Ho Chi Minh City:

The Challenges of Growth

David Dapice
Jose A. Gomez-Ibanez
Nguyen Xuan Thanh

Prepared under UNDP – Harvard Policy Dialogue Papers

“Series on Vietnam’s WTO Accession and International Competitiveness Research”

Policy Dialogue Paper Number 2

Copyright © 2009 by United Nations Development Programme in Vietnam.
All rights reserved.
Publication License No.: 171-2010/CXB/17/02-01/VHTT
Photo: Duc Tri
Layout and design: Phan Huong Giang/UNDP Viet Nam; Phu Sy Printing Co.
Printed in Viet Nam.

1
 Foreword

Viet Nam’s transformation over the past two decades from a least developed, transitional economy to a
middle-income country has taken place hand in hand with a process of rapid urbanization. The most recent
census data for 2009 places Viet Nam’s urban population at well over 25 million, accounting for 29.6 percent
of the total population, up by 40.4 percent from the previous census exercise of 1999. This process of
urbanization poses important challenges to policy makers in a wide range of areas, from the promotion of the
local economy, to the development of urban infrastructures and the provision of social services for its
residents, migrant or registered. Nowhere are these challenges more pressing than in Ho Chi Minh City, the
country’s largest city and its commercial capital.

This policy dialogue paper focuses on two main challenges that Ho Chi Minh City currently faces in
managing the rapid transformation of its urban landscape and population structure: traffic congestion and the
development of new urban areas. The paper examines how interventions in areas such as traffic
management, the improvement and expansion of public transport services and networks, transport
infrastructure development and urban planning can help local authorities in Ho Chi Minh City successfully
address these challenges in the coming decades. It also assesses their implications for public investment,
public financial management and decentralization policies in Viet Nam.

Although the focus of this paper is on the specific case of Ho Chi Minh City, the findings reported in this
study are also applicable to other expanding urban areas in Viet Nam. This obviously includes Ha Noi and its
surrounding areas, the second largest urban concentration in Vietnam after HCMC, but also cities such as
Hai Phong, Can Tho or Dan Nang, which are also experiencing an important transformation of their
urban landscapes.

While the ideas expressed in the paper do not represent the official views of UNDP, we hope that the
recommendations presented here will stimulate discussion and debate among researchers and policy
makers in Viet Nam and elsewhere on this important topic for Viet Nam’s development.

I would like to take this opportunity to thank the research from Harvard University’s John F. Kennedy School
of Governance’s Vietnam Programme for their insightful analysis and constructive policy recommendations. I
would also like to acknowledge the generous support received from the UK’s Department for International
Development (DfID) and the Spanish Agency for International Cooperation and Development (AECID) to
UNDP’s policy advisory work in Viet Nam, without which this paper would not have been possible.

Setsuko Yamazaki
UNDP Country Director

2
 Abstract

HCMC is the largest city in Vietnam and is the economic center of the country. Its official population in 2007
was 6.6 million, a 3.1 percent annual increase from 2002. However, the total population including migrant
workers may be as high as 8.7 million. Independent estimates based on enterprise employment, motorcycle
registration, and construction permit data also suggest an absolute increase of 400,000-500,000 people in
the city a year. Since infrastructure, beyond some minimum, should be related to population growth, the
current amount of major urban investment is lower than it should be. As a result, the city is facing two critical
and related challenges, namely, traffic congestion and development of new urban areas. These challenges
have prompted the city government to develop a modern transportation system, including expressway and
metro rail transit (MRT) systems, and a land use and urban development plan to accommodate the growing
population and the higher income lifestyles that city residents could increasingly afford.

Recognizing the fact that motorcycles are fairly efficient users of street space, even if not as efficient as
buses, the main priority in traffic policy is to discourage or delay the shift to the automobile. Bus use could
and should be increased, but the prospect for dramatic relief from this quarter is probably limited especially
now that HCMC residents have become accustomed to the convenience of a motorcycle. Nevertheless,
exclusive lanes for buses and bus rapid transit should be explored. Despite all of these measures, the city
must invest in added transportation capacity to accommodate its rapid growth. Both the MRT and
expressway systems are expensive so that it is unlikely that they will be completed in their entirety.
Therefore, it is important that the most valuable projects be built first with priorities being driven more by
objective cost and benefit analysis than by funding opportunities.

To accommodate the growing population, development of large-scale new urban areas is inevitable. The city
government must make sure that new urban developments be located on land appropriate for building,
avoiding low-lying wetlands. The historic core of the city in Districts 1 and 3 should be protected, so that the
city remains the Pearl of the Orient. Toward those ends, it is appropriate to develop satellite centers to
disperse central business district functions and reduce pressure on the historic neighborhoods. However,
these developments are not financially viable given the extremely high cost of land. When land costs tens of
thousands of dollars for even a modest official plot, it puts ownership of urban housing beyond all but a very
few. Hence, the government should seek to bring down the price of land through more transparent
conversion of agricultural land into urban land at low transaction costs, and introduction of a local property
tax system.

Finally, HCMC should be allowed to retain a larger portion of its tax revenue to finance its new transport and
urban development projects. A new division of tax revenues between rich and poor provinces is needed to
reflect this. However, simply allowing richer provinces to keep a larger share of their tax payments would not
be productive without major enhancements to project planning and implementation.

The views expressed in this paper are those of the authors alone and do not necessarily reflect those of the
United Nations or UNDP, the Fulbright Economics Teaching Program nor the Harvard Kennedy School.

3
 Table of Contents

Foreword
Abstract
Introduction .......................................................................................................................... ….. .................... 1
The Importance of Cities ............................................................................................................ .................... 2
The Actual Population and Population Growth in the HCMC Area ............................................ .................... 3
The Two Key Challenges ........................................................................................................... .................... 6
HCMC Transport: Avoiding Gridlock on the Streets ................................................................... .................... 8
Development of New Urban Areas............................................................................................. .................. 14
Urban Finance ........................................................................................................................... .................. 17
Conclusion ................................................................................................................................. .................. 21
References ................................................................................................................................ .................. 22

List of table
Table 1. Growth in Gross Regional Product, HCMC Province and Vietnam................................................. 23
Table 2. Population and Motor Vehicle Registrations in Ho Chi Minh City ................................................... 23
Table 3. Percentage of Trips in Ho Chi Minh City by Mode Used, 1996 and 2002 ....................................... 24
Table 4. Demand Forecasts for 2025 with 161 Kilometer Rail Transit Network ............................................ 24

List of figure
Figure 1. Map of Ho Chi Minh City ............................................................................................................... 25
Figure 2. Map of Ho Chin Minh City’s Inner Districts .................................................................................... 26
Figure 3. Plan for Four Elevated Expressways ............................................................................................ 27
Figure 4. MRT Plan as of 2008 .................................................................................................................... 28
Figure 5. Location of Business Centers in 2008 Master Plan ....................................................................... 29

1
 Introduction

Vietnam faces important challenges in managing the explosive growth of its urban areas, and nowhere more
than in Ho Chi Minh City (HCMC), the country’s largest city and its commercial capital. Cities are central to
Vietnam’s economic success, since they are the locus of a disproportionate share of the country’s economic
growth and activities. But the growth of HCMC has pushed up real estate prices to high levels, making
commercial and residential space less affordable for businesses and households. And development
pressures are threatening the historic French colonial core of the city and the wetlands to the east and
southeast. Traffic congestion is increasing as the city is registering 1,300 motorcycles and 150 cars every
day. The city, which the French had christened the “Pearl of the Orient”, is now referred to by some as the
“motorcycle capital of the world.”

According to official government statistics HCMC’s population grew at an average rate of 3.5 percent per
year over the last decade, reaching 6.6 million by 2007. The official statistics seriously understate HCMC’s
growth, however, since as many as 2 million migrants are thought not to be accurately registered with the
city. Rural provinces do not have an incentive to report their population losses accurately since the central
government redistributes tax revenues from wealthy provinces to poor provinces using a formula that takes
population into account. And HCMC, as a wealthy area, is not highly motivated to report all the population it
is gaining because the added services it would have to provide to the migrants might cost more than the
additional tax revenues it would be allowed to retain. As a result, infrastructure investment, hobbled by
transfers of 70 percent of tax collections to the central government, is completely inadequate to prevent
severe congestion that will slow growth. While equity and political bargaining will always dictate some level of
subsidy from richer to poorer areas, the level of transfer (or of borrowing for poorer regions) should not be so
high as to strangle the growth in the most dynamic regions.

In March 2008, the People’s Committee of HCMC approved a revised master plan designed to guide the
development of the city through 2025. The master plan designates areas where growth is to be encouraged
and includes a list of transportation and other infrastructure projects and policies designed to support the
desired developments. Not all of the elements of the plan seem consistent with its goals, however, and the
list of infrastructure projects is so ambitious—including four elevated expressways and six mass rapid transit
(MRT) lines—that it is very unlikely that all will be built on schedule. Moreover, there is little reason to believe
that the current plan will not suffer the fate of its predecessor, which was never implemented in part because
the agencies responsible lacked the technical capacity and the political support.

HCMC must make some difficult choices about what policies and projects it should pursue. The city will
never be able to expand expressway or MRT capacity rapidly enough to keep pace with the explosion of
auto ownership and use that is likely to begin soon. As a result, city leaders must think carefully both about
how to allocate existing street capacity between autos, motorcycles and buses, and about which of the
expressways and MRT lines will be most effective in reducing congestion. At the same time the city needs to
find a way to accommodate the strong pressures to expand the downtown and develop new urban areas
without compromising significantly on the goals of protecting the historic center and preventing flooding.
Finally, the city must strengthen the capacity of its government agencies to administer plans and policies.

1
 The Importance of Cities

HCMC and Vietnam’s other cities are growing because of the economic opportunities they offer for residents
and migrants. Cities exist because businesses and workers are more productive when they are located in
close proximity. Economists call these gains from co-location agglomeration economies, and they arise in a
variety of ways (Duranton and Puga, 2004). When firms from the same industry locate together they can
better support specialized input or service providers or make customer search easier, for example, and when
firms from different industries congregate they may benefit from knowledge spillovers transference and from
a larger and more diverse labor pool. The higher productivity of urban labor and capital results in higher
incomes, and thus it is no surprise that as a country develops it also urbanizes.

Vietnam is no exception—while agricultural output and productivity have increased significantly in the two
decades since Doi Moi, urban output has risen even faster. The Vietnamese economy has grown at an
average annual rate of roughly 7 percent over the last decade, but HCMC’s economy has grown at an
annual rate of around 11 percent, or roughly 50 percent faster (See Table 1). The concentration of
production in Vietnam is vividly illustrated by realized foreign direct investment (FDI). The area in and around
HCMC has over 55 percent of all FDI, while the area in and around Hanoi-Haiphong has another 25 percent.
1
So, over 80 percent of realized FDI through July 2007 is in and around major cities. Of course, FDI is only
part of total activity. If 2005 enterprise employment is considered, taking all types of enterprises, then the
southern urban group has about 40 percent of national employees and the northern urban group has 21
percent - so, in total, over 60 percent of all enterprise employees are in and around the major cities. Similar
ratios are observed for exports, capital and tax payments.2 In 2006, the official population of these two
regions was 23 million or 27 percent of national population, split evenly between the north and south.

