Determinants of urban poverty: the case of medium sized city in Pakistan.
Iqbal, Nasir ; Awan, Masood Sarwar
Urban poverty, which is distinct from rural poverty due to
demographic, economic and political aspects remain hitherto unexplored,
at the city level in Pakistan. We have examined the determinants of
urban poverty in Sargodha, a medium-size city of Pakistan. The analysis
is based on the survey of 330 households. Results suggest that
employment in public sector, investment in human capital and access to
public amenities reduce poverty while employment in informal sector,
greater household size and female dominated households increase poverty.
We recommend greater investment in human capital and public amenities as
a strategy for poverty alleviation.
JEL Classification: 1310,1320, R200
Keywords: Determinants, Urban Poverty, Pakistan
1. INTRODUCTION
The process of urbanisation has dual impact on the development
process of an economy. Initially, it encourages the workers to switch
from low productive sector i.e. agriculture to high productive sectors
like services and manufacturing [Becker, et al. (1994)]. Subsequently,
it generates formidable problems for residents by depriving them of
access to essential basic needs [Egziabher (2000)]. It is also observed
that the poor try to urbanise faster as compared to the whole population
[Ravallion (2007)] and this urbanisation process leads toward the
emergence of urban poverty. Urban poverty is distinct from the rural
poverty with respect to its incidence, economic, demographic and
political aspects. The urban poverty can be controlled by developing the
clear understanding of its nature, magnitude and intensity.
It is estimated that the urban population of Pakistan is 35 percent
of the total population and its annual average growth rate is 3.4
percent (1990-2005) which is much higher as compared to South
Asia's figure of 2.8 percent in the same period [World Bank
(2007)]. Such expansion of urbanisation formulates a daunting task of
peering at the issues of urban poverty. In Pakistan, the phenomenon of
poverty is moving like a business cycle. It was high in 1960s and came
down in 1980s, but again moved upward in 1990s before falling rapidly
after 2000. Urban poverty fell from 22.7 percent in 2000-01 to 13.1
percent in 2005-06 [Pakistan (2008)]. This rapid fall of urban poverty
is linked with strong economic growth, rise in per capita income, large
inflow of remittances, and better economic and social policies of last
government [Chaudhry, et al. (2006)]. Recently, high inflation eroded
the gain made in poverty reduction by pushing people clustered close to
the poverty line to the below the poverty line [Anwar (2008)].
Given the changing level of poverty and emergence of new forms of
urban poverty, it is necessary to examine urban poverty especially at
city level. City Level Poverty Assessment (CLPA) is tool for acquiring
up-to-date information on a city's poverty and its social
development. Poverty profile at city level will provide a snapshot about
who is poor, where they live in the city, their access to services,
their living standards and so forth, thereby contributing to the
effective targeting of poverty by policy measures. Keeping this in view,
the objective of this paper is to estimate the poverty level and its
determinants at city level.
The sample city, chosen in this study to analyse the urban poverty
and its determinants is 'Sargodha' which is 10th largest city
of Pakistan. The estimated population of the city was 0.57 million in
2007 where 0.464 million people resided in municipal jurisdiction and
almost 0.106 million dwelt in cantonment area [Punjab (2007)]. Sargodha
city is the central hub of the district's agriculture and
industrial activities. The major crops of this area are wheat, rice,
sugarcane and citrus. Moreover, the district has gained immense fame in
citrus production especially the oranges (kinnows) of Sargodha which
have earned worldwide praise in taste. Hosiery, Textiles, Chemicals and
Soap are major manufactures of this area. Sargodha has grown at a very
rapid rate and become a major urbanised area in Punjab. It is the
industrial, commercial, financial and service centre of the country. In
recent years, the urban infrastructure has become overburdened and the
city has been subjected to considerable urban strife. Keeping the
importance of the city in view, it becomes essential to conduct a detail
study on poverty.
The rest of paper is organised as follow: review of literature is
presented in Section 2 and socio-economic characteristics of the city
followed by this section. Section 4 and 5 consists on methodology and
poverty profile of the city respectively. Section 6 explains the
determinants of the poverty and last section concludes the paper and
tries to present some piece of policy recommendations.
