however, with a reform aimed at liberalizing economic activity and management and integrating into the regional and world economy, the country has been enjoying development conditions somewhat similar to those in the countries in the region The pattern of regional development in Vietnam should follow a similar manner suggested by the cited scholars The research

hypotheses then are aimed at examining this pattern as well as its magnitude

Although steady economic performance in Vietnam during the transition from a planned economy to a more market-based economy has often been credited to the

working of market mechanisms, it also poses quite a puzzle since there has beena lack of

‘what is often considered to be crucial prerequisite for a market economy: clear private | property rights backed up by a relatively autonomous legal system (Fforde 1999: 45) It,

however, can be argued here that the fastest growing sector in Vietnam’s economy during this period is international trade and foreign investment, access to which had been

previously denied on ideological grounds With respect to provinces and regions where the real growth occurs, there is a clear gap in trade volume and in the returns from trade

due to the presence or absence of factors —known as comparative advantages— such as

natural resources, favorable location, existing social and.physical infrastructures, and

Trang 30

Hypothesis H-1: International trade and foreign investment have been the most important determinants of Vietnam’s economic growth in the period

1990-2000

Hypothesis H-2: In the 1990s, regions and provinces that have

comparative advantages in international trade and foreign investment have

higher economic’ growth rates than regions and provinces that lack these _

conditions

b It is generally acknowledged that Vietnam has achieved impressive results in fulfilling basic needs since reforms began However, there are very few studies to examine whether these achievements are evenly distributed across regions and provinces A major goal of this research is to examine interregional gaps in Vietnam since introducing the new economic policy and detect if these gaps are widening under the new pattern of economic - development, i.¢., one that is efficiency-oriented and based on foreign investment and

export-oriented industrialization In fact, one of the major problems of the type of | development based on foreign investment is that developing countries cannot really dictate the location and ‘sector of investment to foreign investors (Sklair 1990) This is

because, at certain points of development, the pressure of economic growth and the need

for creating employment would prevail over all other considerations, such as immediate

and long-term consequences in many realms of sustained development such as

environment and spatial equality The need for growth and other short-term _

Trang 31

tends to concentrate in certain sectors and geographic locations furthering regional

disparities This is possibly the case in Vietnam in the 1990s The research hypotheses

associated with this pattern of growth are:

Hypothesis H-3: Regional and provincial disparities in social-economic development in Vietnam have increased in the year 2000 compared to 1990

Hypothesis H-4: The rural-urban gap in development has increased in the year 2000 compared to 1990

This group of hypotheses is aimed at examining Kuznets and Williamson’s thesis about

the initial widening gap in development, which states that there are large and growing regional disparities at early stages of development and that, in the long run, regions’

development indicators tend to converge —faster in the developed economies and slower -

in the less developed ones This argument is consistent with both Myrdal and

Hirschman’s economic development theory Generally, Myrdal and Hirschman, and - neoclassical economists’ such as Kuznets (1955) argument about short-term widening gaps between regions appears to be well supported by several empirical studies

(Williamson 1965); but their long-term regional convergence proposition is debatable

c As was noted previously, the relationship between city and region and managed urbanization in developing countries remains debatable Many studies have shown how big cities in the developing world have literally a parasitic effect on the regional and national economy (Friedmann 1966: 35) Yet other studies saw no evidence of any

Trang 32

economic development within a region or country (Alonso 1968) This argument is also reinforced by the fact that economically advanced regions are generally associated with one or several central cities that also play a vital role in the national economy and national

urban system (Cusack 1999: 38)

In any case, development in the contemporary perspective places a heavy focus on

the shift from the primary sector to the urban-industrial sector, i.e., the secondary and tertiary sectors At the early stage of growth, these sectors tend to round themselves in core cities due to the agglomeration effect and economies of scales and in the regions that

can provide agricultural surpluses to support industrial development and trade In

consequence, the core obtains a new development momentum —often at the expense of the

region and rural sector

With respect to Vietnam as a small developing country with relatively well-

developed internal complementarities, a highly hierarchical urban system and

accompanied economic regionalism may actually have negative consequences for overall

national development, at least at the current stage of growth when development options

such as import substitution may play an important role The research hypothesis

associated with the city-region relationship economic consolidation of the core areas is: Hypothesis H-5: Provinces and regions with a large capital city have

registered greater development compared to those associated with smaller

towns.”

