REGIONAL POVERTY DISPARITY IN VIETNAM

Vu Tuan Anh Socio-Economic Development Centre

Hanoi, Vietnam

Vietnam has 82 million inhabitants, living in different regions, which have different

geographical, climate, economical and social characteristics. Economic growth, livelihood,

income, living conditions and poverty of population are varried by regions.

Vietnam has made cosiderable progress in poverty reduction. The poverty rate in the whole

country has halved after less than ten years. However, speed of poverty reduction is still

low in some regions. Identification of regional disparities of multidimensional poverty

provides background for right targeting to the poor and elaboration of appropriate poverty

policies in each specific region.

In some last years, the Vietnamese CBMS project has co-operated with local partners in 5

provinces to conduct poverty studies. These 5 provinces are located in 5 different regions.

This paper presents results of CBMS implementation in Vietnam's localities and using

CBMS data for analysis of regional disparities of poverty. The main objective of the

analysis is to find out the possible explanatory factors affecting the disparity. Based on

results of analysis of regional disparity in poverty, some poverty alleviation policies and

proposal of application of a multidimensional poverty index are recommended.

This paper consists of three parts. The first part gives an overview of poverty reduction and

regional poverty disparity in Vietnam. The second part examines regional disparity in

multi-dimenssional aspects of poverty, basing on analysis of CBMS data. The third part

suggests a regional poverty index, which might be used for regional poverty comparison.

I. Economic Growth and Regional Poverty DisparityEconomic growth is first and essential factor for improvement in the living standards of the

population and to reduce absolute poverty. Where there is fast and stable econonomic

growth, poverty is less and easierly alleviated. It is through the process of direct economic

impacts on population's employment and income generation, and the process of trickle

down that growth benefits percolate to the lowest strata of the society. The increased

disparities in the distribution of living benefits both across social strata and between

2

different regions, which are widely experienced in many developing countries reflect the

failure of distribution policies, inappropriate social and political institutions.

Regional disparity in living standard is used to be measured by difference of income,

expenditure of population in different regions. It is measured also by indicators of specific

aspects of welfare, such as education, health, etc. Regional disparity in poverty is measured

by difference of poverty rates of regions.

In the 1990s, Vietnam witnessed acceleration in the growth rate of GDP. It registered an

average annual growth rate of GDP of 7.6% in the last 16 years (1990-2006). During this

period, Vietnam’s population was increased 118 %, GDP grew to 322%. This caused GDP

per capita to grow to 253% or 6% annually.

Vietnam’s per capita GDP was US$288 in 1995, $639 in 2005, and $835 in 2007.

Thank the stable economic growth, poverty rate has reduced significantly. The poverty rate

measured by Vietnam Living Standard Surveys reduced continuerly from 58.1% in 1993 to

16,0% in 2006. Poverty rate in 2002 has halved, comparing to that in 1993, and poverty

rate in 2006 has also halved, comparing to that in 1998. In average poverty rate halved in

every 8-9 years. This means the first goal of MDGs is completed. Vietnam is considered

as a successful case of poverty reduction among developing countries.

Figure 1: GDP and poverty rate in 1993-2006

Sources: General Statistical Office, Vietnam Statistic Yearbooks.

27.1

13.2

35.1

45.5

5558.1

37.4

28.9

19.516

0

10

20

30

40

50

60

1993 1998 2002 2004 20060

10

20

30

40

50

60

70

GDP (bill. US$) Poverty rate (%)

3

The Human Poverty Index for developing countries (HPI-1) developed by UNDP is an

indication of the standard of living in a country . It is a measure of the extent to which

people in a country are not benefitting from development. While Human Development

Index consists of three essential dimensions of human life: longevity, knowledge and

standard of living, and assesses these components as development; HPI assesses the same

three components from an opposite point of view to take into account factors that HDI

does not include1. HPI of Vietnam has significantly reduced in the last decade (Table 1).

Table 1: Human Poverty Index (HPI-1) of some Asian countries

HPI Rank

1998 2005 1998 2005

Malaysia 14.0 8.3 18 16

Thailand 18.7 10.0 29 24

China 19.0 11.7 30 29

Vietnam 28.2 15.2 47 36

Philippines 16.1 15.3 22 37

Indonesia 27.7 18.2 46 47

India 34.6 31.3 58 62

Source: UNDP - Human Development Reports 2000, 2007/08.

Despite of these successes, poverty reduction faces certain limitations and remains a major

concern for Vietnamese society, namely:

- Poverty reduction is still fragile, unsustainable. A large proportion of people has low

income, which closed to poverty line; therefore they easily fell to poverty when natual

disasters, economic crisises happen, or even when a member in their families get serious

sick. The probability of falling again into poverty is still common.

- The disparity in income and living standard between rural and urban areas, between

different strata, between the poor and the rich provinces tends to increase. The income gap

1 Three components of HPI are following:

1) Longevity - measured by the proportion of the population not expected to survive to the age of 40 years.

2) Knowledge - measured by the adult illiteracy rate.

3) Standard of living - a composite value measured by the proportion of the population without access toclean water, health services, and the proportion of children under the age of 5 years who are underweight.

4

between the richest quintile and the poorest quintile doubled in 15 years (According to

living standard survey of Vietnam's General Statistical Office, it was 4.2 in 1990 and 8.37

in 2006). The Gini index based on income indicator grew up from 0.35 in 1994 to 0.42 in

2006, while Gini index based on expenditure indicator was 0.34 in 1993, but only 0.37 in

20062. There are two opposite tendences: in one hand the poverty reduces, but in other

hand the inequality increases.

- Regional poverty disparity is extending despite of reduction of poverty rates in all regions

of the country. This paper will study more deeply on this tendence.

Regional poverty disparity is reflected through some following features.

First, poverty has declined significantly in all major regions in the country, but at different

rates.

According to division of the General

Statistical Office, there are 8 main

geo-economic regions:

1. Red River Delta (11 provinces,population 18.4 mil.)

2. Northeast Mountains (11 provinces,population 9.5 mil.)

3. Northwest Mountains (4 provinces,population 2.7 mil.)

4. North Central Coat (6 provinces,population 10.7 mil.)

5. South Central Coast (6 provinces,population 7.2 mil.)

6. Central Highlands (5 provinces,population 4.9 mil.)

7. Southeast Region (8 provinces,population 14.2 mil.)

8. Mekong River Delta (13 provinces,

provinces, population 17.5 mil.)

Map 1: Vietnam's 8 regions

Among these 8 regions, four regions Northwest, Northeast, North Central Coat and Central

Highlands are less developed in term of economic level. They are uplands and most of

2 General Statistical Office, Statistics Yeabook 2007. Hanoi 2008.

5

ethnic minority population lives there. They face many constrains in development process,

including a difficult physical environment, hinders access to infrastructure and low

educational level of population. The poverty rate areas is still high in these regions:

- The Northwest Region has though small population, but the highest poverty rate. During

the 8-year period of 1998-2006, it fell by 24.4 percentage points (from 73.4% in 1998 to

49.0% in 2006).