2. REVIEW OF LITERATURE
Poverty is a multidimensional phenomenon. Generally, concept of
absolute poverty is used to measure the poverty. Absolute poverty is
based on defining minimum calorie intake for food need and minimum
non-food allowance for human need required for physical functioning and
daily activities and this approach requires assessment of a minimum
amount necessary to meet each of these needs [Anwar (2006)]. For this
purpose, the most prominent approach used in Pakistan is calorie-based
approach [Naseem (1977); Irfan and Amjad (1984); Cheema and Malik
(1984); Malik (1988)]. In this approach, the poverty line is set as the
average food expenditure of those households who consume in the region
of the minimum required calorific intake. Ercelawn (1990) used calorie
consumption function to derive expected total expenditure of those
households who consume minimum required calorific intake. This method
derives expected expenditure for potential (2550) calorific intake
[Sherazi (1993)]. Subsequently, this method was modified by adjusting
for non-food expenditures [Jafari and Khattak (1995); Ali (1995); Amjad
and Kemal (1997)]. These studies used 2550 calories per day per adult as
the calorific cut-off point for estimation of absolute poverty. This
calorie norm was recommended by Pakistan Planning Commission (1985) and
supplemented by recommendations of FAO/WHO. The nutrition cell of
Planning Commission, Government of Pakistan reduced the calorie cut-off
point for Pakistan to 2150 calories per person per day per adult in 2002
but revised this threshold level to 2350 calories per adult equivalent
per day in July 2002 [Anwar (2006)]. Recently, there are number of
studies conducted in Pakistan by different institutions and authors to
examine the true picture of poverty in Pakistan. These studies used 2350
calories per adult equivalent per day as threshold point by including
food and non-food items for measuring absolute poverty [World Bank
(2006); Anwar and Qureshi (2003); Anwar, et al. (2004); Anwar (2006);
Jamal (2005); Jamal (2007) and Planning Commission and CRPR1D (2006)].
Natural population growth, rural to urban migration and the
reclassification of rural to urban areas works as deeper determinants of
urban poverty. It is estimated that rural to urban migration and
reclassification of areas are responsible of 40 to 50 percent of urban
population growth [UN (2005)]. Role of informal sector could not be
ignored in explaining the phenomenon of urban poverty. Informal sector
absorbs a large part of gigantic population of developing countries.
Hence informal sector, a dominant part of urban areas, assimilates a lot
of workers which are constantly becoming the part of urban population
due to rising urban population, rural-urban migration and
reclassification of areas. Over the year, absorption of labour force in
informal sector of the economy increases from 60.2 percent (1999-00) of
the total labour force to 66.1 percent in 2006-07 in urban areas of
Pakistan [Pakistan (2008)]. The poor section of the urban population can
be divided into the 'working poor' category and
'unemployed poor' category whereas the informal sector is
dominated by the working poor category but at the same time the
destitution of unemployed cannot be ignored [Manda and Odhiambo (2003)].
Poverty dynamics are closely linked with demographic
characteristics of the household especially family size, dependency
ration, sex of the head of the household, age composition and literacy
of the head of the household. Household size is prime demographic factor
and it is generally positively related with the poverty status [Qureshi
and Arif (2001); Chaudhry (2009)]. Large family size is likely to put
extra burden on a household's assets and resource [McKay and Lawson
(2002)]. Education of household head is the significant determinant of
household poverty [Qureshi and Arif (2001)] and the literate head of
household reduces the probability of being poor [Chaudhry (2009)]. Jamal
(2005) showed that in urban areas dependency ratio is also positively
related with the poverty status of the household.
Human capital acts as fundamental determinant in enhancing the
income level and hence in poverty reduction. Pakistan has owned the
poverty reduction strategy paper in which one of the main pillars of
poverty reduction is human capital. Without human capital formulation,
the goal of development or poverty elimination is inevitable. Human
capital accumulation is largely based upon education and skills
attainment. Nasir and Nazli (2000) found that monthly earnings of an
individual worker increased by 7.3 percent with an additional year of
schooling. Earnings will be increased by 37 percent with the attainment
of ten years of schooling against no education. They also found that
quality of schooling has significant effect upon earnings where quality
is here defined as schooling at private schools. Hence education can
increase the earnings potential of the poor. Thus investment in human
capital of the poor in the form of additional schooling can make them
productive. Siddiqui (2001) concluded that improvement in human capital
formation can be important in increasing women's economic
involvement and a reduction in gender based poverty. Jamal (2005) showed
that in urban areas the education of the head of the household is
negatively related with poverty. Haq (2005) found that poor persons of
Pakistan have low level of human capital and education clearly reduces
the probability of being poor because the role of education is important
in the labour market as those with higher education are more likely to
get employment and have higher wages. Wages and productivity in non-farm
activities rise with education at an increasing rate as education rises
[Kurosaki and Khan (2006)].