Trang 33

be analyzed in relation with regional development, which is consistent with Gore’s (1984) framework Further, the urban system in Vietnam will also be examined in terms of its

deviation from the rank-size distribution rule and in terms of its primacy problem, and its

effect on regional and national development Comparison with other urban systems in the Asian-Pacific region will be made Generally, it is acknowledged that the rank-size regularities have been associated with the existence of an integrated system of cities, which is beneficial to both regional and national development The primate cities, on the other hand, have often been associated with less developed countries (LDCs)

3 Quantitative Methods of Analysis and Data Sources ‘3.1 Data Sources and Limitations

This research uses the secondary data source for 1990 and 2000 published by Vietnam General Statistical Office, VGSO, in the various editions of Vietnam Statistical

Yearbooks Most of the indicators in these Statistical Yearbooks refer to material

production and a few to social services such as education and healthcare It should be

noted thatthe first national census was conducted only in 1989 and funded by international institutions Collected data were gradually released in later Statistical

Yearbook editions, which implies that certain discrepancies exist in the number of items between various editions making comparison between two time periods more difficult

Some indicators like per capita income adjusted for purchasing power parity (PPP), infant mortality rate, and poverty rate are taken from the United Nation Development Program’s

Trang 34

Further, although Vietnam Statistical Yearbook is an official data source and is therefore generally more credible than other data sources, there are certain caveats and

limitations in using it; and some of these come from the methods of data collection and

some from item definition Most serious limitations for regional/provincial comparisons

are several Firstly, the data on income, sectoral employment, and exports/imports at the provincial level are lacking or very limited, scattered, and inconsistent Secondly, central government activities in the tertiary sector and social services such as healthcare, etc.,

which are quite sizable, are not broken down to the provincial level where these activities actually take place Another limitation is that adjustment mechanisms such as annual inflation indexes and implicit price deflators to compare time series data are not readily

available; and this makes comparison much less reliable

3.2 Quantitative Methods in Interregional Disparity Measurements

Either as a result of deliberate national and regional development policy or market forces’

effect, spatial differentiation in economic activities and resultant interregional inequalities can be measured using various univariate and multivariate statistical methods In this

study, because of their simplicity and ease of interpretation, univariate methods such as | the Gini coefficients and the coefficients of variation shall be applied Use will also be

_ made of multivariate methods such as factor analysis and regression analysis where it is _

Trang 35

uncover specifically where those disparities occur (Cole 1981; Anand 1983) In this study, therefore, they both are used for the purpose of spatial interpretation as well as international comparison

In either the method of the Gini coefficients, Gs, or the coefficients of variation, "CV, indicators like per capita income, per capita gross industrial and agricultural output,

per capita retail sales, etc are used singly Although the method of Gs is originally and

mainly applied in measuring income inequality among population strata, it can be adequately used in the similar manner in detecting inequality in any other indicator between localities, be it industry, agriculture, education, other social services, and natural

endowments (Cole 1981) By its construction, Gs fluctuate between zero and 1; Gs close to unitary indicate large inequality between spatial units in an indicator and Gs close to

zero indicate equality between them as measured by the indicator Comparing Gs between two time periods helps to determine if progress toward equality is made with respect to a given indicator: income, social services, or economic production

To measure the differentiation between localities in income and other indicators,

the coefficient of variation is also commonly applied (Williamson 1965; Bielasiak 1980;

Koschatzky 1988; Tsui 1990; Wang and Hu 1999: 43-44; Westlund 2000): The unweighted coefficient of variation, CVu, is:

` - ¬\2 n

CVu = yi ml _ Std.Deviation

Trang 36

The weighted coefficient of variation, CVw, is:

sứ

3i~Y y f | Weighted.Std Deviation

CWw =->—— y = Mean Moar

where fi = population of the ith region/province; f= total national population; n = number of regions/provinces; yi = per capita income of the ith region; and y = national per capita

income If the regional populations are approximately equal, the unweighted coefficient

CVu is close to the weighted coefficient CVw In this case, using either CVu or CVw

does not seem to make a big difference In practice, however, it is often the case that regions/provinces vary greatly in terms of population size, im which case CVw should be

applied (Williamson 1965; Wang and Hu 1999: 43-74)

£

CVw indicate how regions deviate from the national average: the higher the CVw,

the greater the size of geographic differentials, For instance, the dispersion of the regional

per capita income levels relative to the national average is measured while each regional deviation is weighted by its share in the national population Like Gs, examining CVw trend from the base year to the current year helps to determine if a progress toward equality is made with respect to any of the indicators: income, industrial or agricultural gross output, social services such as healthcare and education, etc

Strictly speaking, however, there is no sound theoretical study to show that the

method of Gini coefficients or coefficients of variation is more advantageous than other measures of regional inequality (Tsui 1990) Possibly, these coefficients have been

applied more often just because they would make international comparisons valid and