This concentration of economic activity is reflected in the movement of people. Vietnam’s urban population is
growing at an official 3.6 percent a year (1999-2005) but probably closer to 5 percent if unofficial urban
migrants and urban sprawl into “rural” areas adjacent to cities were included.3 The proportion of people who
should be counted as urban is certainly over 30 percent and may be closer to 35 percent, while the official
count is 27 percent.4 With total national population growing just over 1 million a year, urban and near – urban
population growth are probably now absorbing all population growth and some rural migrants each year.
Given the low returns in agriculture and the tendency for younger educated workers to prefer non-farm jobs,
it is likely that rural population is already declining as the formal sector continues to expand – and largely
in urban areas.

But if there are advantages to concentrating activities in cities there are also costs, mainly in the form of
congestion, pollution and competition for scarce centrally located sites. One of the most important functions
of city government is to reduce or ameliorate these costs through the provision of infrastructure and land use
planning. Roads and public transportation systems can reduce the congestion caused by large numbers of
people living and working in close proximity, while drinking water, waste water and drainage systems can
reduce the pubic health and flooding risks of concentrated activity. Thoughtful land use planning can
contribute as well by locating activities on sites and at densities which generate substantial agglomeration
benefits while keeping the congestion and pollution costs manageable. Housing, manufacturing and offices
generate different levels of congestion and pollution, for example, and finding the appropriate match between
uses and sites can be important since some sites are easier to service than others.

1
The HCMC area includes Ba Ria-Vung Tau, Dong Nai. Binh Duong and Long An. The Hanoi-Haiphong area includes Hai Duong, Ha
Tay, Bac Ninh and Vinh Phuc.
2
In 2000, these provinces had 53% of national GDP. The proportion is surely higher now.
3
For example, the 2004 Living Standards Survey found over 70% of workers in the Southeast (HCMC area) had nonfarm jobs,
compared to 84% for urban and 33% for rural. The ratio has risen since then.
4
In China, the official urban count based on registration cards was under 30%, but was moved up to 40% when actual residence was
used. In Vietnam, the share of labor force in farming and fishing fell from 65% in 2000 to 55.7% in 2006, caused by a drop of over two
million moving out of farming, while the total labor force grew by almost six million workers. Non-farm activities absorbed all growth in
the labor force plus the 2.3 million who stopped farming. (Statistical Yearbook of Vietnam 2006, pp. 50-51)

2
 The Actual Population and Population Growth
in the HCMC Area

HCMC is the largest city in Vietnam and is the economic center of the country. Its official population in 2007
was 6.6 million, a 3.1 percent annual increase from 2002. However, it is widely believed that the actual
resident population is higher than that officially counted. The “floating population” might live and work in
HCMC for most of the year, but be counted as coming from their home province. It is no easy matter to count
these people as they often live in “illegal” housing on agricultural land. This section discusses several
alternative approaches to estimating the actual population and its growth in HCMC.

Enterprise Employment vs. Population Growth

The key to a better estimate is to identify independent data sources that seem to be more reliable than the
official population estimates. One such possible source is the annual Enterprise Surveys. Between 2002 and
2005 the official population of HCMC grew by 7.5 percent while the number of workers employed in
enterprises increased by 39 percent:

In millions of people
Official Population Enterprise Employees

2002 5.48 1.08

2005 5.89 1.50

% Change 7.5% 39.0%

Absolute Change 0.41 0.42

If both the population and the employment statistics are correct then the proportion of HCMC residents with a
job in the formal sector increased from about one in five in 2002 to one in four in 2005. This is quite a
dramatic shift to occur in only three years, and it seems more likely to an artifact of under reporting
population growth. If the new jobs in the formal sector were taken by workers migrating to HCMC rather than
by existing residents switching to the formal sector, moreover, then the official figures imply that all of the
reported population growth came from new workers migrating without their families. This later possibility is
highly improbable given that part of the population increase must have been due to births to existing
residents and that at least some workers are likely to migrate with their families. An enterprise employee
normally has a spouse and/or dependents who are often not formal sector employees – the
population/formal employee ratio is similar to that of family size. If all the new employees were migrants who
were accompanied by an average of only two additional family members each then the population should
have increased by 1.26 million persons over the three year period rather than 420 thousand. This seems like
a high growth rate, but it is also quite possible that the 2002 actual population was higher too. If the actual
2002 population started at 6.5 million, for example, the 1.26 million increase would translate into a population
growth rate of about 6% a year instead of the reported 2.5% per year. It would also suggest a 2007
population in the city of 8.7 million! If migrants were accompanied by an average of three family members
that would imply a population increase of 1.68 million during 2002-05 and nearly 8% annual growth (using a
2002 population of 6.5 million as the base).

Motorcycles

Motorcycles are very popular in HCMC and almost a necessity for getting around, though buses, cars and
bicycles are also used to some extent. In the 2004 Living Standards Survey in the Southeast there were
reportedly 71% of all households owning motorcycles. The Houtrans person-trip survey in 2002 showed that
more than 90% of HCMC households owned motorcycles and that 53% owned two or more motorcycles
(JICA 2004). According to the HCMC Department of Transport and Public Works, 1,300 motorcycles a

3
working day are being registered and added in the city. If we assume 1.5 people per motorcycle (children
and some spouses do not own them; a few household heads may not use them), that would imply 487,500
new people each year being added. If there were 8.7 million people, this would imply 5.6 percent a year
population growth in HCMC. This is an upper bound for the population growth rate in as much as the number
of motorcycles per capita is increasing in HCMC and some older motorcycles are presumably retired or sold
to surrounding provinces.

Registered Construction

Newly constructed private homes and apartments have to get a construction permit in order to be built. Of
course, some smaller units on agricultural land may not get a permit, but most units do. There is a rule of
thumb that there are 10 square meters per person, although experts expect this to rise over time to 12-15
meters per capita, at least for legal residents.5 The official registered permits for new home/apartment
construction (excluding major repairs and maintenance) in recent years have been6:

2002 2003 2004 2005 2006 2007

Million square meters
1.79 2.72 3.35 4.75 4.71 5.73
of permits
Implied increase in population
179 272 335 475 471 573
(in thousands)

It is true that some of the permits for new home construction are actually for upgrading, and living space per
capita tends to increase over time. However, these unobservable effects are offset by the unofficial units and
unregistered expansions, which are also unobservable. According to the above table, the increase in implied
population from the end of 2002 to the end of 2007 is 2.1 million or 425,000 people a year on average.

Comparing Results

This attempt to use other information to estimate population growth has resulted in each independent
estimate yielding higher population growth numbers than the official data. The official data state that
population growth in HCMC was only 198,000 a year from 2002 to 2007. Let us compare the findings for
HCMC annual population growth from 2002 to 2007 using alternative methods:

Official data: 198,000

Enterprise data: 400-500,000 (at a ratio of 1 to 2 or 1 to 3 enterprise jobs to population)
Motorcycles: 487.000 (at 1.5 people/motorcycle; good for recent years only)
Construction: 425,000 (at 10 meters square per capita excluding informal)

The surprising implication of this exercise is that the official data are missing over half and perhaps as much
as two-thirds of the actual population growth.

The 2006 Population Change, Labor Force and Family Planning Survey found that out of 13.3 million people
counted as living in the Southeast, there had been 588,000 births from 2003 to 2006. That is 196 thousand a
year or a birth rate of 1.47% [births/population a year]. If the population of HCMC were 7 million, then we
would expect to observe 103,000 births a year. The crude death rate is about 4.3 per 1000 or for seven
million a projected 29,000 deaths, giving a natural net increase of only 74,000. In addition, the same survey
found net migration in the twelve months to April 2006 of 142.5 thousand into the Southeast. If HCMC were a
bit more than half of that – say 75,000 – then the “official” net natural increase and recorded migration would
match the official population increase fairly well. So, the statistical net is consistent with respect to observed
births, deaths and migrants as compared to official population growth.

However, consistency is not the same as accuracy. If our three alternative estimates have any validity, the
official statistical net is particularly poor at picking up “informal” or “floating” workers and their families. It

5 2
According to the HCMC Department of Construction, the average housing area per capital in HCMC in 2008 is 13.1 m .
6
HCMC Statistical Office, HCMC Statistical Yearbook 2007.

4
would also mean that infrastructure investment in these heavy migrant inflow areas is grossly inadequate –
especially given the dual redistribution of remittances to home provinces and tax transfers to deficit
provinces. Actual urban growth of 5-7% a year instead of 2.5% to 3% has many implications for land policy,
road and water and power investments, and school and clinic buildings and staffing. It raises serious
questions about the current pattern of transfers of tax revenue from major cities and fast-growing provinces
to other regions.

The estimates so far have only analyzed HCMC and not the surrounding provinces. It is likely that many of
the same issues arise, to a greater or lesser degree. One informal source said that while the official 2006
population in Binh Duong was 964 thousand, the province believed that actual population was about 1.4
million, or 45% more than officially counted. (If HCMC had 8 million actual inhabitants in 2006, it would have
been 31% over its official population.) If the four thriving adjacent provinces had 30% more than their official
population, they would have not the official 5.5 million but 7.2 million in population. This is of course merely a
guess, needing much more serious effort to make a secure estimate. However, if there were not the official
11.6 million in these provinces in 2006 but instead an actual 15.2 million, that would mean that planning had
“missed” over 3.5 million people or nearly one-quarter of the possible actual population, and 4% of Vietnam’s
total population! Even if the actual population is lower than hypothesized here, it is very likely that there is a
large understatement of actual population in the region if official statistics are used. This would also suggest
a significant underinvestment in infrastructure, since most infrastructure investment relates to population and
output growth rather than levels.

At this point, it is speculative to estimate the region’s population level and growth. However, if we believe the
provinces around HCMC (excluding Tay Ninh) also had 30% more population than officially counted in 2002
(that is, 12 million instead of 9.2 million) and if they were growing at 6%, then by 2007 their population would
be 16 million, or four million growth in five years. The entire national population growth from 2002 to 2007
was 5.5 million. This would imply that nearly three-quarters of national population growth was concentrating
in these provinces. Of course, this calculation is no better than the assumptions it is based on, and they are
speculative. If the 2002 population were only 20% higher than officially recorded (11 million instead of 9.2
million) and if population growth was only 5%, the actual population gain would have been three million, or
55% of the national population growth. The region’s actual share of national population growth over the last
five years is probably between 55% and 75%.