- The Northeast was the second poor region in 1998, but in 2006 it ranked at the fourth

place. The poverty incidence fell by 37.0 percentage points (from 62% to 25%).

- The Central Highland region was the third poor in 1998 and did not changed it's rank in

2006, despite it's poverty fell 23.8 percentage points (from 52.4% to 28.6%).

- The North Central Coast was in 1998 at the fourth rank, but it became the second poor

region in 2006. The poverty incidence has declined by 19 percentage points (from 48.1%

to 29.1%).

The rest four regions have reduced their poverty at different rates: the Red River Delta

declines by 20.5% percentage points, the Mekong Delta by 26.6 percentage points, the

South Central Coast by 21.9% percentage points, and the Southeast by only 6.4 percentage

points, but it's original poverty rate was already low 12.2% in 1998. (Table 1)

Second, as results of the different speeds of poverty reduction, there exists a big

diffenrence between poverty incidences of regions and this gap is widening for the poorest

regions. The gap of poverty incidence has been reduced in most of regions, except three of

the above four poorest regions. Compared to the lowest poverty rate in Southeast region,

the poverty difference of the Northwest incresed from 6.0 times in 1998 to 8.4 times in

2006; the North Central Coast from 3.9 to 5.0 and the Central Highlands from 4.3 to 4.9

times. In the same time, the difference of Northeast has decreased from 5.1 to 4.3. (Figure

2)

6

Table 1: Difference of poverty rate

1998 2002 2006

Rate(%)

Difference(time)

Rate(%)

Difference(time)

Rate(%)

Difference(time)

Whole country 37.4 28.9 16.0

Red River Delta 29.3 2.4 22.4 2.1 8.8 1.5

Northeast Mountains 62.0 5.1 38.4 3.6 25.0 4.3

Northwest Mountains 73.4 6.0 68.0 6.4 49.0 8.4

North Central Coast 48.1 3.9 43.9 4.1 29.1 5.0

South Central Coast 34.5 2.8 25.2 2.4 12.6 2.2

Central Highland 52.4 4.3 51.8 4.9 28.6 4.9

Southeast Region 12.2 1.0 10.6 1.0 5.8 1.0

Mekong River Delta 36.9 3.0 23.4 2.2 10.3 1.8

Source: General Statistical Office, "Vietnam Statistics Yearbook 2007".

Notes: 1) Poverty rate was calculated by poverty line which is measured by per capita expendituresin a month as follow: 1998: 149,000 VND; 2002: 160,000 VND, and 2006: 213,000 VND.

2) Regional poverty gap is the difference between poverty rates of other regions to the SoutheastRegion which has the lowest poverty rate.

Figure 2: Poverty gaps between regions (Southeast Region = 1)

Source: General Statistical Office, "Vietnam Statistics Yearbook 2007".

0

1

2

3

4

5

6

7

8

9

1998 2002 2006

Northwest Mountains

Central Highland

North Central Coat

Northeast Mountains

South Central Coast

Mekong River Delta

Red River Delta

7

II. Regional Poverty Disparity Measured by CBMS Data

During two last years 2006-2007, the CBMS-Vietnam project team has been supporting

local partners in 5 provinces, which are representatives for 4 regions, to conduct CBMS in

50 communes (45 rural communes and 5 urban wards) of 14 districts.

Table 2: Scope of CBMS implementation in 2006-2007

Region Province Numberof

districts

Number ofrural

communes

Numberof urbanwards

Number ofhouseholds

Red River Delta Ha Tay 10 9 1 10,016

Ninh Binh 1 24 1 16,725

Northern Mountains Yen Bai 1 3 2 6,314

Southern Central Coast Quang Ngai 1 5 0 6,382

Central Highlands Lam Dong 1 4 1 3,500

TOTAL 5 14 45 5 42,937

In the surveyed localities, there are 3 whole districts (in Yen Bai, Ninh Binh and Lam

Dong). Except Quang Ngai, all surveyed localities consist of one or two urban wards and

several rural communes. Although the structure of population in term of rural-urban is not

similar in surveyed localities, but the collected data can be used as examples for comparing

localitires in different regions.

Using CBMS-approach, in 2004-2005 in the framework of a research project of the

Vietnam Academy of Social Sciences we have conducted a nation-wide sample household

survey. The sample of this survey covered 14,044 households in 133 rural communes and

urban wards of 63 provinces (of which 11,740 rural households in 60 provinces and 2304

urban households in 16 cities). In each province two communes /wards were selected, and

in each commune/ward approximately 100 households were randomly selected for

interview. The results of this survey showed changes of socio-economic situation of

households and communities. [Vu Tuan Anh & Nguyen Xuan Mai (2007)]. In this paper

we use also data of this survey, especially data of rural households for analysis of regional

disparity.

8

The indicator set used in CBMS is modified in regions and provinces to adapt to

circumstances of localities. However, a number of core indicators are the same in all

surveyed localities. In this paper, we use these common core indicators for examining

regional disparity in different aspects of socio-economic situation, which closely related to

poverty. These aspects are: (1) household structure, (2) income, (3) dwelling, (4) property,

(5) education, and (6) health care.

2.1. Population and household structure:

There is no significant difference in the sex structure of the population of the surveyed

localities. The ratio of female population is a bit higher than that of male population. It's

happenes now too in whole Vietnam (50.86% to 49.14% in 2007) as well as in almost all

regions, except in Northwest and North Central Coast.

Regarding the age structure, the significant reduction in fertility and the gradual increase

in life expectancy have resulted in the ageing population in Vietnam, with a smaller

proportion of young population vis-a-vis the greater proportion of old population. The

proportion of population aged less than 15 years old has reduced from 39% in 1989 to 33%

in 1999 and further down to 26% in 2007. In the same time, due to higher and higher index

of life expectancy, the proportion of population aged 65 and over in the country increased

from 5% in 1989, 6% in 1999 to 7% in 2007. [GSO (2008a)].

Data of age structure of the surveyed households in all regions reflects this tendency in

Vietnam's population change. However, data of regions also shows that the poorest regions

(Northwest, North Central Coast, and Central Highlands) have higher proportions of young

population (0-14 year old), and accordingly a lower proportion of people in labourable age

and also old-aged comparing to other regions. The cause is the fact that in these poor

regions, the family planning works have weak effects on fertility behaviour and having

many children is still a popular phenomenon, especially among some ethnic minority

communities .

The age structure of the population is used to calculate the total dependency ratio. This

indicator reflects the relationship of age, fertility and mortality levels with labour force in

the country. The dependency ratio is an indicator used to assess the quality of the

population and it reflects the burden of the working age population able to work. Total

dependency ratio is defined as the percentage of the number of people under 15 years old

(0-14) plus old people (65 years old and over) per people aged 15-64.