Provision of public services in the vicinity of the household is
also critical in determining the status of the household. Haq (2005)
found that the human poverty indicators, like housing, health, drinking
water, sanitation facilities and garbage collection system, are in
deplorable conditions in poor areas of city. Poor persons have low
standard of housing, majority suffered from chronic diseases, mostly use
the open well as a source of water, open drain system is prevalent in
poor persons and almost no garbage collection system is present for the
community. Arif and Iqbal (2009) found that access to electricity and
provision of education facilities for girls and health facilities in the
public sector play an important role in explaining the differences in
poverty levels. Investing more in provision of education and health
services is thus key to an increase in overall income of the population
and hence to reduce the poverty.
These studies clearly depict the multidimensional nature of poverty
and only knowledge about the absolute number is not sufficient to design
the effective poverty reduction strategy. Rather than focusing on
national and regional level poverty estimate, there is needed to conduct
detail study at city level to acquire the true picture of poor people.
To fulfil this gap in literature, this study explores these factors at
medium sized city in Pakistan such as Sargodha.
3. SOCIO-ECONOMIC CHARACTERISTICS OF CITY
In this section the descriptive analysis of the socio-economic
characteristics of Sargodha city is presented. This profile is based on
survey conducted for this study.
Education
Education is an important component of human capital and it is very
much effective in poverty reduction. Analysis shows that 14.2 percent
individuals never attended educational institutions whereas 55 percent
availed the education facility in past and 30.8 percent are presently
enrolled in educational institutions. Regarding the absorption of
educational institutions we have seen that out of the total students who
were enrolled or presently studying 73.4 percent are students of
government institutions and 25.2 percent are students of private
institutions (Table 1). It shows that in city, public sector is still
providing the educational facility to many students.
Occupational Status with Sectoral Composition
The occupation status shows that 18.6 percent of total individuals
are employed in government category. In this category people employed in
government departments/institutions and semi-government institutions are
included and 23.4 percent are working in private corporate sector.
Analysis also indicate that the largest occupation is own business/firm
category (30.5 percent). The workers who are getting salaries on daily
wages are 13.7 percent and 5.2 percent are overseas Pakistanis. Those
whose income is based on pension are only 1.8 percent. The individuals
in house-job are 0.6 percent and 0.9 percent are searching jobs in
labour market, while only 0.5 percent is not eligible to be employed
(Figure 1).
Sectoral composition indicates that 5.1 percent people are working
in agriculture sector, which is very low because sample only covers the
city region of Sargodha. 2.2 percent fall in the category of mining and
quarrying and 6.2 percent are working in manufacturing sector while 6.7
percent in construction related activities. Analysis also shows that 2.2
percent are involved in the distribution of services such as gas and
electricity, 3.0 percent are engaged in storage and communication sector
and 15.8 percent are in wholesale and retail trade. Persons in finance
and insurance, ownership of dwellings and public administration and
defense are 6.4 percent, 0.9 percent and 5.0 percent respectively.
Social service is the second highest sector having 21.0 percent of
working people (Table 2).
Dwelling Types and Status
Analysis shows that out of the total dwelling 93.6 percent houses
are independent houses and only 0.9 percent is apartment or flat. This
low figure is correct in the sense that in Sargodha city there is no
such flat-culture and most of the people have independent houses. Result
indicates that 5.5 percent dwell in a facility which is part of the
large unit and 87.9 percent houses are occupied by the owners. Only 3.9
percent are in the category of owner occupied (self-hired). The houses
on rent, subsidised rent and free of rent are 7.0 percent, 0.9 percent
and 0.3 percent respectively. Most of the houses have three rooms (24.5
percent of the total houses). Houses with two rooms are 16.4 percent and
residences with four and five rooms are 14.5 percent each (Table 3). It
is seen that houses with one room and with above six rooms are in low
proportion.