5
 The Two Key Challenges

HCMC’s rapid growth has lead to two main policy initiatives to develop master plans for the city’s land use
and infrastructure. The first planning effort was initiated in 1994 with financial support from the World Bank
and produced an urban master plan for the year 2020, which was approved in final form by the Prime
Minister in 1998. The 2020 plan proved ineffective, however, in part because it provided only general
guidance on where growth should be encouraged and lacked the specifics on permissible land uses and
densities. This defect of vagueness was reportedly compounded by weak and confused implementation. The
city’s Office of the Chief Architect and the Department of Land and Housing, the two agencies nominally
responsible for implementing the plan, had few qualified staff. And the land permitting process was very
cumbersome, and involved controls at both district and city levels. Lacking better guidance and control,
development was often opportunistic, where real estate promoters and city officials believed the profits would
be highest and resistance lowest.7 Within a decade of the plan’s completion it was already obsolete. Some
areas of the city already had more population than was forecast for 2020, while other areas that were slated
for growth were lagging.

The development that occurred during this period heightened concerns about two challenges. One challenge
is traffic congestion, which increased rapidly because of an explosion in motorcycle ownership. The number
of motorcycles in Vietnam grew steadily during the 1990s but this pace increased around 2000 (Table 2).
During this period not only were incomes growing rapidly but the government reduced the import tariffs on
motorcycles and the Chinese began to export inexpensive motorcycles to Vietnam. Prices for a 100-cc
Japanese motorcycle fell by roughly half, from over $2000 to a little over $1000, while one could buy a new
Chinese motorcycle for as little as $500. Between 2000 and 2007, the number of motorcycles in the country
8
tripled from 6.4 million to 20.2 million , equivalent to a national ratio of one motorcycle for every 4 people.
HCMC, being more affluent, had 3.4 million motorcycles by the end of 2007, or one for every two people. As
early as 2002, 92 percent of all HCMC households owned one or more motorcycles, 2.6 percent of
households owned cars, and only 5 percent of households were entirely dependent on buses or bicycles.9

A second challenge that emerged during this period is the development of new urban areas. HCMC has
approximately 1 million housing units and flats in urban areas and about 140,000 houses in the suburb.10
And this housing stock has to support 8.7 million people as estimated above. Currently, there is a conflict
between the need to expand the city’s downtown and the desire to protect the historic French colonial
neighborhoods that surround it. Much of the city’s population and job growth has been on the city’s
periphery, particularly to northwest of the city center in the Tan Binh and Go Vap districts near the airport, but
also to the west near the Binh Tan district and to the north around the Thu Duc district. However, developers
have begun to build office towers and residential apartment buildings in the downtown, on the west bank of
the Saigon River in District 1. The fact that Vietnamese economy is growing so rapidly and that HCMC is
Vietnam’s commercial capital has stimulated a surge in demand for downtown office space. Meanwhile the
cumbersome and complicated land use permitting system and the fact that some of the best downtown sites
are occupied by government agencies has slowed the response of property developers. By 2007 the
mismatch between demand and supply had driven the annual rents for class A offices—modern offices of the
kind that multinational and large domestic corporations insist on having—to more than US$400 per square
meter, a figure comparable to New York City and major cities of continental Europe. And the prospect of
such high rents is increasing pressures to expand the downtown into the surrounding historic neighborhoods
of French-era villas and shady tree-lined streets.

To accommodate a growing population and higher incomes without encroaching on the historic
neighborhoods around the downtown, the city government is developing an ambitious housing development
plan with several new urban areas in the periphery (Figure 5). However, these efforts are thwarted by the

7
For criticisms of the 2020 plan see Urban Planning Institute of HCMC and Nikken Sekkki (consultants), “The Study on the Adjustment
of HCMC Master Plan up to 2025”, report to the Ho Chi Minh City People’s Committee, 2008, pp. 2-23 to 2-34; and Huong Ha and Tai-
Chee Wong, “Economic Reforms and the New Master Plan of Ho Chi Minh City, Vietnam: Implementation Issues and Policy
Recommendations”, GeoJournal 49 (1999), pp. 301-309.
8
Ministry of Transportation, National Committee on Traffic Safety
9
HOUTRANS data as cited by Systra MVA, Ho Chi Minh City Metro Rail System Study, HCMC Master Plan Ridership and Revenue
Forecast Study, Final Report, January 2008, p. 8.
10
Nguyen Van Hiep, “The track records and challenges of housing problem in Ho Chi Minh City at present and up to 2010”, Mega City
Research Project, Ho Chi Minh City.

6
extremely high cost of land. While housing construction costs for basic housing are reasonable – about $100
to $150 per square meter – the costs of land in the range of $4,000-10,000 per square meter in the central
business district (i.e. Districts 1 and 3), $2,000-4,000 in other inner city districts, and $850-4,000 in the newly
developed eastern and southern peripheries11 are so high that housing costs are well beyond the reach of
most Vietnamese. As result new housing projects that are being developed are extremely skewed towards
villas and luxury apartments serving high-income earners.

These two problems— traffic congestion, and expensive housing development —stimulated a second effort
to prepare plans for the city. The cornerstone is a land use master plan for 2025 drafted by the Japanese
planning firm Nikken Sekki and approved by the People’s Committee of HCMC in March 2008. Two
complementary transportation plans were also prepared, one covering the entire transportation system, both
highway and rail, and a second detailing the plans for a 161-kilometer system of MRT and light rail transit
lines. The preamble to the new 2025 land use plan establishes several laudable objectives. The plan should
accommodate the growing population and the higher income lifestyles that city residents could increasingly
afford. New developments should be located on land appropriate for building, avoiding low-lying wetlands.
The historic core of the city in Districts 1 and 3 should be protected, so that the city remains the Pearl of the
Orient. And the plan should be more specific and implemented more conscientiously. But the new plan
leaves a number of key transportation and land use issues unresolved. These are examined in the sections
that follow.

11
VietRees, Vietnam Real Estate Weekly Newsletters, 2009.

7
 HCMC Transport: Avoiding Gridlock on the Streets

Current Traffic Conditions

The 2025 plan places great emphasis on building MRT lines and expressways as an answer to congestion.
But, as we will explain later, even if the new rail and expressway systems are built as planned most of
HCMC’s passenger trips will still be made on the existing streets, in motorcycles, buses or cars. Thus getting
the most out of the existing street system is essential if HCMC is to avoid gridlock.

Any effort to deal with traffic congestion must cope with the city population’s heavy reliance on the
motorcycle and the pending explosion in car ownership. HCMC differs from most other large metropolitan
areas in developing Asia in relying so heavily on the motorcycle and so little on public buses. The dominant
mode of urban transportation changes as economies develop and per capita incomes increase. The usual
trajectory is from human or animal powered forms of transportation—such as walking, bicycles or ox carts—
to motorized public transportation—in the form of the bus or minibus—and finally to motorized private
transportation—usually in the form of the automobile. HCMC has largely skipped over the motorized public
stage, however, shifting directly from bicycles and walking to motorcycles. In other cities at similar states of
economic development the bus typically accounts for 50 to 60 percent of urban trips, walking and bicycles
roughly 20 to 30 percent, and private automobiles and motorcycles the remaining 20 or 30 percent. In
HCMC, by contrast, motorcycles account for 74 percent of all non-walking trips with the bus only 5 percent,
12
private auto 2 percent and most of the remaining percentage on bicycles.

Bus ridership is low in HCMC in part because the bus system essentially collapsed in the two decades after
reunification, so that there was little bus service available when Vietnamese incomes reached levels where
they could afford to give up their bicycles and turn to a faster and more comfortable mode of transportation.
Currently, HCMC’s bus services are provided by the Saigon Bus Company, a public enterprise with two
affiliates that operate about 900 buses, and by 28 privately-owned cooperatives which operate about 2,300
buses, many of them small size. This mixed public-private system is a legacy of the several private bus
companies and the many owner-operators of lambros (a three-wheeled motorcycle with a tiny bus body on
the back), bon banhs (a four-wheeled lambro) and mini-buses that existed before reunification. All public
transport operators were nationalized on reunification but services deteriorated and in 1988 the government
returned the vehicles that had belonged to small owner-operators to their original owners who were allowed
to offer services as cooperatives. Neither the public enterprise nor the cooperatives received much support
from the government, however, and bus service was extremely limited.13

Efforts to revitalize the bus system began in 1994 but have been only moderately successful. In that year the
government formed a joint venture between Australian investors and the Saigon Bus Company to provide a
skeleton network of four routes radiating out of the Ben Thanh, a large market in the center of the downtown.
A further impetus for reform came a few years later, when the 2020 HCMC master plan concluded that the
city had to develop a public transportation system. The Prime Minister endorsed the recommendation in
2000, and after two years of study the city’s Transport and Urban Public Works Services (TUPWS) office
implemented the “model bus” and “pilot route” reforms. The intention was to revitalize the bus system by
helping to finance the purchase of a fleet of roughly 1,300 full-size, air-conditioned buses for both the Saigon
Bus Company and the cooperatives, and by establishing “pilot routes” which the company or a specific
cooperative would agree to service in return for government subsidies. The pilot route contracts specify the
frequency of service to be provided and TUPWS uses a formula to calculate how much the service should
cost and the subsidy required. Approximately 1,500 buses operated on pilot routes as of early 2006. The rest
of the buses, mostly owned by cooperatives, operated on regular routes and received no subsidy. At some
point during the reforms the lambros and bon banhs were banned from the city.14 The model bus reforms
resulted in a six-fold increase in ridership in just five years, from 57 million trips in 2002 to 380 million trips in

12
Systra MVA, Ho Chi Minh City Metro Rail System Study, HCMC Master Plan Ridership and Revenue Forecast Study, Final Report,
January 2008.
13
Based on interviews between the authors and officials at HCMC’s Transport and Urban Public Works Services.
14
The reforms are described in more detail in MVA Asia Limited, “Technical Assistance for Consolidation and Development of a Bus
System in Ho Chi Minh City, Vietnam; Task 1 Report: Review of Status of Bus System”, March 2006. The pre-reform system is
described in Japan Bank of International Cooperation, “Urban Public Transportation in Vietnam: Improving Regulatory Framework” JBIC
Research Paper no. 4, December 1999, p. 18 on

8
2007 (Table 5). The buses’ share of trips in the city increased from 2 percent to only an estimated 5 percent,
however, because the city’s population and trip making rates were increasing as well.

Automobile registrations are beginning to pick up in HCMC, but so far only a small fraction of HCMC
households own an automobile and many of the automobiles operated are official vehicles. Private
automobile ownership is discouraged by heavy import duties and registration fees, which roughly triple the
landed price of an automobile. For example, a new Toyota Camry that might cost US$ 20,000 to land at the
dock costs nearly US$ 60,000 after taxes and registration fees.15 However, small Chinese cars are
beginning to be imported for around US$ 20,000, including import duties. And if the current rate of economic
growth continues it will not be too long before Vietnamese incomes reach the levels where auto ownership
takes off.