9

Table 3: Population structure of the regions

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Surveyed households 11.740 2.001 2.695 731 999 999 1.014 901 2.400

Surveyed population 54.526 8.478 12.122 3.656 5.035 4.448 4.911 4.223 11.653

1. Sex structure

- Male 50.5 50.1 50.6 49.8 49.9 51.1 51.2 50.4 50.5

- Female 49.5 49.9 49.4 50.2 50.1 48.9 48.8 49.6 49.5

2. Age structure (%)

0 - 14 27.6 24.7 27.8 30.6 33.5 28.2 33.3 27.6 23.1

15 - 64 67.3 68.5 66.9 64.8 61.9 66 63.4 68.4 71.8

65 and over 5.1 6.8 5.3 4.6 4.6 5.8 3.4 3.8 5.2

3. Dependency ratio (%)

- Total dependency 48.6 46.0 49.5 54.3 61.6 51.5 57.9 45.9 39.4

- Child dependency 41.0 36.1 41.6 47.2 54.1 42.7 52.5 40.4 32.2

- Old dependency 7.6 9.9 7.9 7.1 7.4 8.8 5.4 5.6 7.2

The decline in total dependency ratio of Vietnam has been contributed mainly by the

reduction of the child dependency ratio (0-14). This could be the result of the effective

family planning programmes. The child dependency ratio of the whole Vietnam has

declined over the same period from 84 in 1979 down to 39 in 2007. That is, after 26 years,

the child dependency ratio has reduced to more than a half. Meanwhile the old dependency

ratio has continuously increased but slightly. Data of the regions resembles to the national

one. Again here, three poorest regions have the highest dependency ratios.

The average household size also reflects the differences in population change. While the

average household size in the country is 4.6, the sizes of households varies in regions and

three poorest regions have larger household size. Accordingly, large-scalled households (6

persons and over) in these regions occuppy bigger proportions. Households with 3-5

members are the most popular type in all regions. Especially, the proportion of households

with 4 members - nuclea family with 2 children - is about one third of total number of

households.

Poor households have larger household size. This tendency is precised for all regions. The

witness is larger size of households belonging to low-income quintiles compared to that of

higher-income quintiles.

10

Table 4: Structure of households by size (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Average HH size(person)

4.6 4.2 4.5 5.0 5.0 4.5 4.8 4.7 4.9

Of which: Labourable 3.1 2.9 3.0 3.2 3.1 2.9 3.1 3.2 3.5

Dependent 1.5 1.3 1.5 1.8 1.9 1.6 1.8 1.5 1.4

Number of dependentper labourable

1.5 1.5 1.5 1.5 1.6 1.6 1.6 1.5 1.4

HH size by income quintile

* Quintile 1 5.0 4.5 4.7 5.0 5.2 5.1 5.5 5.0 5.3

* Quintile 2 4.9 4.5 4.8 5.4 5.5 4.7 4.9 4.9 5.1

* Quintile 3 4.6 4.2 4.5 5.2 5.0 4.4 4.7 4.9 4.8

* Quintile 4 4.5 4.1 4.4 5.0 4.8 4.3 4.8 4.6 4.8

* Quintile 5 4.2 3.9 4.0 4.4 4.7 4.0 4.3 4.0 4.3

Structure of HH by size (%)

* 1 person 1.1 1.6 1.4 0.4 0.7 1.5 1.3 0.8 0.8

* 2 persons 5.9 8.4 5.9 4.4 4.6 8.0 5.3 5.7 4.1

* 3 persons 14.4 14.8 15.6 10.1 10.9 15.3 14.8 16.2 14.4

* 4 persons 31.0 38.2 34.5 33.1 23.5 26.1 24.8 27.1 29.6

* 5 persons 22.3 21.7 22.0 21.2 24.3 21.7 23.6 23.5 21.8

* 6 persons and over 25.4 15.2 20.7 30.9 35.9 27.3 30.3 26.7 29.4

2.2. Income

Income is the major indicator, which is used by Vietnamese governmental authorities for

identification of the poor. The Ministry of Labour, Invalids and Social Affairs has defined

the national poverty line for every 5-year period, using monthly per capita income

indicator. The local authorities use this poverty line for identifying the poor and

distributing support and benefits to them. Household's income is also used for analysis of

living standards, poverty and social stratification.

Two opposite tendencies exist in Vietnam during the fast economic growth: in one hand

poverty rate reduces, but in other hand inequality in income distribution among social

groups and regions increases. According to the national household surveys made by the

General Statistical Office, while poverty rate halved during the last 7-8 years, the income

difference between the rishes andthe poorest quintile increased from 7.3 times in 1996 to

11

8.37 times in 2006. The Gini index of income distribution increased from 0.36 to 0.42 in

the same time.

Implemented CBMS in provinces of Vietnam consideres income as one of the most

important indicators for comparison of regions and communities. Calculating average

value, annual turnover per household in the whole country is 22.4 million Vietnam Dong

(VND).Households in the Mekong River Delta have the largest scope of economic

turnover with 33.7 mill. VND. Following are Southeast (30.7 mill. VND), Central

Hishlands (25.8 mill. VND), South Central Coast (21.5 mill.VND), North east (17.7

mill.VND), the Red River Delta (16.8 mill. VND), Northwest (15.4 mill.VND) and lastly

North Cenral Coast 913.8 mill. VND).

Comparing to the lowest turnover level of North Central Coast, the largest scope is

equivalent to 2.4 times, and the average scope 1.6 times.

After deduction of expences from total turnover, the rest value is net income of

households. In average, income per households in whole country is 18.4 mill. VND. The

highest level is achieved in Southeast (27.2 mill. VND), then the Mekong River Delta

(25.3), Central Highlands (20.8), South Central Coast (19.1), Northeast (15.6), the Red

River Delta (14.1), Northwest (13.5) andat the last rank North Central Coast (10.8).

Figure 4: Average household's turnover and income in regions (mill. VND)

Comparing the lowest income level, the highest level is 2.52 times higher, and the average

level is 1.7 times. Namely:

• North Central Coast = 1,00 • South Central Coast = 1,77• Northwest = 1,25 • Central Highlands = 1,93• Red River Delta = 1,31 • Mekong River Delta = 2,34• Northeast = 1,44 • Southeast = 2,52

22.4

16.8 17.715.4 13.8

21.525.8

30.733.7

18.414.1 15.6 13.5

10.8

19.120.8

27.2 25.3

Country Red RiverDelta

Northeast Northwest NorthCentralCoast

SouthCentralCoast

CentralHighlands

Southeast MekongRiver Delta

Turnover Income

12

The difference of 2.52 times between the highest and the lowest average income shows

that income disparity of regions is quite significant. However, according to the national

household surveys conducted by GSO, this regional disparity is much wider - 2.85 times.

The average income per capita in a month is 351 thousands VND. The regional difference

is quit big. Taken the lowest income of the North Central Coast as 1, indexes of the other

regions are as follows: Northwest 1.29, Red River Delta 1.57, Northeast 1.65, South

Central Coast 2.04, Central Highlands 2.10, Mekong River Delta 2.49, and Southeast 2.87.