Trang 37

coefficients and coefficients of variation is that they use indicator(s) singly, one by one,

whereas development and a region’s well-being are multi-faceted Further, using the Gini ' coefficients one at a time and its magnitude to interpret the degree of unevenness in

development may have certain limitations And using per capita income to measure interregional inequality, for instance, may also have several drawbacks For one, per capita GDP may misrepresent the productive capacities and social service facilities of different regions Further, per capita income often does not fully reflect the extent of

polarized development in less developed countries in the modem world of globalized production, for within these economies there may operate enclaves or areas or even regions that have quite modern export sectors or industries and associated high-income population segments that are quite sizable in magnitude, if not in the share of the total

national population It also does not capture the activities in the informal sector that is

sizable in any developing economy Further, various empirical studies show that the

quality of human life in a region such as physical environment and life expectancy may not be highly correlated with its per capita income (Richardson 1977; Wang and Hu

1999: 79) |

To complement these shortcomings and to increase the probability of arriving at

more accurate conclusions about the real state of affairs, factor analysis is widely used to

treat a large number of indicators of development (Flohloff-Kulke 1988) In essence,

factor analysis is a method used to reduce a large number of initial and related indicators

to a relatively small number of main independent components, or factors, that account for

Trang 38

the original indicators in an intelligible manner (Gorsuch 1983: 2-4) Each of the resultant

factors can be quantified in the form of factor scores, which are a numerical value for each region or province and which can then be used directly to group spatial units or

indirectly as input variables in other types of statistical analysis such as regression and cluster analysis Factor analysis, in turn, is not without problems For instance, the number of main components extracted can be quite sensitive to the number and types of

the original variables The main components’ interpretation is therefore could be

problematic

Factor analytic techniques have been successfully applied in studying the spatial

distribution of many demographic, political, social, economic, and cultural processes

Jonassen and Peres (1960), for example, using Census data examined 82 community

variables from 88 counties in Ohio and found just 7 factors representing all of the original -

82 social-economic-demographic variables That is, the set of diverse indicators of a community or an area’s social-economic health and activity can be reduced to a new set

consisting of just some underlying factors This clearly helps policy makers understand and prioritize problems, and concentrate resources to solve the problems Another

empirical study worth mentioning is by Ohuallachain (1992) who used factor analysis to identify many recognized types of industry clusters in 150 largest SMSAs in the U.S.,

such as ports, petrochemical complexes, resort areas, and centers for traditional

manufacturing and resource processing industries Analyzing extensive data on U.S

Trang 39

manufacturing, and corporate administration as opposed to the traditional industry clustering under purely economic forces In short, factor analysis helps to derive underlying structure of regional development

3.3 Hypothesis Testing

a The unit of analysis in this study is province There are 61 provinces in Vietnam

grouped into 8 major economic development regions The number of cases, 1.¢., 61

provinces, seems to be proper for various types of statistical analysis such as factor analysis and regression analysis The provinces in Vietnam are, however, small in area

and population size compared to other regional economies For the international

comparison purposes, regions will also be analyzed (see Vietnam map at the end of this Chapter) Hypothesis H-1 on the national growth determinants will be examined

’ descriptively Specifically, indicators such as gross outputs in industry, agriculture,

service, private industry, foreign investment, etc will be examined to determine the contribution of each of these sectors to national economic growth Based on this national

growth pattern, Hypothesis H-2 then aims at specifying the growth pattern at the level of

individual provinces

b Regional uneven development can be measured based on comparing indicators such as per capita gross output of industrial and agricultural production, and service sector, and other indicators of social services such as number of physicians per thousand residents,

etc These indicators relatively dependably reflect the spatial distribution of physical

infrastructure, productive facilities, and social facilities These are readily available in the

Trang 40

attendant inequalities between provinces and regions (Hypotheses H-3 and H-4) will be detected by constructing the Lorenz curve, computing and comparing the Gini

coefficients and coefficients of variation for 1990 as the base year and 2000 as the reform

year Regional disparities can find reflection in many dimensions such as economic

development and human development Although a number of indicators can be used to

measures those disparities, income is one of the most commonly used indicators In the

absence of income statistics at the provincial level, provinces’ per capita gross output of

industry, primary sector, and retail sales will be used as a substitute for per capita income

In fact, these two indicators are strongly correlated, with Pearson’s bivariate correlation coefficient equal 0.96 in 1999 —the only year for which provincial income data are available -

As noted in the above, empirical analysis of regional growth and disparities in a - number of indicators can be conducted using factor analysis, which helps examine and

identify ageregated socioeconomic variables that underlie the differences between regions and provinces Factor analyzing a large set of variables in the year 2000 will complement the trend in regional disparities detected in Hypothesis H-3 In this study, 28 variables have been used in factor analysis These range from population characteristics (such as densities, level of urbanization, etc.) to principal economic activity (such as industry, agriculture, transportation, retails, services, foreign investment, etc.) to social

development indicators (such as numbers of students, physicians, hospital beds,