The Advantages of the Motorcycle

HCMC’s reliance on the motorcycle is more tolerable because they are fairly efficient users of street space,
even if not as efficient as buses. Traffic engineers measure the amount of street capacity different types of
vehicles require relative to the capacity required by a passenger car. Thus, for example, a bus in mixed
traffic typically needs between 2 and 4 passenger car equivalents (PCEs) of capacity. A bus requires more
street capacity than a car, not so much because the bus is physically larger but because it accelerates and
decelerates more slowly and is less maneuverable. A bus operating express service, without stopping to pick
up and drop off passengers, might have a PCE of 2 or less, for example. But a bus providing local service
might have a PCE of 4 or more because the bus often blocks other traffic as it pulls in and out of curbside
bus stops. However, an average bus can carry as many as 40 people, compensating for their high PCE ratio.
The number of PCEs required by a motorcycle has not been extensively studied but seems to depend on the
proportion of motorcycles in the traffic and the width of the traffic lanes. On streets used mainly by cars and
trucks and with few motorcycles, studies suggest that a motorcycle requires two-fifths or one-third of a PCE.
But on urban streets with a high proportion of motorcycles and lanes at least 3 meters wide, a motorcycle
requires as little as one-fifth to one-tenth of a PCE.16 In such conditions motorcyclists can often ride several
abreast in a lane and maneuver past a queue of cars stopped at a red light to wait in front of the stop line,
ready to take off more quickly than a car is able to when the light turns green. The municipal traffic engineer
of HCMC believes that a motorcycle requires about one-sixth of a PCE under local conditions.17

These PCE estimates imply that while it would be advantageous to shift motorcyclists to buses, it is more
critical to discourage motorcyclists from shifting to cars. For example, 100 people traveling on motorcycles
would require 13.8 PCEs of street capacity (assuming each motorcycle requires one-sixth of a PCE and
carries 1.2 passengers). The same 100 people carried on buses might require only between 4 to 8 PCEs of
street capacity (assuming 2 to 4 PCEs and 50 passengers per bus). But the same 100 people in automobiles
would require as many as 83 PCEs of capacity (assuming an average of 1.2 passengers per automobile).
The shift from bicycles to motorcycles in the 1980s and 1990s alone probably contributed little to street
congestion in HCMC because a bicycle probably takes up about as much street capacity as a motorcycle.
The main source of added congestion in the last few decades was probably the increase in the total number
of trips by either bicycle or motorcycle caused by the growth in the city’s population and per capita incomes.
A shift to buses would provide some margin to accommodate further population and income growth, but a
significant shift to automobiles would quickly overwhelm the existing street capacity creating gridlock.

The Difficulties of Promoting Buses

Bus use could and should be increased, but the prospect for dramatic relief from this quarter is probably
limited especially now that HCMC residents have become accustomed to the convenience of a motorcycle.
Motorcycles have a disadvantage in that riders are exposed to weather and pollution, and the danger of
traffic injuries is much higher. But the climate is relatively mild in Vietnam, and the motorcycle offers many of
the conveniences of a private automobile, such as on-demand, door-to-door service, and the ability to make
intermediate stops and carry packages and extra passengers. Many Vietnamese have more than one job
and children to drop off and pick up at school, and thus value the flexibility of a motorcycle.

15
New automobiles are subject to import, special consumption, and value-added taxes whose rates were 70, 50, and 10 percent
respectively. These three taxes are imposed in a cascading structure, for an overall tax rate of 180.5 percent
16
Ing Husu, Ahamad Frahan Mohd Sadullah, Nguyen Xuan Dao, “A Comparison Study on Motorcycle Traffic Development in Some
Asian Countries: Case of Taiwan, Malaysia and Vietnam”, paper prepared for the Eastern Asia Society for Transportation Studies,
October 2003, pp. 38-39.
17
Interview with Le Minh Triet, Traffic Management Department, Transport and Urban Public Works Service, People’s Committee of Ho
Chi Minh City, May 9, 2007.

9
TUPWS officials argue that expanding the role of buses requires increased financial support from
government and protection from congestion. The average bus fare (including transfers) is relatively
affordable at 1,800 VND (US$ 0.11) compared to monthly per capita incomes above 1.2 million VND (US$
75).18 Fare revenues cover less than 60 percent of the bus operators’ costs, however, with the government
paying the balance; and government subsidies have increased from 40 billion VND (US$ 2.5 million) in 2002
to 528 billion VND in (US$ 33 million) in 2007. TUPWS argues that the buses are already fairly heavily
loaded—carrying an average of 37 passengers per trip in vehicles with an average capacity of 50
passengers—so that more subsidies will be needed to carry more passengers, since expanding the role of
buses will require purchasing a significant number of new vehicles.

But protection from congestion is more important than government subsidy. Most arterial streets in HCMC
have two or three lanes in each direction. By law the curbside lane is reserved for slower moving vehicles,
which means motorcycles and bicycles, the inner lane (or lanes) is for faster moving cars and light trucks,
while buses can use any lane. (Heavy trucks are allowed on city streets only between the hours of 9:30 pm
and 6:00 am.) These laws are largely followed by cars and trucks, although in traffic jams it is often difficult
for cars turning onto a street to force their way into the inner lane. The motorcycles are less disciplined,
typically ready to enter the inner lane when there is a gap in the auto traffic.

TUPWS officials argue that the difficulties of maneuvering in curbside lanes filled with motorcycles slows the
buses greatly, adding to bus operating costs and reducing the attractiveness of buses to potential riders.
There are no statistics available on the relative speeds of buses and motorcycles in traffic, but the popular
impression is that a bus trip takes twice as long as the same trip on a motorcycle because the motorcycle is
more maneuverable, has better acceleration, and does not stop to pick up passengers. As a result, busses
tend to be used primarily for longer trips, where riding a motorcycle is too stressful. Thirty seven percent of
bus riders are students, some of whom are probably too young to drive or unable afford the expense of a
motorcycle.19

TUPWS experimented with exclusive lanes for buses to protect them from congestion nearly 10 years ago,
but only a short lane at a bottleneck in District 5 remains in operation. The traffic police were opposed on the
grounds that exclusive lanes might cause safety problems. But the more important objections came from the
general public, which protested that traffic congestion on the remaining lanes became substantially worse.
Motorcycles typically outnumber other vehicles by 10 or 20 to 1 on downtown streets, and thus forcing them
all into the inner lanes with cars and trucks is likely to significantly increase congestion unless many
motorcyclists switch to buses. And the improvement in bus services was apparently not enough to induce
many motorcyclists to switch.

Discouraging Auto Use

As important as promoting bus patronage may be, discouraging travelers from shifting from motorcycles to
private automobiles is more critical. The existing preventive policy is the high import tariff and special
consumption tax on automobiles. However, it should be noted that the taxes may raise the income threshold
needed to make car ownership affordable for most families, but eventually, as the economy grows, that
threshold will be breached. Moreover, new car taxes and registration fees directly affect only ownership and
do little to discourage those who own a car from driving in the rush hours on congested roads. As a result,
other measures need to be instituted.

A relatively simple and effective strategy the government could use to discourage automobile use during
congested periods is to maintain the current prohibition on automobiles in the curbside lanes. The
government of HCMC has proposed to allow automobiles to use the curbside lanes on the grounds that the
inside lanes are becoming seriously congested with the increase in auto traffic. But allowing automobiles in
the curbside lanes would only slow the bus and motorcycle travel times relative to those for automobiles, and
thereby accelerate the switch from buses and motorcycles to cars, with disastrous results for congestion.
Tolerating higher congestion in the inner rather than the curbside lanes is desirable to reduce the travel times
of motorcycles and buses relative to autos and thereby discourage further shifting to autos on congested
roads at peak hours. It will be easier to continue to keep cars out of the curbside lanes now then to remove
them from the lanes after auto drivers have come to rely on them.

18
Systra MVA’s “willingness to pay” survey revealed that on average commuters between District 1 and 2 would be prepared to pay
6,100 VND for a better transport mode
19
Systra MVA, Ho Chi Minh City Metro Rail System Study, HCMC Master Plan Ridership and Revenue Forecast Study, Final Report,
January 2008, p. 11.

10
Another strategy is to charge motorists a fee to drive in congested central areas during the peak hours, much
as Singapore and London do. These schemes need not be very complicated to administer as Singapore
demonstrated in 1975 when it required that motorists display a simple paper license on their windscreen as
proof that they had paid the special fee to enter the central area during rush hour. But such pricing schemes
require careful planning and are often controversial, especially if they are implemented only after many
travelers have already switched to cars. HCMC already has in place an effective scheme for limiting car use
in the central area—the prohibition on automobiles in the curbside lanes—and maintaining that strategy
probably would be far easier than shifting to a new policy of charging for downtown street use.

If HCMC is to continue to rely heavily on motorcycles it should try to reduce the associated safety and
pollution risks. Vietnam has a very high traffic fatality rate thanks in part to its many motorcycles. The
fatalities are probably disproportionately on rural roads, however, where speeds were higher. And the
government took an important step in December 2007 in enforcing a law requiring motorcyclists to wear
helmets. Virtually all Vietnamese motorcycles are 150-cc’s or less, since larger motorcycles require a special
permit and license, and the small size presumably helps reduce emissions and may cut injuries as well. But
the government should investigate the possibility of taking further steps to promote safety and reduce
emissions.

The Hopes for New Rail Transit Lines and Expressways

HCMC is growing so rapidly that it must invest in added transportation capacity even if it succeeds in
discouraging or delaying the shift to the automobile. The greatest challenge will be to provide capacity in and
around the already densely built up areas, and particularly the downtown. During the late 1990s and the
early 2000s donors funded a series of studies of individual rail transit lines and expressways, with little effort
toward establishing an overall plan or priorities.20 To its credit, the government commissioned local and
international consultants to prepare two transportation plans that were to be consistent with the 2025 land
use master plan. The first of these plans, completed in 2007, was a transportation master plan covering both
highways and mass transit. The second, completed in 2008, was a more detailed plan for a rail transit
system. Much of the attention has focused on the rail transit plan, which calls proposals for 161-kilometer
system including six new MRT lines plus a tram line and a monorail. But the transportation master plan also
calls for a roughly 40-kilometer network of four elevated expressways that approach and encircle the existing
center of the city. These expressways are inspired in part by the system of elevated highways that Bangkok
built in the 1990s.