Figure 5: Average monthly per capita income (thousand VND)

Comparing the lowest income level (North Central Coast), the highest level is 2.87 times

higher, and the average level is 1.9 times. Namely:

• North Central Coast = 1,00 • South Central Coast = 2,05

• Northwest = 1,29 • Central Highlands = 2,11

• Red River Delta = 1,57 • Mekong River Delta = 2,49

• Northeast = 1,66 • Southeast = 2,87

A deeper analysis of income differenciation of quintiles helps to identify tendencies of

social changes and to supply background for policy making process.

Data shows that the poverty rate and depth in Vietnam's rural areas are quit high. Because

of this, the income differentiation in rural areas is sharp. Taken the poverty line of 200

thousand VND defined by MOLISA as criterium for poverty rate calculation for rural

areas, one can remark that all households of the Quintile 1 and most of the Quintile 2 are

poor. The poverty rate is 30-35%, which is similar as result of the national surveys in

2003-2004.

352290 307

239185

379 390

531461

Cou

ntry

Red

Riv

erD

elta

Nor

thea

st

Nor

thw

est

Nor

thC

entra

lC

oast

Sout

hC

entra

lC

oast

Cen

tral

Hig

hlan

ds

Sout

heas

t

Mek

ong

Riv

erD

elta

13

The income difference of quintiles between the poorest (North Central Coast) and the

richest (Southeast) regions increases from 1.31 times in Quintile 1 to 1.89 in Quintile 5.

That means the poverty situation is similar, but the richness has different levels in different

regions.

Table 5: Monthly per capita income by region and quintile (thousand VND)

Wholecountry

Quintile1

Quintile2

Quintile3

Quintile4

Quintile5

Quintile 5 :Quintile 1

Rural area 352 88 164 248 381 877 10.0

- Red River Delta 290 85 157 226 321 664 7.8

- Northeast 307 83 149 221 322 761 9.1

- Northwest 239 62 111 1687 272 584 9.4

- North Central Coast 185 54 103 152 214 405 7.5

- South Central Coast 379 103 187 272 404 928 9.0

- Central Highlands 390 68 148 257 478 1001 14.7

- Southeast 531 109 207 348 549 1439 13.2

- Mekong River Delta 461 112 208 311 488 1186 10.6

Max (Southeast) : Min (NorthCentral Coast)

1.73 1.31 1.39 1.58 1.70 1.89

The income distribution has not significant difference between regions. As data shows, in rural

areas of the whole country 20% households belonging to Quintile 1 possesses only 5,7% of total

income, while 20% households belonging to Quintile 5 possesses 46,8% of total income. The

poorer regions have smaller gap between the rich and the poor. In the poorest region the gap

between income share possessed by Quintile 5 and that of Quintile 1 is 6.8 times, while in the

richest regionthis index is 10.1.

Table 6: Income distribution (% total income)

Total Quintile1

Quintile2

Quintile3

Quintile4

Quintile5

Quintile 5 :Quintile 1

Rural area 100.0 5.7 10.4 14.9 22.2 46.8 8.3

- Red River Delta 100.0 6.4 11.9 16.0 22.4 43.3 6.7

- Northeast 100.0 5.9 10.6 14.7 20.8 47.9 8.1

- Northwest 100.0 5.5 10.5 15.7 23.8 44.4 8.0

- North Central Coast 100.0 6.2 12.5 16.8 22.5 42.0 6.8

- South Central Coast 100.0 6.5 10.9 15.2 21.9 45.4 7.0

- Central Highlands 100.0 4.2 8.5 13.8 25.3 48.2 11.4

- Southeast 100.0 4.8 9.1 15.2 22.3 48.6 10.1

- Mekong River Delta 100.0 5.6 10.2 14.1 21.9 48.2 8.6

Max - Min 2.3 4.0 3.0 4.5 6.6

14

2.3. Dwelling

Most of households in Vietnam ownes a house. In rural areas, almost all households have

their piece of residential land and house. In the urban areas 90% households ownes a

housing place (house, apartment or room), and only 10% rents dwelling [Vu Tuan Anh &

Nguyen Xuan Mai (2007)].

The types of dwelling are diversified by regions because of the difference climate

conditions, traditional housing habits of ethnic groups and living standards. For

identification of living standards, however, one can classify dwelling in some types by

criteria of solidity of construction and conveniences. There are three types of dwelling

which are used by localities for identifying living standards and poverty: (i) Permanent

houses, which consist of brick/betone houses (multi-storey or one-storey) and good

wooden houses; (ii) Semi-permanent houses (wooden, brick dwelling of low quality); (iii)

Temporary houses (bamboo houses, tent, dilapidated houses). Poor households owne

certainly the third type of dwelling, and many of them possess also the second type of

dwelling. Besides, a small number of young families and migrants do not have their owned

dwelling.

In the National Program for Poverty Reduction, supporting poor households in "rubing

out" temporary tents, degraded bamboo houses and constructing new dwelling is one

important activity, despite of small proportion of temporary dwelling left in localities. Til

the beginning of 2005, there were 4 provinces completed this objectives and in the

beginning of 2006, 17 provinces more achieved the goal of in "rubing out" temporary

dwellings.

There is still a noticeable number of households with temporary dwelling in the poor

regions. The proportion of households which owne temporary dwelling is 17.7% in the

whole country. The regions having high percentage of this dwelling type are Mekong River

Delta (29.3%), North Central Coast (24.4%), Southeast (24.2%), Northwest (23.6%),

Central Highlands (19.2%). Only the Red River Delta has low persentage of this dwelling

type (2.8%). Regions in the North have lower proportion of temporary dwelling than

regions in the South, because the climate in the North is cold in winter. (Table 7)

15

Table 7: Types of dwelling (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Proportion of households possessing dwelling (%)

* Multi-storeypermanent house

4.2 12.2 5.5 1.7 0.6 1.2 2.3 2.1 1.2

* One-storeypermanent house

31.2 55.3 23.7 24.9 44.8 59.0 11.5 20.9 17.1

* Semi-permanentdwelling

46.1 29.2 56.5 49.9 29.9 26.7 64.7 52.5 51.7

* Temporary dwelling 17.7 2.8 14.1 23.6 24.4 10.7 19.2 24.2 29.3

* No owned dwelling 0.7 0.5 0.2 0.0 0.3 2.4 2.3 0.3 0.7

Average living area of main dwelling (m2)

* Multi-storeypermanent house

80 51 94 82 78 91 120 177 147

* One-storeypermanent house

67 43 69 62 56 74 106 119 93

* Semi-permanentdwelling

82 29 124 69 47 51 70 87 78

* Temporary dwelling 55 29 80 65 46 29 34 53 54

Figure 6: Structure of dwelling types in regions

0%

20%

40%

60%

80%

100%

Cou

ntry

Red

Riv

erD

elta

Nor

thea

st

Nor

thw

est

Nor

thC

entra

l

Sou

thC

entra

l

Cen

tral

Hig

hlan

ds

Sou

thea

st

Mek

ong

Riv

er

No owned

Temporary

Semi-permanent

Permanent

16

Poor households and young families possess mostly temporary dwelling. Approximately

1/3 of households belonging to the quintile 1 (the poorest) have this dwelling type. The

proportion of this dwelling type decreases with increasing income. (Table 8)

The proportion of households possessing permanent dwelling increases with increase of

income. This tendency is available in all regions. The Red River Delta has the highest

proportion of this dwelling type with 55.8% households in Quintile 1 and 75.8% in

Quintile 5. In the other regions, the number of households having permanent dwelling has

trend to increase.

