DSpace at VNU: Inequality income in Hanoi''s peri-urban areas : evidence from household survey data

DSpace at VNU: Inequality income in Hanoi''''s peri-urban areas : evidence from household survey data tài liệu, giáo án, bà...

INEQUALITY INCOME IN H A N O I’S PERI-URBAN AREAS:

EVIDENCE FROM H O U SEH O L D S URV E Y DATA

Tran Quattg Tuven

Steven Lim

1 In tro d u c tio n

A number o f studies indicate that d ive rsifica tion in nonfarm activities, including wage and self-em ploym ent, results in poverty alleviation and therefore increases liv in g standards for a large number o f Vietnamese rural households

A ccording to W B , income from wage em ploym ent played the most im portant role

in income grow th o f the poor in 2004 Wage income contributed almost ha:f o f overall income grow th o f the poorest households between 2002 and 2004 ( V B , 2006) Van de W alle and Cratty (2004) found that the p ro b a b ility o f falling in poverty is substantially higher among households w ho do not participate in non:arm self-em ploym ent activities Also, their evidence based on cross-sectional data indicated that there is a causal relation between d iv e rsifica tio n out o f a g ricu tu re and poverty reduction M oreover, Pham, B ui, and Dao (2010) estimated that on

average and ceteris p a rib u s , a household that moves from a pure a g ricu tu re

production unit to a pure non-agriculture one w ill increase its expenditure per capita, and this outcom e tends to steadily increase over tim e

However, w h ile d iversification in nonfarm income sources has an increasing effect on household w elfare, it exaccrbates incom c inequality in Vietnam U ỉin g dala sets from tw o V ietnam Household L iv in g Standard Survey (V H L S S ) in '993 and 1998, G allup (2002) shows that between 1993 to 1998, wage em ploym ent contributes an approxim ately same amount to overall incom e inequality as ith e r non farm em ploym ent sources (household enterprise and remittances, m a h lv )

A g ricu ltu ra l income, on the contrary, has a decreasing effect on overall income inequality A s im ila r fin d in g from V H I,S S 2004 is also found in Van Cao and A kita

* MA University o f E con om ics and Business, Vietnam National University, Hanoi PhD student at the School o f Management, the University o f Waikato, N e w Zealand.

** Ph.D Waikato Management School, the University o f Waikato, N e w Zealand.

IN F Q U A IIT Y INCOM E IN H A N O I’S PFRI URBAN A R EAS .

(2008), w hich shows that agricultural incomc served to decrease overall inequality whereas nonfann self-em ploym ent and wage income served to increase overall inequality

Nevertheless, when looking at incomc distrib u tio n w ith in rural and urban sectors, Van Cao and A k ita (2008) found that the contribution o f each income source to overall inequality is quite different between tw o sectors W hile agricultural income is inequality-decreasing in the urban sector, it is inequality- incrcasing in the rural sector Nonfarm self-em ploym ent seemed to be not related to inequality in the urban scctor but contributed to the rise in overall inequality in the rural scctor Wage incom e is found to decrease overall inequality w ilh in each sector F in a lly , other income sources served to increase overall inequality in both see tors As noted in Babatunde (2008), by decomposing income inequality one can clucidatc the structure and dynamics o f incomc w ith in d iffe re nt socio-econom ic groups M easuring die contribution o f each income source to overall inequality coupled w ith re a lizin g the relation between socio-economic characteristics and overall inequality, can be useful to analyzing economic policies and designing poverty alleviation programmes,

U sing a novel data set from our own household survey in Hanoi's peri-urban areas, this study contributed new empirical findings to the existing literature on income inequality in Vietnam 's peri-urban scctor This paper is structured as

follows: the subsequent section provides a brief description about the study area

M ethodology is m entioned in Section 3, and estimation results are discusses in Section 4, fo llo w e d by conclusion in Section 5

2 D e s c rip tio n o f study area

Our research was conducted in Hoai Due, a peri-urban district o f Hanoi Hoai Due is located nn the northwest side o f Hanoi, 19 km from the Centra] Business

D istrict (C R D ) (W B , 2011) The district has an extrem ely favourable geographical position, w hich is surrounded by various im portant roads namely Thang 1 one highw ay (the co u n try’ s longest and most modem highway) National Way 32 and

in close p ro x im ity to industrial zones, new urban areas and Bao Son Paradise Park (the biggest entertainm ent and tourism complex in N orth Vietnam ) Consequently,

in the period 2006-2010, around 15,600,000 rrT o f farmland were com pulsorily acquired by the State fo r 85 projects (LH , 2010)

Before r 1 August 2008, Hoai Due used lo he a d is tric t o f Ha Tay Province a neighbouring province o f Hanoi, which was merged into Hanoi on P ' August 2008 The; cistrict occupies 8,247 hectares o f land, o f w hich agriculture land accounts fo r