Both the MRT and expressway systems are expensive, however, so that it is unlikely that they will be
completed in their entirety. The consultants for the transit plan offered an “order-of-magnitude” cost estimate
that the entire network of 161 kilometers could be built for US$ 9.7 billion, or an average of US$ 60 million
per kilometer.21 This figure is a little low by international standards, especially if it includes rolling stock and
equipment as well as civil works. Indeed the first line that is now under construction will reportedly cost twice
the estimate in the plan. The MRT cost estimates are also for capital expenditures only and exclude
operating costs. Most mass transit systems require significant and continuing government subsidies because
they cannot recover their operating costs, never mind their capital costs, from passenger fares.

The expressways are projected to cost around US$ 1.4 billion, or about US$ 30 million to 40 million per
kilometer, a figure that apparently excludes land acquisition costs. The proposed alignments are largely over
existing streets and canals and the government hoped that the takings required would be minimal. But in
some places the roads are too narrow and the canals too serpentine to avoid significant land takings. The
government has experienced substantial delays and cost overruns in previous infrastructure construction
22
projects, which further fuels concerns that these estimates will prove optimistic.

Even if these new systems are completed on time and on budget, moreover, they will not provide enough
additional capacity to make it unnecessary to manage the existing streets carefully. The 2025 mass transit

20
Among the studies were the “HOUTRANS” study funded by the Japan International Cooperation Agency, the “UMRT Line 1” study
funded by the Japan Bank for International Cooperation, the 2003 and 2005 “Feasibility Study for the Two Priority Lines in the
Metropolitan Rail System (METRAS)” study funded by the German government, the “MRT Lines 1, 2 and 3” study financed by the
Russian government, a tramway study financed by the French government, a 1998 demand model financed by Britain’s Department for
International Development, and a bus rapid transit study sponsored by the World Bank.
21
Forty-eight kilometers would be in tunnel (at an estimated cost of US$90 million per kilometer) while 98 kilometers would be elevated
(at US$ 50 million per kilometer) and only the 15 kilometers of tram at grade (at US$ 20 million per kilometer).
22
Institute for Economic Research and United Nations Development Program, Paying for Urban Infrastructure and Services: A
Comparative Study of Municipal Finance in Ho Chi Minh City, Shanghai and Jakarta, UNDP, Hanoi, June 2007, pp. 16-18.

11
plan projects that the proposed 161-kilometer rail system will increase public transportation’s share of trips
carried to 44 percent, a dramatic improvement over the current 5 percent share. The consultants who
produced these forecasts note, however, that only 21 percent of the public transportation trips would be by
rail. If these figures are correct then 87 percent of all trips would still be made on the existing streets in
busses, cars and motorcycles, although the share presumably would be lower on the approaches to the
downtown (Table 4). Moreover, the consultants note that substantial improvements to the bus system, strong
policies to discourage the growth in auto ownership, the efficient integration of bus and rail services and the
concentration of development around rail corridors would all be needed to achieve the forecasts. Indeed in
their supplementary report on ridership, the consultants warn that assumptions about bus speeds, parking
charges and auto ownership “have been have been adjusted” to conform to the master plan’s target of a 45
percent public transport share by 2025.23 With a continuation of current policies toward bus priority on
streets and auto ownership, bus and rail’s share would increase to only 22 percent of all trips and 91 percent
of all trips would be carried on the city’s streets.

Furthermore, the elevated expressways will provide relief to some parts of the street system but increase
pressures on others. The four expressways are designed to relieve four congested radial arterials on the
northern, western and southern approaches to the downtown as well as to provide a bypass around Districts
1 and 3 (Figure 3). The effects of these new roads depend heavily on their exact alignments and on the
locations of the on and off ramps, which are still being determined. The bypass is likely to relieve pressure by
on downtown streets by providing an alternate route for through trips but the radial spokes are likely to
encourage more motorists to drive into the downtown, dumping additional traffic on streets that are already
highly congested. And allowing the expressways to extend beyond the bypass and penetrate the downtown
would threaten serious damage to the historic districts surrounding the downtown.

In this context, if in the end only some of the proposed rail lines and expressways are likely to be built, it is
important that the most valuable projects be built first. Unfortunately, implementation priorities are more
likely to be driven primarily by funding opportunities rather than by transportation needs. Each MRT line is
associated with a donor country which has sponsored preliminary studies of the line, for example, and the
order in which the lines are built is likely to be dictated by the order in which the countries are willing to offer
concessionary financing. The first rail line under construction is the eastern half of Line 1, which runs from
the downtown at Ben Thanh market across the Saigon River to an area in the northeast where the
government is attempting to attract development with a new amusement park, an industrial park and a new
site for the university. Constructing the western half of Line 1 would have brought more immediate
transportation benefit since it connects the already heavily built up northwest of the city with Ben Thanh. But
the eastern half was the segment that the Japanese had studied and their government offered a US$ 1
billion loan to build it, although reportedly the cost estimate has doubled to US$ 2 billion. The government
agency responsible for the rail system is hoping that the western half of Line 2 will be built next, since it has
been studied by the Germans and offers more immediate congestion relief by connecting the built up south
western part of the city with Ben Thanh. The estimated cost of this segment is US$ 1.2 billion, however, and
the German government has offered only US$ 400 million in low interest loans, enough to cover rolling stock
and equipment (which German firms would supply) but not the civil works. The government has approached
the Asian Development Bank for the balance.

The possibility that funding opportunities may distort transportation investments is even more likely with the
elevated expressways. The government signed memoranda of understanding in 2007 and 2008 with three
big construction firms from South Korea, Malaysia and Vietnam to develop feasibility studies for the four
expressways. The construction companies are to propose a final alignment and financing scheme for each
road. The onset of the world financial crisis shortly after the signing made financing more difficult and
property values more uncertain, thereby delaying the construction companies’ proposals. But most
observers expect that the roads are unlikely to generate enough toll revenues to recover all their construction
costs, and that the construction companies will compensate by requesting additional land from the
government along the right-of-way to develop for housing or commercial uses. It is possible that this
procedure will generate imaginative proposals that coordinate the development of the expressways with
adjacent parcels. But it is also possible that each company’s desire to maximize profits from its expressway
and parcels will lead them to ignore undesirable impacts on others. And the lack of some kind of competitive
process for soliciting and reviewing the proposals raises will make it harder to judge whether the proposed
schemes are as beneficial as they might be.

23
Systra MVA, “Ho Chi Minh City Metro Rail System Study, HCMC Master Plan Ridership and Revenue Forecast Study, Final Report”,
January 2008, p. 36.

12
Finally, given likely funding constraints it is unfortunate that the 2025 transportation plan does not include the
relatively low cost option of bus rapid transit, or BRT. BRT systems attempt to provide service comparable in
quality to MRT lines by operating high frequency buses on exclusive bus lanes and stations. In many BRT
systems boarding and dwell times are reduced by collecting fares in stations before boarding, raising the
station platform to the level of the bus floor and buying buses with multiple doors. The World Bank pushed
Vietnam to consider BRT as a cost-effective alternative to MRT and arranged funding for a study to identify
pilot BRT lines in HCMC and Hanoi. The suggested line in HCMC was a little north of the alignment of the
western half of MRT Line 2, running 17.5 kilometers from Ben Thanh market northwest to the airport and
then beyond to An Suong. The estimated cost was US$ 58 million, which included US$ 7 million for a fleet of
30 specially designed buses. On most of the route the streets were wide enough to fit in BRT and still
maintain two lanes in each direction for general traffic. When the line approached the downtown, however, it
had to use some streets that had only two traffic lanes in both directions. The planners proposed a one-way
loop that would maintain a lane for local access by general traffic but the city rejected the idea as
too disruptive.24

24
MVA Asia Ltd, “Bus Rapid Transit”, task 2 report of “Technical Assistance for Consolidation and Development of a Bus System in Ho
Chi Minh City, Vietnam”, March 2006.

13
 Development of New Urban Areas

One of the most important dilemmas in planning for HCMC is where to expand the central business district
functions currently located in the downtown on the west bank of the Saigon River. Finding additional room for
these activities is important because the city is Vietnam’s commercial and financial center. The need for
financial and business services is growing rapidly as Vietnam’s economy grows. And firms providing
business and financial services typically locate together in the centers of metropolitan areas because they
benefit greatly from close contact with one another. The fact that the rents for modern offices in downtown
HCMC are currently very high by international standards is testimony to the intense demand for and the
limited supply of centrally located offices.

Up to now, only one new urban area has been developed in HCMC. That pioneering project was Saigon
South New Town (also known as Phu My Hung), a 600 hectare residential and commercial development built
on wetlands south of the Te Canal in District 7. The government leased the land in 1996 to a Taiwanese real
estate developer in return for a 30 percent interest in the development. Few believed that much would come
of the land deal because the site was so difficult to build on and was located 5 kilometers south of downtown.
But the developer was experienced and well financed, and gradually built an extremely high-quality master
planned community with over 25,000 residents that became one of the most prestigious residential areas in
HCMC.25 The success of Saigon South has encouraged local developers to build on wetlands nearby.
Much of the land in the area is vacant and owned by the city government, which receives revenues from
leasing the land to developers. (From 2001 to 2004 roughly 15 percent of the HCMC government’s budget
came from sales of land use rights and buildings.26).

Also since the mid 1990s, the city government has been pursuing an urban development strategy of
eventually expanding the downtown to the east bank of the Saigon River in District 2, opposite the existing
downtown. The area, known as Thu Thiem, has ample room and would relieve the pressure on the historic
quarters of the city. But unlike South Saigon, it is densely inhabited by poor families who will have to be
resettled elsewhere. Throughout the late 1990s and early 2000s, the government remained undecided over
bridge building and compensation policy in Thu Thiem while speculators drove up land prices. As a result,
the area has remained undeveloped wetlands over the last 10 years. Most recently, the project has been
give new life with the completion of Thu Thiem bridge, the expected completion of Thu Thiem tunnel and Phu
My bridge, and the government’s decision to triple to the land compensation price. In the long run, the
development of Thu Thiem as an extension of the downtown is inevitable if the historic districts around the
existing downtown are to be preserved. But Thu Thiem has many wetlands and its development threatens to
exacerbate flooding in the city.

The new 2025 land use plan seeks to accommodate a growing population and higher incomes without
building on flood-prone land or encroaching on the historic neighborhoods around the downtown. Toward
those ends it proposes the development of satellite centers to disperse central business district functions and
reduce pressure on the historic neighborhoods. The traditional downtown, Thu Thiem, the Chinatown
business area in District 5 and South Saigon would function together as the new “extended central business
district” (Figure 5). Other suburban business districts would be developed on the periphery of the city,
outside the second ring road. And manufacturing would be encouraged in “growth corridors” with good
transportation mostly located along major city highways leading out of the city.