Table 8: Types of dwelling by regions and income quintile (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Permanent dwelling (%)

Total 35.6 67.5 28.9 26.7 45.9 60.7 14.4 22.8 18.5

* Quintile 1 25.8 55.8 19.4 24.7 36.5 48.5 6.4 12.2 7.7

* Quintile 2 29.5 67.0 19.8 21.9 36.0 53.5 7.9 23.3 10.4

* Quintile 3 33.6 66.5 25.0 25.9 46.0 62.0 9.3 19.4 16.7

* Quintile 4 40.0 72.6 35.3 26.0 51.3 63.5 19.2 26.1 21.7

* Quintile 5 49.3 75.8 45.1 34.9 60.0 75.5 29.2 32.6 36.0

Semi-permanent dwelling (%)

Total 46.1 29.0 56.4 49.5 30.6 24.2 67.9 52.3 51.8

* Quintile 1 40.2 35.0 51.3 20.5 24.5 29.0 69.3 37.8 37.9

* Quintile 2 47.6 31.3 62.2 51.4 32.5 31.5 64.0 46.1 50.0

* Quintile 3 50.8 32.0 61.9 57.1 34.0 23.5 70.1 61.7 58.1

* Quintile 4 47.5 24.4 56.3 63.0 32.7 21.0 66.0 55.6 58.5

* Quintile 5 44.4 22.3 50.5 55.5 29.5 16.0 69.8 60.2 54.3

Temporary dwelling (%)

Total 18.2 2.9 15.3 23.4 26.4 10.4 20.1 24.4 29.4

* Quintile 1 32.8 8.0 27.6 54.8 38.5 21.0 22.8 48.3 53.3

* Quintile 2 22.1 1.5 16.3 26.7 42.5 11.0 23.6 27.2 37.7

* Quintile 3 15.1 2.0 12.8 15.6 21.0 7.0 21.1 18.9 25.2

* Quintile 4 13.2 1.7 12.1 10.3 17.6 8.0 21.2 16.7 20.5

* Quintile 5 8.1 1.3 7.8 9.6 12.5 5.0 11.9 11.0 10.2

17

2.4. Ownership of durable consumer goods

Ownership of valuable consummer goods reflects level of household's living standards. It

also shows level of satisfaction of some basic needs, such as transportation, access to

information, access to living convenience.

Regarding transportation means, bicycle and motocycle are most popular individual

transport means of Vietnamese. 47.8% of rural households possessing at least a motocycle.

In average, there is one motocycle per two households, and every 100 people possesse 12

motocycles. The poorest region (North Central Coast) has the lowest number of per capita

motocycles.

More than ¾ of households possess at least a bicycle. The Mekong River Delta, where

water transport is more popular than land transport, has the lowest percentage of ownership

of bicycles. In average, every 100 people have 27 bicycles.

13.8% of households doesn't possess any motocycle or bicycle.

Regarding equipment for accessing to information, tiviset is popularly used by people.

76.6% of households possess a tivi, of which 62.9% have a color tivi and 13.7% have a

black-white tivi. This type of tivi is used mostly by the poor, or where there is no grid

electricity yet.

The difference between regions in ownership of tivi is significant. The poorest regions

have lower percentages, namely Northwest 59.2%, North Central Coast 65.0%, Northeast

72.1% and Central Highlands 72.8%. In the sametime, the percentage in other regions are

higher: Red River Delta 80,7%, South Central Coast 81,7%, Mekong River Delta 84% and

Southeast 87.1%.

In average, there are 35 video-audio equipments available per 100 people. The highest

number is in Red River Delta, Mekong River Delta and Southeast (about 40 pieces), while

the lowest number is in the poorest region (22 pieces in North Central Coast). 16.4% of

households still do not possess any video-audio equipment to access to information. As

ussual, the poor regions have higher proportion of households,which do not possess any

equipment,namely: Northwest 29%, North Central Coast 26%, Central Highlands 21.5%,

Northeast 20.5%; while the other regions have much lower proportion: South Central

Coast 13.7%, Red River Delta 11.1%, Mekong River Delta 10.6%, and lastly Southeast

7.7%.

18

Table 9: Possession of durable consumer goodsWhole

countryRed

RiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Percentage of households possessing goods (%):

* Radio receiver 36.2 31.0 30.1 37.3 26.9 38.6 37.1 47.7 45.3

* B&W tiviset 13.7 11.6 16.8 13.3 8.9 4.9 12.4 17.0 17.3

* Colour tiviset 62.9 69.1 55.3 46.0 56.1 76.8 60.4 70.1 66.8

* Video, VCD, DVDplayer

32.3 40.4 26.0 22.2 14.2 32.5 40.0 33.4 39.5

* Audio equipments 13.9 7.7 8.9 9.6 5.4 18.4 21.6 17.1 23.3

* Electric fan 76.9 84.6 76.1 59.9 82.2 89.2 49.9 80.2 79.3

* Sewing machine 12.9 8.8 11.3 15.5 3.2 6.8 4.6 6.1 30.1

* Fridge 9.0 12.6 7.1 3.3 3.6 8.1 9.6 12.0 11.3

* Motocycle 47.8 39.3 45.6 40.8 33.5 66.3 56.5 64.3 47.8

* Bicycle 77.4 83.5 75.6 78.2 88.1 81.3 72.6 82.4 68.0

* Telephone 11.6 10.7 7.6 4.2 6.0 14.1 10.6 15.9 19.0

* Washing machine 1.1 0.6 0.4 0.1 0.3 1.2 3.7 1.7 1.4

* Computer 2.3 5.4 0.6 0.0 0.3 1.7 3.5 3.0 2.7

Households notpossessing any video-audio equipment (%)

16.4 11.1 20.5 29.0 26.0 13.7 21.5 7.7 10.6

Households notpossessing anymotocycle, bicycle(%)

13.8 12.1 16.8 15.5 8.0 10.7 11.8 4.3 19.5

Number ofmotocycles per 100people

12.0 10.2 11.4 8.6 7.1 18.4 13.7 17.0 12.3

Number of bicyclesper 100 people

27.1 36.4 26.9 25.0 29.6 29.8 21.9 28.2 20.9

Number of video-audio equipments per100 people

35.0 39.1 31.0 25.9 22.3 39.2 35.9 40.6 40.5

Possession of durable consumer goods is also varried by households belonging to different

income quintiles. Indicators of possession of three goods - tiviset, motocycle and telephone

- show clearly correlation between living standards and satisfaction of needs for transport,

access to information, and communication. (Table 10).