V IỆ T N A M H Ọ C - K Ỷ Y Ế U H Ộ I T H Ả O Q U Ổ C T É L À N T H Ứ T ư

4 272 hectares and 9 ] percent o f this area is used by households and individuals (H oai Due D is tric t People's C om m ittee, 2010a)1 There are 20 adm inistrative units under the district, in clu d in g 19 communes and 1 town Hoai Due has around 50400 households w ith a population o f 193,600 people In the w hole district, the

em ploym ent in the agricultural sector dropped by around 23 percent over the past dccade Nevertheless, a sig n ifica n t proportion o f em ploym ent has s till remained in agriculture accounting fo r around 40 percent o f the total em ploym ent in 2009 The corresponding figures fo r industrial and services sectors are 33 and 27 percent, respectively (Statistics Departm ent o f Hoai Due D istrict, 2010) P rior to r l August

2008 Hoai Due used to be the richest d istrict in Ha Tay Province (M O N R E , 2(07)

In 2009, Hoai Due GDP per capita reached 15 m illio n V N D (H oai Due District People's Com m ittee, 2010b), w hich is less than h a lf o f H a n o i’ s average (32 m i lion

V N D ) (Vietnam Governm ent Web Portal, 2010)

3 Methodology

3.1 D ata collection

Adapted from GSO (2006), a household questionnaire was designed to gather

a set o f quantitative data on household demographic characteristics and income sources from various econom ic activities First, six communes were randomly selected2 and then 80 households, were random ly selected from each commune for a target o f sample size o f 480 The survey was carried out from A p ril to June 2010, and 477 households were successfully interviewed

3.2 Data analysis

Income inequality can be measured using various methods (Babatunde, 2008)

Am ong the d iffe re nt ways o f inequality measurement, the G in i coefficient is popularly used to measure the disparity in the distribution o f income, consumption, and other w elfare indicators (López-Feldman, 2006) F o llo w in g Van Den Berg and

K um bi (2006), we examine the relation between incom c sources and incomc

Y itzh a ki 1985; Shortocks, 1982) A ccording to López-Feldm an (2006), Le-man

1 As laid down in the current l a n d Law o f Vietnam, land is not privately ow ned because it is the collective property o f the entire people, which is representatively ownct and administrated by the State and the land use rights are to be granted to individuals, households, enterprises and other organizations (National Assembly o f Vietnam, 2003'

2 Six communes include Song Phuong, Lai Yen, Kim Chung, An Thuong, Due Thuon>, and Van Con.

1NEQUAII TY INCOME IN HANOI'S P E R I-U R B A N AR EAS

and Y itzhaki (1985) extended the results o f Shorrocks (1982) and indicate that the

G in i coefficient o f total incom c inequality (G) can he denoted as:

where Sk represents for the share o f income source * in total income, Gt is the G ini coefficient o f the income distribution from source *, and Rk is the correlation coefficient between incom e from source * and w ith total income Y Babatunde (2008) shows that GkRk is known as the pseudo-Gini coefficient o f income source *

w h ile the share or contribution o f income source * to lota] income inequality is expressed as:

A s shown in Stark, T a y lo r and Yitzhakj (1986), the income source elasticity o f inequality indicates the percent change in the overall G in i co e fficien t resulting from

a ] percent change in incom e from source * is expressed as:

Where G is the overall G in i coefficient prior to the incom e change As noted

by Van Den Berg and K u m b i (2006), equation (3) is the difference between the share o f source * in the overall G ini coefficient and its share in the total income (Y )

It should he noted {hat the sum o f income source elasticities o f inequality should be zero, w hich means that i f a ll the income sources changed by same percentages, the overall G ini coefficient ((7) w ould remained unchanged

4 E s tim a tio n results

4.1 B a c k g ro u n d o f household income and incom e-generating activities

In previous studies on the decomposition o f income inequality in Vietnam household income is often disaggregated into various sources, including wage income, nonfarm self-em ploym ent income, agricultural income and other income

K

(1)

k=l

(3)

V IỆ T N A M H Ọ C - K Ỷ Y Ế U H ộ ] T H Ả O Q U Ó C T Ế L À N T H Ủ T ư

(.Adger, 1999; G allup, 2002; Van Cao & A k ita , 2008) H ow ever, wage earners often

w ork for various types o f em ployers in Vietnam Nguyen (2010) divided wage workers in V ietnam in to tw o groups, including inform al wage w orkers and formal wage workers Form al wage earners are those w ho w ork fo r other households, often

do not have a labour contract and not participate in social insurance In fo rm a l wage earners are those w o rk in g fo r the State, enterprises and other organizations, and they earned a much higher wage incom e than inform al wage earners F o llo w in g this approach, we further brake down wage income into tw o sub-categories, namely inform al wage and form al wage income Household incom e, therefore, can he contributed by five d iffe re nt sources (Table 1)