The satellite center strategy seems to be an unhappy and perhaps unworkable compromise for the problem
of the downtown. In the first place, the development of Thu Thiem and South Saigon as major business
centers clearly requires building on wetlands, in conflict with the plan’s goals. .Equally important, it is unclear
that the central business district functions can successfully be divided among four satellite centers, especially
since Chinatown and Saigon South are several kilometers apart from the others. Moreover, the
transportation components of the plan seems inconsistent with the satellite approach in that the MRT and
elevated expressway systems are centered on Districts 1 and 3, practically inviting high rise office
development in historic neighborhoods.

25
Seventy percent of the land had been developed as of 2007, and a population of 100,000 was expected when fully built out.
26
Institute for Economic Research and United Nations Development Program, Paying for Urban Infrastructure and Services: A
Comparative Study of Municipal Finance in Ho Chi Minh City, Shanghai and Jakarta, UNDP, Hanoi, June 2007, pp. 16-18.

14
Some of these defects are addressed in two more detailed plans for Thu Thiem and the historic downtown.
The detailed plan for Thu Thiem, prepared by Sasaki, an American architectural and planning firm, calls for
maintaining and enlarging some canals that flow through the site and for focusing most of the development in
the north and center of the area, avoiding the southern edge which has the highest concentration of
wetlands. There has been no detailed hydrological study of the consequences of developing Thu Thiem,
however, so there is little assurance that these measures will be effective in limiting the flooding risk.

The detailed plan for the historic central area is being prepared by Nikken Sekki, the consultants who drafted
the 2025 master plan. The main idea is to concentrate office and residential high-rise development along the
river, and keep it from encroaching on the historic area behind. Toward that end, the consultants propose to
move MRT lines 3 and 4 further east in the central area, so that their stations are closer to the river instead
of in District 3 and the western part of District 1. Equally important is their suggestion to convert the obsolete
port facilities immediately to the north and south of the traditional downtown into downtown uses. A four-lane
boulevard that runs along the western bank of the river would be put underground in spots to strengthen the
link between the downtown and the river and to increase the road’s capacity. The conversion of the ports
could turn underused facilities into valuable assets and buy some time to think through the development of
Thu Thiem.

Even when the technical aspects of the city’s land use plans are carefully considered, the financial feasibility
of the new housing and commercial development is still in doubt given the extraordinarily high cost of land. A
recent article put the cost of land next to major roads in Japan, a crowded and mountainous country with per
capita GDP of nearly $40,000, at $1050 per square meter. Urban land in Vietnam is often more than this and
even cheaper land near major roads is typically several hundred dollars per square meter. With Vietnam’s
per capita income still under $1000, the relative cost of land is completely out of line with incomes27 If land
costs tens of thousands of dollars for even a modest official plot, it puts ownership of urban housing beyond
all but a very few.

For typical housing projects developed recently in District 7, next to Phu My Hung, for example, land
accounted for approximately three quarters of the final cost of townhouses built, ultimately making the units
unaffordable for middle class families. The market price of land in the Phu My Hung area rose to VND 36
million in late 2007, at the height of the real estate bubble, but then declined to VND 16 million by late 2008,
after the bubble had deflated. Even at VND16 million (or $1,000) per square meter, a family would require
VND 960 million to purchase a modest plot of 60 square meters of land in this project. By contrast, the
construction on the site of a townhouse with 80 square meters of floor area would cost only VND 320 million,
bringing the total cost to VND1.28 billion (or $75,300):

Town House Development Cost in Phu My Housing Project, District 7:

2
Land cost: VND16 million * 60 m = VND960 million
Construction cost: VND4 million * 80 m2 = VND320 million
Total acquisition cost: VND1.28 billion ($75,300)
At a 10% interest rate, the family would have to make an annual payment of VND135 million in 30 years to
finance the acquisition of the house through a mortgage. Even an upper middle-class family with available
annual savings of VND60 million (i.e. VND5 million a month) would never be able to finance the purchase of
the house.

The high price of land is not the result of high compensation to farmers or high costs of fees to government
or required infrastructure but rather appears to be caused by a shortage of permitted sites. The table below
shows that the average development cost of land (including compensation, land-use right fee, and
infrastructure) incurred by the developer in the above project was only VND4.72 million (or $278) per square
meter in 2007 prices. The mark up to VND 16 million per square meter must reflect some artificial constraint
on supply, possibly caused by speculators holding back land in the hope of ever higher prices but more likely
because of city officials holding back permissions for the conversion of land to urban uses. During the
expansionary monetary policy period of 2006-2007, speculative activities did play a major part in fueling a
real estate bubble in the newly developed southern and eastern peripheries as commercial banks were on

27
Per capita income is valued at exchange rates, not PPP (purchasing power parity) in this comparison because that is the relevant
concept to use. Real money is used to buy and sell land – it is not an imputed value multiplied to get an equivalent price to the United
States or Japan.

15
an unprecedented real estate lending spree.28 Nevertheless, the fact that land and housing prices have
remained high after the burst of the bubble indicates scarcity as the fundamental factor.

Land Development Cost in Phu My Housing Project, District 729

Project land area: 8.16 hectares

Housing land area: 3.93 hectares
Agricultural land compensation: VND133.14 billion (VND1.63 mil/ m2)
Land-use right fee: VND31.58 billion
Infrastructure cost: VND65.34 bil (VND0.8 mil or $47/m2)
Total cost: VND 185.68 bil
2
Average cost per m of housing land: VND4.72 mil ($278 per m2)
If the land price were VND 4.72 million instead of VND16 million per square meter, the total acquisition cost a
town house would be only VND603 million (or $35,500). The financing cost for the house at a 10% interest
rate would be VND64 million a year (or VND5.3 million a month) in 30 years, which is within the reach of
middle-income families in the city.

If land prices remain as high as VND 16 million per square meter, then the solution will be to develop high-
rise buildings. As shown in the rough calculation below, the cost of development of a 65-square-meter
apartment in a 20-storey building is VND 625 million (or $36,700).

Apartment Development Cost in Phu My Housing Project, District 7:

Ratio of the footprint of the building to the plot area: 50%

Number of floors: 20
Land cost: VND16 million per m2
Apartment construction cost: VND8 million per m2
Average cost of land and construction per m2 of apartment floor area:
VND9.6 million (or $565)30
Cost of a 65-m2 apartment: VND625 million (or $36,700)

However, just like land, there is also a huge difference between the market apartment price and its
development cost. If developers typically incur VND9-10 million per square meter in cost for land as
estimated above, the apartment price is bid up to VND16-18 million a square meter. At this price, apartments
are no longer affordable to middle-income people.

The housing situation has worsened in the past decade. Previously most urban dwellers have received
apartments or land from the state or from their parents. Indeed, in the 1999 Census only 11 thousand
households were listed as not owning their urban living place out of 16.7 million people! However, the rapid
migration of recent years together with strong income growth has increased the demand for housing and
driven up prices. Unofficial migrants more often rent than own, and often illegally. A 2004 study focused on
migrants reported that of the migrants surveyed, 60% reported housing problems.31 Sky-high land prices
mean home ownership, or even affordable rents in decent housing, is difficult or impossible for most
migrants. Their problems are intensified if infrastructure is not built for them.

28
The State Bank of Vietnam’s Banking Report 2008 shows that as of August 2008, nine banks out of the total 41 have more than 30
percent of their loan portfolios in real estate finance (with one bank reporting the proportion of 50 percent). Another nine banks have
property-related lending accounting for more than 20 percent. The real estate prices peaked in January 2008, just a month after the
peak in new loan disbursements (See Fulbright Economics Teaching Program, Policy Discussion Paper No. 3, 18 Sep 2008).
29
Information taken from the ADC Phu My Project Feasibility Study adjusted for 2007 prices.
30 2 2 2
For example, a 20-storey building with a footprint of 2,000 m constructed on a 4,000 m plot would have 40,000 m of total apartment
floor area and VND64 billion of total land cost. The average cost of land per m2 of apartment floor area would be VND1.6 million. The
average cost of land and construction per m2 of apartment floor area would be VND9.6 million
31
The Quality of Life of Migrants in Vietnam, a volume in the 2004 Vietnam Migration Survey, Table 3.1. The survey is a joint General
Statistics Office/UNFPA effort and was published in 2006.

16
 Urban Finance

In the end, the planned improvements to urban transport systems and the development of new urban areas
will depend in part on the availability of municipal financing. According to a UNPD-funded study comparing
municipal finance of HCMC, Jakarta and Shanghai32, HCMC is quite successful in mobilizing tax revenue,
but is permitted to keep only roughly 30% of the taxes it raises. The revenues retained are significantly short
of HCMC’s budgetary needs, and despite creative financing the city has not been able to keep up with the
its rapid growth and concomitant resource needs….Debt financing through the issuance of municipal bonds
has become a significant source of income over the past four years rising from zero to half of all regular local
revenue; other revenue sources have risen only slightly during the same period, so their relative shares of
total local regular revenue have declined. The debt figures significantly underestimate HCMC effective
borrowing in that they exclude the city’s considerable contingent liabilities incurred through its explicit and
implicit guarantees of off-budget borrowing by local government entities….This greatly limits the potential of
a significant increase in HCMC debt financing to pay for future investments in urban infrastructure.33 In
short, HCMC is going deeply into debt to finance its infrastructure, but is falling behind in part because it can
only keep 30% of the taxes it collects. But the current pattern of debt accumulation cannot be continued and
HCMC will soon have to cut back sharply from the already inadequate levels it is investing.

HCMC’s financial situation is in part due to a 2004 fiscal decentralization reform which cut back on central
government transfers to richer cities while maintaining the high share of local taxes those cities must forward
to the central government. In her study of the 2004 reforms Pham Lan Huong of the Central Institute of
Economic Management (CIEM) reports that local governments no longer retain any share of import and
export taxes, VAT and excise taxes on imports, petroleum industry related taxes, and corporate income
taxes on enterprises with whole-unit accounting. Local governments get to retain only all land related taxes
and a share of locally generated non-import VAT, other corporate taxes, the personal income tax, the special
consumption tax, and gasoline and oil fees. The new law also allows provinces to borrow directly under
conditions stipulated by the central government without further approval.34

Even before the 2004 reforms, the richer provinces supported the poorer provinces through a system of
fiscal transfers designed to equalize per capita spending for social services, particularly health and
education. Between 1998 and 2003 the share of supplementary transfers from the central budget in all local
spending rose from 39% to 57%. In 2002, only 6 or 8 of Vietnam’s 64 provinces had to pay out appreciable
transfers35, while the rest retained all their local taxes and received transfers from the central government as
well. Several poorer provinces received over $50 (about 750 thousand dong) per capita in subsidies.