Posession of tiviset is saturated by the rich households, while proportion of poor

households (belonging to quintiles 1 and 2), who do not have any tiviset is still high (40-

50% in the poor regions).

19

Possession of motocycle is at 60-80% by the richer quintiles and at only 20-30% by the

poor quintiles.

20-30% households of the richer quintile have a telephone, while it is a rare equipment for

the poor households.

Table 10: Possession of durable consumer goods by income quintile (% households)Whole

countryRed

RiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Tiviset (%)

Total 76.6 80.7 72.1 59.2 65.0 81.7 72.8 87.1 84.0

* Quintile 1 58.5 66.8 52.4 39.0 45.0 72.5 58.4 65.6 61.5

* Quintile 2 70.2 77.5 64.3 46.6 55.0 82.5 56.7 85.0 79.6

* Quintile 3 78.2 82.8 72.0 59.2 68.5 78.5 71.1 93.9 88.1

* Quintile 4 86.1 87.8 84.0 75.3 73.4 85.0 86.2 93.9 93.3

* Quintile 5 90.1 88.8 88.1 76.0 83.0 89.5 91.6 97.2 97.7

Motocycle (%)

Total 47.8 39.3 45.6 40.8 33.5 66.3 56.5 64.3 47.8

* Quintile 1 26.1 21.5 24.6 16.4 17.0 49.0 35.6 35.0 21.5

* Quintile 2 37.5 30.0 32.2 31.5 25.0 59.5 41.4 60.0 37.5

* Quintile 3 45.8 36.5 41.7 36.1 28.0 64.5 56.4 71.1 46.7

* Quintile 4 59.0 49.1 55.4 49.3 44.7 77.0 69.0 76.7 61.8

* Quintile 5 70.6 59.5 74.3 70.5 53.0 81.0 80.2 78.5 71.7

Telephone (%)

Total 11.6 10.7 7.6 4.2 6.0 14.1 10.6 15.9 19.0

* Quintile 1 4.2 4.5 1.7 1.4 2.0 5.5 5.0 3.3 8.1

* Quintile 2 6.3 7.3 3.0 1.4 3.5 6.5 4.4 12.8 10.2

* Quintile 3 9.7 7.5 6.3 1.4 2.5 9.0 8.3 17.2 18.8

* Quintile 4 14.7 13.5 10.6 6.2 6.5 15.5 16.7 22.2 22.3

* Quintile 5 23.0 21.0 16.8 11.0 15.5 34.0 18.3 23.8 35.6

2.5. Education

We used two indicators for analysis of education situation: illiteracy and child school

enrolment.

The rate of literacy calculated for people from 6 and over is 94.5% in the whole country.

Accordingly, illiteracy rate is 5.5%. The mountainous and poor regions have higher

illiteracy rate, namely: Northeast 6.9%, Central Highlands 9.1%, and Northwest 14.9%.

20

The percentage of people who have primary education (from grade 1 to grade 5) is 31.5%,

lower secondary (6-9 grades) 41.6%, upper secondary (10-12 grades) 18.8%, college and

university 1.8%.

Table 12: Education levels of 6 year-old and over population (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighland

s

Southeast

MekongRiverDelta

Illiteracy 5.5 3.0 6.9 14.9 3.8 3.0 9.1 2.5 4.4

Primary (1-5 grades) 31.5 19.5 25.0 42.0 29.6 34.7 36.3 32.7 40.5

Lower secondary (6-9grades)

41.6 54.0 43.8 31.0 49.6 36.8 34.3 38.3 36.2

Upper secondary (10-12 grades)

18.8 20.8 21.3 11.3 15.3 21.9 17.7 23.1 16.4

Vocational secondary 0.9 1.1 1.2 0.4 0.5 1.0 0.6 1.0 0.7

College, university 1.7 1.6 1.8 0.4 1.1 2.4 1.9 2.1 1.8

Regarding child's schooling, there is a small pcentage of children who give up study or not

go to school at all. According to survey data, 5.5% of children in schooling age (6-14) does

not attend study. The population in the North part of the country has a tradition of paying

more attention on giving education to children. In the Red River Delta, the number is the

smallest, only 2.3%. The Northeast and North Central Coast regions, despite of being poor

regions, have also low percentages (3.4% and 3.6%). The other regions have significantly

higher percentages than the national average level, namely South Central Coast 6.2%,

Mekong River Delta 7.3%, Southeast 7.5%, Northwest 8.2% and Central Highlands 9.5%.

Poverty is an important reason explaining why children do not go to school. 46.3% of

children, who have dropped-out the study, are caused by such reasons like lack of labour in

the family, too high costs of schooling, etc. Invalidity, serious sickness caused 14.5% of

not-schooling children. 27.6% of total number of not-schooling children has dropped out

the study because they have bad results in study. The awareness of parents and children on

necessarity of education caused 9.5% of the cases.

The structure of causes of child's not-schooling is different among the regions.

21

Table 13: Child's schoolingWhole

countryRed

RiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Percentage of 6-14years old children notattending school (%)

5.5 2.3 3.4 8.2 3.6 6.2 9.5 7.5 7.3

* Boys 3.0 1.8 1.9 3.0 2.0 2.9 5.3 4.9 3.8

* Girls 2.5 0.5 1.5 5.3 1.6 3.3 4.3 2.7 3.5

Why not go to school (% responses)

* Lackof labour infamily

21.7 8.8 27.9 29.8 8.3 16.3 34.9 19.4 19.0

* High costs ofeducation

24.4 17.6 21.3 27.7 41.7 40.8 17.5 22.6 19.0

* Bad results in study 27.6 35.3 19.7 23.4 13.9 18.4 28.6 32.3 37.2

* Invalidity, sickness 14.5 38.2 13.1 6.4 33.3 20.4 3.2 16.1 9.1

* No neccessarity tostudy

9.5 0.0 18.0 10.6 2.8 0.0 15.9 6.5 10.7

* Other 2.3 0.0 0.0 2.1 0.0 4.1 0.0 3.2 5.0

2.6. Health

It's difficult to measure the level of satisfaction of people's basic needs in health care. In

CBMS we use some simple indicators which indirectly reflect situation of health care of

households. They are basic household's sanitary facilities, such as supplying safe drinking

water, using sanitary toilet, having a bathroom.

Vietnam's Goverment consideres supply of safe water for population and securing sanitary

living environment as one of the national prioritised targets. The National Targeted

Program of Safe Water Supply and Rural sanitary Environment had invested 7,000 billion

VND (approx. equivalent to 0.5 bil. USD) in the period of 1999-2005. Until the end of

2005, 68% of rural population accessed to safe drinking water. Still 32% of rural

population used unsafe water (from rivers, ponds, lakes, etc.). In the period of 2006-2010,

this National Targeted Program has planned to invest 22,600 bill. VND (1.4 bil. USD).

Poor regions belong to prioritised areas of this program.