Table I : H ousehold incom e by source

crop and livestock production

2 Non-farm Self-employment Self-employment in household businesses (excluding

farm production)

3 Informal wage work Wage work that is unstable, irregular, and without a

formal labour contract, informal wage earners are often manual workers and hired by individuals or households

4 Formal wa%e work

* Wage work that is regular and relatively stable in factories, enterprises, state offices and other organizations with a forma] labour contract and often require skills and higher level o f education

5 Other income Income from other sources such as public and private

transfer, loan interest, house or land rental, etc

Source: Survey data and authors’ compilation from Becker (2004), Maxwell, el a]

(2000), Cling et a] (2009), and Nguyen (2010)

T a b lc 2 provides background in fo rm a tio n about household incom e shares

by source It also indicates how m uch per household and per capita incom e by source The results show that o v e rw h e lm in g m a jo rity o f surveyed households

INEQUALITY INCOME IN H AN O I'S P E R I-U R B A N A R EAS

(around 84 percent) d erive income from fa rm in g (c ro p and live sto ck

p ro d u ctio n ), w h ic h , how ever, ju s t accounts fo r about 27 percent o f to ta l incom e

on average T h is suggests farm ing has rem ained im p o rta n t in term s o f food security and cash incom c to some extent M any households have continued rice

c u ltiv a tio n as a source o f food supply w h ile others produced vegetables and fru its to supply fo r H a n o i’ s urban markets The com m on types o f crop plants consist o f cabbages, tomatoes, w aier m orning g lo ry and va rio u s kin d s o f heans, oranges, g rapefruits and guavas, etc A n im a l husbandry is m a in ly undertaken

hy p ig or p o u ltry breeding sm all-size farms or cow g razing households These

a c tiv itie s , how ever, have considerably declined due to the spreading o f cattle diseases in reccnt years

A lm ost all surveyed households (95 percent) participate in nonfarm activities and income from these sources contributed about two thirds o f total income on average A m ong these activities, informal wage income accounts fo r about one fourth o f total income w ith a participation rate reached around 41 percent This incomc source is often earned from manual labour jobs and the com m on types o f such jobs are carpenters, painters, building workers and other various kinds o f casual jobs Such jobs are often hired by individuals or households w ith low and unstable income, and w ith o u t labour contracts Those who undertake these jobs have be low-average education and are younger Sim ilar figures are observed fo r the case o f non farm self-em ploym ent income About 43 percent o f the household sample reported engaging in non-famn household businesses, and on average around

26 percent o f total income is contributed by this a ctivity Such businesses constituted small-scale trade or production units, using fa m ily labour The households’ business premises are m ainly located at their own houses or residential land plots that have a p rim e location for opening a shop, a w orkshop or a small restaurant

A b o u t 28 percent o f sample households received incom e from form al wage

w ork, accounting fo r 18 percent o f total income on average F o rm a l wage earners are often employees w ho w ork in enterprises and factories, state offices

or other organizations Such jobs are ofien h ig h ly paid w ith stable incom e and labour contracts Those undertaking these jobs tend to have a m uch higher education level and are younger F inally, about one th ird o f surveyed households received other income, in c lu d in g private transfer (g ift and rem ittances) and public transfer (pension and social assistances), rental incom e and interest incom c, and so on T h is source, however, contributes o n ly 6 8 percent o f total incom e on average

V IỆ T N A M H Ọ C - K Ỷ Y Ế U H Ộ I T H À O Q U Ố C T Ế L À N T H Ứ T ư

Table 2: Composition of income, and household participation

in different activities

Per household

Per capita

Share o f total Income (%)

Participation rate (%)

Unit: 1,000 VND Nonfarm income “ (A+B+C) Number o f observations: 477 Standard deviation in parentheses

Source: Authors' own calculation from the survey

4.2 G ini coefficients f o r income inequality

Figure 1 presents the disừibution o f incom e sources over incom e quintilcs As

compared to households in the higher incom e strata (4 and 5), the lo w er incom e

INEQUALITY INCOME IN HANO I'S P E R I-U R B A N AREAS

quintilc households (1 and 2 ) h a \o a higher share o f farm incom e, whereas those in the rich cr groups have a higher share o f non farm s e l f - e m p l o y m e n t and form al wage

incomes This suggests that incomc shares by source arc closely associated w ith the

in c o m e d is tr ib u tio n ; s p e c ific a lly th e re is a positive a s s o c ia tio n b e t w e e n d ie n a n f a r m

s c lf-c m p lo y m e n t in c o m e share, fo rm al w age in co m e share an d p e r c a p ita in c o m e ,

but a negative correlation between the farm and inform al w age incom e shares and per c a p ita i n c o m e