While the fiscal system has been remarkably successful in equalizing public spending per capita for health
and education it may not adequately recognize the degree to which the rapidly growing urban provinces
need to invest in infrastructure, schools and other public facilities. Public infrastructure investment per capita
is higher in the richer and growing provinces. In 2002, for example, 6 provinces invested over 2 million dong
per capita, while another 10 provinces invested between 1 and 2 million dong and the remaining 48
provinces invested less than 1 million dong per capita. But infrastructure investment needs may be a function
of the rate of population growth as much or more than the population level. Even provinces that are losing
population must invest in infrastructure because the existing facilities and services often need to be
upgraded or extended to remote or underserved communities. However, rapidly growing provinces not only
must upgrade the infrastructure that serves the existing population but also build new roads, schools, clinics

32
Institute for Economic Research and United Nations Development Program, Paying for Urban Infrastructure and Services: A
Comparative Study of Municipal Finance in Ho Chi Minh City, Shanghai and Jakarta, UNDP, Hanoi, June 2007.
33
Pages 1 and 2 in Paying for Urban Infrastructure and Services: A Comparative Study of Municipal Finance in HCMC, Shanghai and
Jakarta.
34
“Fiscal Decentralization from Central to Sub-national Government in Vietnam,” Pham Lan Huong, CIEM, December 2006, p. 7. The
new sharing follows the 2002 Budget Law implemented in 2004. The author argues that the VAT and corporate income taxes are
“assigned” to major cities where corporate headquarters are even if the economic activity takes place elsewhere. But if this activity takes
place in other rich provinces, the regressive impact of this is less than if it takes place in poorer provinces.
35
The Ministry of Finance website gives data for the planned 2006 and 2007 budgets. Seven provinces or cities had more than one-third
of the revenues collected in their borders directed on a net basis to the central budget: HCMC in 2007 had local spending/revenues of
18%; Hanoi had 20%; Haiphong had 22%; BR-VT had 4% (oil); Binh Duong had 30% and Dong Nai and Quang Ninh had 33%.

17
and power or water systems to accommodate new residents.36 The close link between investment needs
and population growth is the reason why it is so important to know the actual population growth in each area.
Public investment, after the basics, should be skewed to those areas with growing needs and demands. In
spite of the current distribution of investment, it is unclear that current investment patterns reflect the actual
population or economic growth patterns. There may be “excess” investment in some lagging provinces and
“deficit” investment in some fast-growing ones. Here, the actual population growth data are crucial.

As an aside, there is also huge room for improvement in the spending efficiency of HCMC on its
infrastructure. Too often, the land compensation claims for roads or other infrastructure soar to incredible
levels rather than learning from relevant experience in places such as China. There, the land taken for a new
road extends some distance from the actual new roadway, and those who had land fronting on the old road
get ground level ownership of a multistory building fronting on the new road. Those who had been one row
back from the old road get second story apartments or shops in newly constructed buildings next to the new
road. This reduces considerably the compensation payments, as the land owners are paid in kind, not cash.

It is not unusual for nations experiencing rapid industrialization to increase payments to rural areas. By its
nature, agriculture cannot grow nearly as fast as industry and unless there is a large amount of rural nonfarm
activity, rural areas will tend to have lower incomes and slower growth. This difference in income levels is
often offset in part by domestic remittances as grown children work at better paid urban jobs and send some
money home to their parents or other family members. Unfortunately, little is known about these within-family
transfers.37 Rural-urban disparities can also be ameliorated by fiscal transfers, such as those described
above. It is normal and politically healthy to not allow differences in urban and rural incomes to become
too large.

However, if virtually all of the nation’s population growth is occurring in urban and near-urban areas, then the
country faces a difficult choice between continuing to invest heavily in rural areas after their basic
infrastructure is in place, even if little economic activity results, and supporting the thriving areas which most
need new investment. It may be neither fair nor efficient that HCMC generates 30% of the nation’s
government revenues but spends only 9-11% of them. If stability comes from generating decent jobs for new
workers then investing in thriving areas may be preferable but if reducing inequalities in regional per capita
incomes is more stabilizing, even at the cost of some growth, then it may be better to continue to invest in
slower growing regions.

In 2004, there were about VND 8.1 trillion invested by HCMC or about 1.1 million dong (or $73) per capita in
the local public investment budget. Most of this was financed through debt, land sales/transfers, or other
methods such as BOT (build-operate-transfer). If these sources are limited over time, then only a few
outcomes are possible. One is that the HCMC government finds a way to reduce the unit cost of its capital
investment, making do with less but still getting critical infrastructure built. A second is that spending levels
fall with roughly similar efficiency as now and much less infrastructure is built. A third possibility is that
revenues now directed to poorer provinces are redirected to HCMC, with or without efficiency changes. A
fourth is that the central government takes on debt payments for investments that are now paid for by the
city. (This is similar in effect to option three, but with a different mechanism.) A fifth is an even greater use of
BOT or land swap arrangements for infrastructure. (This negates the assumption that these methods have
limits that are being approached.) What is likely to happen?

Vietnam already has a rather high ratio of government revenues and spending relative to GDP, given its level
of per capita income. Expenditures are 28-31% of GDP in recent years while revenues are 25-27%.38 It is not
likely to be easy to increase these ratios efficiently, unless oil and gas production and revenues increase
markedly or new sources of revenues are developed. (China has revenues and expenditures that are about
19% of GDP.) If indeed, revenues cannot rise much faster than GDP, then the allocation of revenues will
become even more important.

36
The 2004 Living Standards Survey reported that 91.6% of rural households surveyed had electricity, up from 82.7% in 2002. While the
Northwest region had only 72% and the Central Highlands and Mekong Delta had only 86-87%, the 2007 data are certainly even better.
If power availability is a good proxy for infrastructure, the basic extension of services is progressing well and is fairly far along.
37
In the 2002 Living Standards Survey, “other” income is $45 per person a year or $3.57 billion. This amount is within the range of some
estimates of foreign remittances, and as it might include gifts from nearby relatives as well, it is not useful to try to extract rural-urban
payments from this single data point.
38
These data, and those for China, are taken from Key Indicators 2007 of the Asian Development Bank.

18
An Illustrative Calculation

In 2007, the region around HCMC (including HCMC province and five provinces surrounding it) was
budgeted to collect 170 trillion dong in taxes and fees, or 58% of the national total. However, this amount
includes 74 trillion from Ba Ria-Vung Tau, which is largely oil – a clear national resource. Subtracting that
province’s revenues and spending leaves 96 trillion in non-oil revenues compared to 21.4 trillion in spending,
a ratio of 22%. Similarly the region around Hanoi (consisting of the provinces of Hanoi, Haiphong, Quang
Ninh and Vinh Phuc) was budgeted to collect 68.5 trillion in revenues and spend only 16.5 trillion, a 24%
ratio. Together, these richer provinces around HCMC and Hanoi provide 75% of the nation’s non-oil
governmental fees and taxes but account for only 32% of spending. (Ba Ria-Vung Tau is excluded from both
revenue and spending in this calculation.)

That means that all other provinces provided only 25% of non-oil revenues but spent 68% of total provincial
spending. If the richer provinces were allowed to keep one-third of their non-oil revenue, that would transfer
39
17 trillion, or 20% of current spending, from the poorer provinces to the richer ones. This would increase
the resources available to the richer provinces by nearly 45%. Such a hypothetical switch would be
unpopular with the poorer provincial governments who already see themselves as hard-pressed. Yet if the
cutbacks focused on less productive capital projects rather than on equalizing health and education, the real
disadvantage might be minimal. Of course, having large projects in one’s own province is politically popular.
If many workers are going elsewhere (as seems to be the case), however, perhaps more of the “rich
province” taxes should stay where they were collected to provide needed services and infrastructure.

It bears repeating, however, that the quid pro quo for allowing the richer provinces to retain more revenue
should be greater efficiency in their investments. Serious transportation and urban plans that minimize
distances from home to workplace or school would be one place to start. Avoiding drainage problems from
filling in wetlands, and treating land compensation more rationally would be other steps. Setting aside land
for urban rail and roads might be another idea. The point is that these “extra” funds, while needed, will do
little good if wasteful and conflicting plans and projects continue. Vietnam cannot afford highly expensive and
inefficient urbanization. Too many workers want to live in the vibrant economic regions. If these regions
include a number of provinces adjacent to the major cities, the problems of crowding and congestion could
be contained. Indeed, successful development tends to spread outward, helping ever more remote
provinces. A more rational and efficient distribution of infrastructure projects would help this outcome.

For those who argue that the large disparities between richer and poorer provinces should require more
spending in the poorer areas, the question must be how much good the existing pattern of public investment
actually accomplishes. Recall that with its oil revenues, Vietnam already has a 50% higher ratio of
government spending to GDP than China does. This high level of spending is skewed towards the poorer
provinces. Basic infrastructure has been put in place. More needs to be done with rural roads and other
improvements, but the temptation is to continue construction beyond what is productive.

There is currently little good information on the comparative advantages of public investment in various
provinces. Of course, there are projections based on wishful thinking about growth, but these are seldom
reliable enough to be taken seriously. Proponents of a particular project seldom investigate whether other
approaches might satisfy the need at a lower cost. The result is that most choices are political and
administrative, with only weak input based on probable productivity.

For projects that could be built with private funding and should recover costs, the obvious way to check for
efficiency is to require cost recovery with reasonable charges. Extending electrical transmission and
distribution lines40, building ports and airports, putting in place urban water and sanitation, or constructing
major bridges, tunnels and expressways could all be based on this principle. Having private developers
involved with some of their own capital at risk (without a sovereign guaranty – based only on project
revenues for success) is sometimes a good way to ensure that the revenue projections will be serious rather
than fictional.

39
The current system provides for some taxes that go entirely to the locality and some that are subject to a split. This is a good idea, for
if all taxes went largely to the center, there would be little incentive for the richer locality to collect taxes, and more fees and other
charges would appear.
40
There has also been a tendency to prefer hydroelectric generation over other types, even though this contributes to dry season supply
shortages and is often more capital-intensive than alternatives. The issue of choice of electrical generating type is a complex subject
beyond the scope of this paper.

19
However, there is a large class of projects that cannot reasonably be expected to recover costs. Rural and
minor urban roads, small bridges, drainage and rural water supply (among others) need to be financed and
maintained out of general revenues. Some electrical or telephone transmission to very remote areas might
also fit in this category. In such cases, it is best to develop comparative standards and build to a certain
density of service consistent with funding availability. This ensures fairness and minimum
infrastructure standards.

For industrial estates, the tendency has been to take land cheaply and then put in infrastructure with a “build
it and they will come” attitude. The trouble is, “they” have not come in many cases. Many industrial estates
are underutilized and have wasted land and money, besides displacing many thousands of families. In these
cases, less investment is desirable. Most of these estates already have the minimum needed infrastructure.
Without more indication of need – that is, interest from non-state investors – the argument for central funding
of these estates is weak. This would be an obvious place to reduce capital spending.