Sources of water are diversified in the regions. One used to considere piped water, rain

water, water from deep-driled wells as certainly safe. Water from dug wells may be safe or

unsafe depending on concrete conditions of locations and seasons. Water from natural

surface sources like rivers, lakes, ponds, streams is considered as unsafe.

22

Accepting working definition that water from dug well is considered as unsafe, we can

remark that only 50% of population access to safe water. In the mountainous regions like

Northwest, Northeast, Central Highlands and in the Southeast, more than 70% of

population still use unsafe water sources. (Figure 7).

Table 14: Access to safe water for dringking and cooking (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Piped water, publicreservoir

8.0 10.0 7.3 1.8 9.8 5.5 0.1 11.2 11.3

Rain water 21.6 65.0 0.3 13.8 1.3 0.0 1.0 1.4 42.3

Deep-drilled well 20.0 20.4 15.7 1.0 37.4 27.0 1.4 13.9 29.2

Dug well 37.7 4.1 55.5 43.8 45.8 67.0 91.5 61.7 1.3

Pond, lake, river 12.6 0.5 21.1 39.6 5.7 0.6 6.0 11.8 15.9

Note: Sum of all types might be over 100% because one household may use several water

sources.

Figure 7: Access to safe drinking water (%)

The poor households face more difficulties in access to safe water sources. Taken two

types of sources: deep-drilled well and natural surface sources and comparing percentage

of users by income quintiles, one can see that the poorer quintiles have smaller percentages

using the deep-drilled water, for which households have to invest a certain amount money

0%10%20%30%40%50%60%70%80%90%

100%

Cou

ntry

Red

Riv

erD

elta

Nor

thea

st

Nor

thw

est

Nor

th C

entra

lC

oast

Sout

h C

entra

lC

oast

Cen

tral

Hig

hlan

ds

Sout

heas

t

Mek

ong

Riv

erD

elta

Unsafe

Safe

23

to equip; and they have larger percentages using the "free-of-charge" but unsafe water from

natural surface sources. (Table 15)

Table 15: Access to safe and unsafe water by income quintile (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Deep-drilled well (%)

* Quintile 1 17,9 17,8 14,3 0,0 29,0 26,0 1,0 10,6 29,4

* Quintile 2 18,4 19,8 13,9 1,4 30,0 23,5 0,5 12,8 30,2

* Quintile 3 20,0 22,0 13,3 0,7 37,5 27,5 1,0 14,4 31,7

* Quintile 4 21,7 20,2 14,9 1,4 43,2 30,5 3,0 16,7 33,8

* Quintile 5 24,2 22,8 19,2 2,1 48,5 30,0 1,5 15,5 38,3

Pond, lake, river (%)

* Quintile 1 11,0 0,5 18,7 52,1 5,0 0,5 9,9 10,6 6,0

* Quintile 2 10,3 0,5 18,9 41,8 3,5 0,0 15,8 11,1 4,0

* Quintile 3 8,1 0,3 16,1 40,1 3,5 0,5 3,4 11,7 1,5

* Quintile 4 6,8 0,2 11,3 37,0 2,5 0,5 0,5 11,7 3,1

* Quintile 5 5,0 0,3 8,2 30,8 1,5 0,0 0,0 10,5 1,0

There is only 41.5% of households in the whole survey sample possessing a sanitary toilet.

This figure is similar to the result of the health survey done by the Ministry of Health in

2003. Among regions, the Red River Delta has the highest percentage (64.7%), then

follow South Central Coast (54.9%), Southeast (49.1%), Northeast (46.3%), North Central

Coast (45.5%). The Northwest has only 22.9%, Central Highlands 23.3%, and Mekong

River Delta 20.0%. These figures are too low, compared with the target of the National

Strategy of Safe Water Supply and Sanitary Environment, in which 70% of households

will possess sanitary toilets till 2010.

Table 16: Types of toilet (%)

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

Sanitary types 41.5 64.7 46.3 22.9 45.5 54.8 23.3 49.1 20

Unsanitary types 46.9 31.7 42.3 71.2 30.8 27 53 49.1 69.2

No toilet 11.6 3.6 11.5 5.9 23.7 18.1 23.6 1.8 10.8

24

III. Proposal of a Composite Poverty Index

Poverty rate measured by value indicators such as income and expenditures of households

and individuals gives a general picture of poverty, but doesn't reflect concrete aspects of

living in which people are in shortage. Poverty is a multi-dimensional phenomenon.

Poverty is foremostly unsatification of human basic needs such as food, cloth, housing,

education, healthcare, information, etc. However, if indicators of basic needs are used

separately, one cannot define overall poverty rate, as well as compare poverty of different

regions and in times.

In the CBMS, which was piloted by the Vietnam research project, poverty is

comprehensively reflected by a set of indicators that include both value indicators (income)

and the basic households needs (e.g., food intake, clothing, housing, transportation,

education, healthcare). A study has been also done for identifying a composite poverty

indicator for Vietnam [L.M. Asselin (2002); L.M. Asselin & Vu Tuan Anh (2005)]. In this

study, eight simple non-monetary, categorical indicators of human and physical assets

developed in CBMS research in Vietnam, have been analysed and aggregated in a

composite indicator using the factorial technique. These indicators reflect the following

groups of basic needs of population: (1) income generation (underemployment); (2) health

(chronic sickness, access to safe drinking water, sanitary facilities); (3) education (adult

illeracy, child under schooling); (4) housing (types of dwelling); and (5) access to

information (ownership of tiviset, radio receiver).

The comparison of this multidimensional approach to poverty measurement with the

moneymetric one base on total household expenditures shows that the CBMS type

indicators present a strong analytical potential for multidimensional poverty analysis, being

complementary to the more standard moneymetric analysis. In addition, due to their low

cost, they should be looked at to meet the objective of regularly producing largely

disaggregated poverty profiles for a more efficient monitoring of poverty reduction

policies and programs.

During implementation of the proposed composite poverty index in localities, a problem is

bared itself. It seems that the factorial analysis technique, precisely the multiple

correspondence analysis (MCA) is implemented with difficulties, since local partners,

especially people from district and commune levels are not able to understand and to use it.

In order to follow one of fundamental principles of CBMS, namely the simplicity of

25

indicators and methods, we propose to develop a composite poverty index,which is

constructed withour weighting primary indicators. In practise, this type of composite index

has been implemented widely. Human Development Index (HDI), Human Poverty Index

(HPI), MDGs Index and some other indexes developed by UNDP belongs to this type. The

Human Poverty Index fordeveloping countries (HPI-1), for example, attempts to capture

deprivations in three essential dimensions of human life already reflected in the HDI. It

includes in itself three components:

- Longevity - measured by the proportion of the population not expected to survive to the

age of 40 years (P1).

- Knowledge - measured by the adult illiteracy rate (P2).

- Standard of living - a composite value measured by the proportion of the population

without access to clean water(P31), health services (P32), and the proportion of children

under the age of 5 years who are underweight (P33).

The composite variable P3 is constructed by taking a simple average of the three variables

P31, P32 and P33. Thus: P3= (P31 + P32 + P33) / 3

The HPI-1 is computed by the following formula:

HPI-1 = [1/3(P13+P2

3+P33)]1/3

The composite MDGs index includes 8 goals divided into 18 targets with 48 indicators.