F i g u r e 2 s h o w s t h e d i s t r i b u t i o n o f i n c o m e s o u r c e o v e r f a r m l a n d

h o ld ing s A s reve ale d in th is figure, households in the h ig h e r l a n d h o l d i n g stratu m s have a m uch h ig h e r percentage o f farm in c o m e hut have a lo w e r

share o f n o n fa rm s e lf-e m p lo y m e n t, form al wage incom es and o th e r incom e

By contrast, the lo w e r la n d h o ld in g stratum households re ce ive m ore incom e from nonfarm s e lf-e m p lo y m e n t and manual labour jo b s , w h ic h im p lie s that households w ith lim ite d farm land m ig h t be pushed in to these a c tiv itie s as a way to co m p le m e n t th e ir incom e F in a lly , the share o f fo rm a l wage incom e

appears not to be c o rre la te d w ith the d i s t r i b u t i o n o f fa rm la n d , su g g estin g that

t h is in c o m e s o u r c e m a y b e a s s o c ia te d w it h o th e r f a c t o r s , s u c h as e d u c a t io n ,

rath er than fa rm la n d h o ld in g

Figure 1: Income shares hy source and income quintiies

■Non-fami « Formal wa^e ■Iufomiahvage ■ Other iiK C r t iie

100%

1

•ệ te 60 %

3

0

X

•3 40%

©

w

1 20%

* Firm

III com qulntlltc (incomcper capita)

V IỆ T N A M H Ọ C - K Ỷ Y Ế U H Ộ I T H Ả O Q U Ỏ C T Ế L A N T H Ứ T ư

I

8

Ề

5

Ô

5

I

Jm

1

i

V0

i

w

I

Farm land holding qulnlllM (farmland size p a bcHiiebold)

Table 3 presents the G in i decom position o f incom e inequality by income source The overall G in i co e fficien t fo r the sample households was 0.267, w h ic h is much low er than the G in i co e fficien t o f 0.434 in the w hole country and 0.4 ] ] in the

Red River Delta reported by GSO (2008) This indicates a quite low degree o f

income inequality among the sample households Such a lo w e r inequality at d istrict level than that in larger areas is also found in V ietnam by M in o t, Baulch, and Epprecht (2006), w ho explained that, s im ila r 10 other measurements o f inequality,

there is a ứend tow ard sm aller G in i coefficients fo r sm aller regions, such as provinces or districts, than for the country as a whole T h is is due to the fact that

households in a sm all region are lik e ly to have m ore sim ila ritie s than households

across the w hole country

B y decomposing the total household income inequality into various income

sources, the results reveal that nonfarm self-em ploym en t, form al w ag e incom c and other income becom e the m ajo r contributors to the overall incom e in eq u ality am ong sample households Taken together, they account for 93 percent o f the total income

inequality B y contrast, farm and in fo rm al wage income reduced the inequality; the pseudo-Gini coefficients o f these income sources are much low er than the total G in i

coefficient, whereas the pseudo-Gini coefficients for nonfarm self-em p lo ym en t,

INEQUALITY INCOME IN H AN O I'S P E R I-U R B A N AR EAS

form al w a g e in c o m c and other income so u rc c s arc much higher, s p e c ific a lly , a 10

p c r c c n t increase in incom c from fa rm and inform al wage activities w ould have led

10 a 1.7 percent and 1.9 pcrccnt decline in the o verall incom e in e q u a lity ,

respectively Whereas, the same increase in nonfarm self-em ploym ent, form al wage

incom c and other incom e sources w o u ld have resulted in a 1.4 percent, 1.6 percent and 0.57 percent increase in the overall income inequality, respectively.

T a b l e 3 : G ini decomposition o f incomc inequality by income source

income source

Income share Sk

Gini Gk

Correlation with the distribution

of total income Rk

Pseudo-Gini GkRk

Share lo tola] income inequality

(RkGkSkVG

Source elasticity

of total inequality (RkGkSkV G-Sk

Non farm

Self-employment

N o te E stim ates are based on annual per capita incom es N '-4 7 7

L o o kin g at the third and fourth column in Table 3, the results show that the

inequali(y o f farm and in fo r m a l wage incomes among households is low er than the

in e q u a lity o f nonfarm self-em ploym ent, formai w age incom e and other incomes

among households In addition, as c o m p a r e d to nonfarm self-em p lo ym en t income,

form al w age incomes and other income, farm and in fo r m a l w age incomes have a

much low er correlation w ith the d istribution o f total income Consequently, income

from farm and i n fo r m a l w age w ork had an eq ualizing e ffe c t on the incom e

distribution Basically, this fin d in g is in a c c o r d a n c e w ith G allup (2002) and Van

C a o a n d A k it a ( 2 0 0 8 ) , w h o fo u n d th a t w h ile agricultural in c o m e a c tu a lly re d u c e d