Finally, there are “breakthrough” or “strategic” public investments such as the proposed $56 billion high
speed rail line between HCMC and Hanoi. There is no realistic prospect of cost recovery in such cases.
European and Japanese experience indicates that for trips over four hours, passengers prefer to fly. Given
that the only two major origin and destination points would be Hanoi and HCMC, and these would be a day’s
trip, the prospect of attracting large numbers is limited. The existing flights are frequent and cheap – cheaper
than the cost of a rail ticket would be. For proponents, the lack of cost recovery is only proof of how bold the
project is. One response must be that the real cost of such projects is not the dollars spent, but the
congestion, time wasted, and slower growth in major cities that are starved of needed infrastructure. While
prestige projects are observed everywhere, they are best kept to a small fraction of total spending and
subsidiary to essential investments.

20
 Conclusion

This paper has argued that the actual population levels and particularly the rate of population growth in
HCMC are higher than officially measured. In the face of the explosion in population growth, two key
challenges have emerged, namely, traffic congestion and development of new urban areas. It is critical that
the government agencies responsible for developing and implementing city’s transportation and land use
policies have the capacity to do so. The preparation of the 2025 land use and transportation plans was an
extremely useful exercise in forcing the city to confront the issues raised by economic growth. But in many
respects planning is just the beginning of the effort.

Even with the plans, there are still many important decisions to be made. In the area of transportation, for
example, which of the many MRT lines and expressways should be built first given that there may not be
funds to build them all? How can the alignments of the expressways and MRT lines be adjusted to minimize
the unsightliness of the facilities themselves, and to better support development where it is desired and
discourage development where it is not? How should the capacity of existing streets be managed given that
the new facilities will provide never provide all the capacity needed? Similarly in the area of land use, exactly
how do the general guidelines in the master plan translate into standards for the density and uses allowed on
particular tracts of land? How should the expansion of the traditional downtown be managed, especially if
Chinatown and Saigon South cannot attract central business functions? And are there reasonable strategies
to cope with the flooding threats posed by the development of Thu Thiem and Saigon South?

And where the decisions have been made, the capacity for implementation will be essential. The cost of new
MRT lines and expressways can inflate rapidly, for example, if procurement is poorly managed. And policies
to manage existing street capacity typically depend on the willingness of police to enforce them. Similarly in
the benefits of land use controls depend on the ability of the agencies in charge of permitting new
developments to issue and enforce the supporting regulations.

Finally, since infrastructure, beyond some minimum, should be related to population growth, the current
amount of major urban investment is lower than it should be. This has been offset by non-sustainable debt
and land sales, but these expedients are temporary. A new division of tax revenues between rich and poor
provinces is needed to reflect this.

However, simply allowing richer provinces to keep a larger share of their tax payments would not be
productive without major enhancements to project planning and implementation. Many public projects are
currently being built well ahead of demand, too expensively, or inferior to alternatives.

The first step is to count the people actually living in each province. The second is to agree upon a rational
division of tax revenues. The third is to develop a stronger review process for selecting and implementing
public investments, including ways to recover costs when feasible. If these three things are accomplished,
the urbanization and development of Vietnam will be more rapid, sustainable and equitable than if cities are
starved of essential investments while slowly growing rural areas are subsidized beyond a point where the
investments are productive.

21
 References

Duranton, Gilles and Diego Puga (2000). “Diversity and Specialization in Cities: Why, Where and When
Does it Matter?” Urban Studies 37: 297-308.

HCMC Statistical Office, HCMC Statistical Yearbook 2007.

Huong Ha and Tai-Chee Wong, “Economic Reforms and the New Master Plan of Ho Chi Minh City, Vietnam:
Implementation Issues and Policy Recommendations”, GeoJournal 49 (1999)

Ing Husu, Ahamad Frahan Mohd Sadullah, Nguyen Xuan Dao, “A Comparison Study on Motorcycle Traffic
Development in Some Asian Countries: Case of Taiwan, Malaysia and Vietnam”, paper prepared for the
Eastern Asia Society for Transportation Studies, October 2003.

Institute for Economic Research and United Nations Development Program, Paying for Urban Infrastructure
and Services: A Comparative Study of Municipal Finance in Ho Chi Minh City, Shanghai and Jakarta, UNDP,
Hanoi, June 2007.

Japan International Cooperation Agency (2004) “Study on Urban Transport Master Plan and Feasibility
Study in Ho Chi Minh Metropolitan Area (Houtrans)” undertaken for the Ministry of Transport and the Ho Chi
Minh City People's Committee with support from the Japan International Cooperation Agency. Prepared by
ALMEC Consultants, Japan.

Ministry of Labor, War Invalids, and Social Affairs, Statistical Data of Labor-Employment in Vietnam, 2004.

MVA Asia Ltd, “Bus Rapid Transit”, task 2 report of “Technical Assistance for Consolidation and
Development of a Bus System in Ho Chi Minh City, Vietnam”, March 2006.

Nguyen Van Hiep, “The track records and challenges of housing problem in Ho Chi Minh City at present and
up to 2010”, Mega City Research Project, Ho Chi Minh City.

Pham Lan Huong, “Fiscal Decentralization from Central to Sub-national Government in Vietnam,”, Central
Institute for Economic Management, December 2006.

Systra MVA, Ho Chi Minh City Metro Rail System Study, HCMC Master Plan Ridership and Revenue
Forecast Study, Final Report, January 2008.

Urban Planning Institute of HCMC and Nikken Sekkki (consultants), “The Study on the Adjustment of HCMC
Master Plan up to 2025”, report to the Ho Chi Minh City People’s Committee, 2008.

22
 Table 1. Growth in Gross Regional Product, HCMC Province and Vietnam

Year HCMC Vietnam

1997 14.6% 8.2%
1998 12.1% 5.8%
1999 6.2% 4.8%
2000 9.0% 6.8%
2001 9.5% 6.9%
2002 10.2% 7.1%
2003 11.4% 7.3%
2004 11.7% 7.8%
2005 12.2% 8.4%
2006 12.2% 8.2%
2007 12.6% 8.5%

Source: General Statistical Office, Vietnam Statistical Yearbook 2007; Ho Chi Minh City Statistical Office,
Ho Chi Minh City Statistical Yearbook 2007

Table 2. Population and Motor Vehicle Registrations in Ho Chi Minh City

Population Motorcycles Cars

Year (thousands) (thousands) (thousands)
1990 4,118 n.a. 12.3
1991 4,259 500 13.5
1992 4,426 n.a. 14.7
1993 4,531 n.a. 16.1
1994 4,582 844 17.5
1995 4,640 n.a. 19.4
1996 4,749 n.a. 38.0
1997 4,853 1,200 n.a.
1998 4,958 n.a. n.a.
1999 5,064 n.a. n.a.
2000 5,249 1,571 131.0
2001 5,449 1,970
2002 5,659 2,285 158.2
2003 5,867 2,306 221.7
2004 6,063 2,429 253.0 (115.0 private)
2005 6,240 2,623 267.8
2006 6,425 2,903 296.1
2007 6,651 3,406 399.0 (202.0 private)
N.a. indicates not available.

Sources: Population figures from Ho Chi Minh City Statistical Office, “Ho Chi Minh City Statistical Yearbook 2007”, pp. 22
and its website http://www.pso.hochiminhcity.gov.vn. Vehicle registrations for 1990-1996 from Japan Bank of
International Cooperation, “Urban Public Transportation in Vietnam: Improving Regulatory Framework”
JBIC Research Paper no. 4, December 1999, pp. 5 and 8. Vehicle registrations for 2000-2007
from Ho Chi Minh City Road Traffic Police Department.

23
 Table 3. Percentage of Trips in Ho Chi Minh City by Mode Used, 1996 and 2002

1996, without 2002, without 2002, with

walking walking walking
Walking --- Walking --- 17.1%
Bicycle 20.4% Bicycle 17.4% 14.4%
Bus 0.2% Bus 1.7% 1.4%
Motorcycle 76.5% Motorcycle 74.5% 61.8%
Car 2.2% Car 1.4% 1.2%
Conventional taxi and
Truck 0.7% 4.1% 3.4%
other
All modes 100% Motorcycle taxi (Xe Om) 0.8% 0.7%
All modes 100% 100%

Trips per weekday

8.2 11.9
(millions)

Sources: 1996 figures from DFID 1998 Ho Chi Minh City Transportation Study as cited by Japan Bank for International
Cooperation, “Urban Public Transportation in Vietnam: Improving Regulatory Framework, JBIC Research Paper no. 4,
December 1999, pp. iii and 12. 2002 figures estimates from Almec Corporation, “The Study on Transportation Modes
and Feasibility Study in HCM Metropolitan Area (HOURTANS)”, vol. 1, Summary, June 2004, p. 2-17.

Table 4. Demand Forecasts for 2025 with 161 Kilometer Rail Transit Network

Forecast consistent with Forecast consistent with current

master plan goals policies toward bus speeds, auto
parking and auto ownership
Share of trips by mode All modes Public only All modes Public only
Motorcycle and bicycle 38% 49%
Car 18% 29%
Public transport
Rail 9.2% 21% 6.8% 30.7%
Bus 30.8% 70% 12.9% 58.5%
Ferry 0.4% 1% 0.1% 0.4%
Intercity rail and bus 3.5% 8% 2.3% 10.4%
Subtotal public 44% 100% 22% 100%
All modes 100% 100%
Modes on urban streets 87% 91%
Modes not on urban 13% 9%
streets
All modes 100% 100%
Forecast consistent with Forecast consistent with trends
master plan goals
Peak load per Peak load per hour
Riders by line Weekday hour Weekday
Line 1 863,000 29,200 716,000 28,300
Line 2 551,000 24,900 405,000 27,900
Line 3 474,000 14,300 313,000 11,200
Line 4 366,000 12,000 260,000 11,600
Line 5 525,000 8,000 320,000 5,500
Line 6 69,000 2,700 52,000 2,000
Tram 386,000 9,100 225,0000 5,800
Monorail 22,000 1,500 20,000 1,400
Total 3,256,000 2,310,000

Source: Systra MVA, “Ho Chi Minh City Metro Rail System Study,
HCMC Master Plan Ridership and Revenue Forecast Study, Final Report”, January 2008,pp. 47,60, 63, 66

24
 Figure 1. Map of Ho Chi Minh City

Source: Government of Vietnam Tourism Office

25
 Figure 2. Map of Ho Chin Minh City’s Inner Districts

Source: Government of Vietnam Tourism Office

26
Figure 3. Plan for Four Elevated Expressways

27
 Figure 4. MRT Plan as of 2008

28
 Figure 5. Location of Business Centers in 2008 Master Plan

Source: Urban Planning Institute of HCMC and Nikken Sekkki (consultants), “The Study on the Adjustment of HCMC
Master Plan up to 2025”, report to the Ho Chi Minh City People’s Committee, 2008.

29