Despite of it's complexity, this index is calculated by a similar method as that of HPI.

Implementing this computation method of composite indexes, we propose to compute a

CBMS Composite Poverty Index (CBMS-CPI), using set of CBMS indicators, namely:

1- Food poverty: percentage of households which have income below food poverty line

(P1).

2- Dwelling poverty: percentage of households which have temporary dwelling and not

have owned dwelling (P2).

3- Information poverty: Percentage of households which do not possess any audio-video

equipments (P3).

4. Communication poverty: Percentage of households which do not possess any motocycle

and bicycle (P4).

5. Knowledge poverty: Simple average of adult illiteracy rate and childt under-schooling

rate (P5 = ½[P51 + P52])

26

6- Health poverty: Simple average of percentage of households, which do not access to

safe drinking water and not possess a sanitary toilet (P6 = ½[P61 + P62]). In a better option,

where data is available, we can add other indicators, which reflect fundamental situation of

health poverty, like percentage of child malnutrition.

The CBMS Composite Poverty Index is simple average of the 6 above poverty indicators:

CBMS-CPI = 1/6(P1+P2+P3+P4+P5+P6)

CBMS-CPI has some following advantages:

- It containes the major aspects of human poverty, therefore it is a multi-dimensional

poverty indicator. One can use it for measuring and comparing poverty across time and

regions.

- Major aspects of poverty reflect most targets of the national poverty reduction program.

Therefore one can use CBMS-CPI for monitoring poverty reduction activities and

programs.

- Computing method is very simple and easily to be understood, so that everybody at

grassrot levels can use.

- Excepts income, all rest indicators base on simply to be collected data. However, income

is the data, which communities have to colect and to monitor regularly poverty by

MOLISA poverty line. This poverty line is approximately at the level of food poverty.

Therefore we can use this data for pomputing CBMS-CPI.

To test CBMS-CPI, we use CBMS data for computing and comparing CBMS-CPI of

regions. (Table 17 and Figure 8)

Comparing CBMS-CPI with income poverty P1, one can see how the multi-dimensions

poverty and purely value poverty are different. (Figure 9).

27

Table 17: Computing CBMS-CPI of regions

Wholecountry

RedRiverDelta

North

east

Northwest

NorthCentralCoast

SouthCentralCoast

CentralHighlan

ds

Southeast

MekongRiverDelta

P1: Incomepoverty *

15.5 12.9 23.2 46.1 29.4 21.3 29.2 6.1 15.3

P2: Dwellingpoverty

18.4 3.3 14.3 23.6 24.7 13.1 21.5 24.5 30

P3: Informationpoverty

16.4 11.1 20.5 29.0 26.0 13.7 21.5 7.7 10.6

P4:Communicationpoverty

13.8 12.1 16.8 15.5 8.0 10.7 11.8 4.3 19.5

P5: Knowledgepoverty

5.5 2.7 5.2 11.65 3.7 4.6 9.3 5.0 5.9

Illiteracy 5.5 3.0 6.9 14.9 3.8 3.0 9.1 2.5 4.4

Childunderschooling

5.5 2.3 3.4 8.2 3.6 6.2 9.5 7.5 7.3

P6: Healthpoverty

54.4 20.0 65.2 80.3 53.0 56.4 87.1 62.2 48.6

No safe water 50.3 4.6 76.6 83.4 51.5 67.6 97.5 73.5 17.2

No sanitarytoilet

58.5 35.3 53.8 77.1 54.5 45.1 76.6 50.9 80.0

CBMS-CPI 20.7 10.4 24.2 34.4 24.1 20.0 30.1 18.3 21.7

Note: * Data of poverty rates is taken from GSO national household survey in 2006 [GSO (2008)].Poverty rates have been measured by monthly average income per capita according to the lateststandard of the Government for the period 2006-2010 with different standards as follows: 260thous. VND for urban; 200 thous. VND for rural (excluding effect of price index).

Figure 9: Comparison of income poverty rate and CBMS-CPI

05

101520253035404550

Cou

ntry

Red

Riv

erD

elta

Nor

thea

st

Nor

thw

est

Nor

thC

entra

lC

oast

Sout

hC

entra

lC

oast

Cen

tral

Hig

hlan

ds

Sout

heas

t

Mek

ong

Riv

er D

elta

Income poverty CBMS-CPI

28

Figure 8: Regional CBMS-CPI and it's components

IV. Conclussions1. Regional disparity in poverty is one of the key challenges, which Vietnam faces on it's

current development path. Despite of remarkable achievements in economic growth and

poverty reduction, regional disparity and social inequality may be strong factors hampering

the socio-economic progress in the future. Vietnam has to pay more attention on policies of

inequality reduction towards regions, social groups and ethnic groups.

2. CBMS can be used as an appropriate tools for poverty monitoring, especially by local

authorities, social organisations and communities. Using CBMS data, one can analyse

diversified aspects of human life, including poverty, and do comparison across regions and

times.

29

3. A CBMS Composite Poverty Index reflects the approach of multidimensional poverty. It

can be used by local communities for analysis of different aspects of poverty, as well as for

comparison of multidimensional poverty accross regions, localities and times. A

Composite Poverty Index constructed by a simple method and based on community-based

survey data is feasible to be implemented widely in poverty monitoring and evaluation of

poverty reduction activities.

REFERENCES

• Asselin Louis-Marie (2002). Multidimensional Poverty. Theory. IDRC. in MIMAPTraining Session on Multidimensional Poverty. Quebec. June 2002.

• General Statistical Office. GSO (2008a). The 2007 Population Change and FamilyPlanning Survey: Major Findings. Hanoi.

• General Statistical Office. GSO (2008b). Statistics Yearbook of Vietnam 2007. Hanoi.

• Louis-Marie Asselin & Vu Tuan Anh (2005). Multidimensional Poverty Monitoring: AMethodology and Implementation in Vietnam. Vietnam’s Socio-Economic DevelopmentReview. No. 41. Hanoi. March 2005.

• Louis-Marie Asselin & Vu Tuan Anh (2008). Multidimensional Poverty and MultipleCorrespondence Analysis. In "Quantitative Approaches to Multidimensional PovertyMeasurement" edited by Nanak Kakwani and Jacques Silber. Palgrave Macmillan. NewYork.

• Nguyen Xuan Mai & Vu Tuan Anh (2007a). Reduction of Urban Poverty. Vietnam’sSocio-Economic Development Review. No. 51. September 2005. Hanoi.

• UNDP. Human Development Reports.

• Vu Tuan Anh & Vu Van Toan (2004). Implementaion of the community-based povertymonitoring in Vietnam. Vietnam’s Socio-Economic Development Review. No. 39. Hanoi.September 2004.

• Vu Tuan Anh & Nguyen Xuan Mai (2007). Socio-Economic Changes of Households.Publisher of Social Sciences. Hanoi.