Provision of Public Amenities
As the area of analysis is urban region therefore regarding the
provision of basic infrastructure services such as electricity, gas,
telephone and sewerage, it is expected that urban dwellers are enjoying
better facilities. Result shows that almost 99.7 percent houses have
electricity connections and only 0.3 percent is deprived of this service
while 82.6 percent houses have gas connections and 17.4 percent are
without it. Regarding land-line facility, it is noted that 58.2 percent
houses have the land-line phone service against 41.8 percent who are
without it. It is also a noticeable fact that recent boom in cellular
mobile companies effected the monopoly of government land-line phone
service. Water supply facility is availed by 85.8 percent of the total
community and 95.1 percent houses have the sewerage system and only 4.9
percent are deprived of it (Table 4). It is observed that 89.1 percent
houses connected with underground drains, 4.2 percent with just covered
drains, 6.1 percent with open drains and 0.6 percent have no such
system. 91.8 percent houses have flush connected to public sewerage, 5.2
percent houses have flush connected to pit and only 3.0 percent houses
with flush connected to open drain.
For drinking water, 56.7 percent houses rely upon motorised
pumping/tube-well, 27.6 percent use piped water in their houses, 11.8
percent use hand pump and only 3.9 percent use other sources of water
for drinking purposes. It is also observed that 85.3 percent houses have
water in tap for 24 hours. From this figure we cannot conclude that
water facility of municipal administration is efficient because large
number of houses depend upon motorised pumping to use under ground
water. We have observe low figure for less than 24 hours and only 5.2
percent houses have less than one hour water available in their taps.
Water charges are very negligible in country and also paid by very less
proportion that use this facility. Only 33.6 percent of the total houses
pay for water supply and 66.4 percent do not.
4. DATA AND METHODOLOGY
Data Source and Data Collection Procedure
Poverty analysis is generally based upon primary data at household
level. For this study, primary data are collected under the joint survey
'Assessment of Poverty in Sargodha City' by the Pakistan
Institute of Development Economics (PIDE) Islamabad and the University
of Sargodha (UOS) Sargodha in May 2008. Sargodha city is mainly divided
into 22 union councils. The information is taken through randomly
selecting 11 union councils and then interviewed 30 households at random
in each selected union council. For selecting union councils and
household, we used the information provided by Federal Bureau of
Statistics. This activity provides the detailed information of 330
households in the city on major components required for poverty
estimation, including roster of the household, income of the household,
expenditure of household on food items, fuel and utilities, housing,
frequent non-food expenses and other non-food expenses like clothes,
footwear, education, and health related expenses. (1) It also contains
information on socio-economic indicator of the household.
Definition of Poverty Line
Poverty estimates are measured by using three different poverty
lines. First; official poverty line, estimated by the Planning
Commission of Pakistan is used. By using the Pakistan Integrated
Household Survey (PIHS) 1998-99 data, the Planning Commission estimated
absolute poverty line as Rs 673.5 per month per adult equivalent. This
poverty line is adjusted by consumer price index (CPI) to get the
adjusted poverty line for 2008. The Commission has already adjusted the
poverty line for the 2000-01, 2004-05 and 2005-06 periods using the
Consumer Price Index (CPI). In 2004-05, the official poverty line was Rs
878.64 per month per adult equivalent and in 2005-06 the inflation
adjusted official poverty line was Rs 944.5 per month per adult
equivalent [Pakistan (2008)]. Adjusted official poverty line, for
2007-08, used in this study is Rs 1140 per month per adult equivalent.
Anwar (2006) estimated poverty line by using latest PSLM data for
2004-05 and applying 2350 calories per adult equivalent per day as a
cut-off point. Poverty line based new estimate was Rs 933 per month per
adult equivalent for 2004-05. Adjusted poverty line is Rs 1211 per month
per adult equivalent for 2007-08. This poverty line also validates the
findings of World Bank (2006) about head count ratio in Pakistan. To
make these two poverty lines compatible with urban areas, these lines
were adjusted by rural urban food price differentials. The focus of this
study is to investigate the poverty in urban area, so to strengthen the
result and make them more suitable for urban area, this study also used
urban specific poverty line to get clearer picture of the poverty.
Qureshi and Arif used the Food Energy Intake (FEI) method to compute
separate poverty lines for both rural and urban areas. The cost of food
component of this basket was equal to the food poverty line determined
by estimating the cost of food consistent with a calorie intake of 2550
per adult equivalent per day for rural areas and 2295 calories per adult
equivalent for urban areas. They used 'Pakistan Socio-economic
Survey' (PSES) 1998-99 data for estimation of urban poverty line.
The estimated urban poverty line was Rs 874.1 per month per adult
equivalent for 1998-99 [Qureshi and Arif (2001)]. The adjusted urban
poverty line is Rs 1476 per month per adult equivalent (2) for 2007-08.
Measures of Poverty
By using these poverty lines based on the total expenditure
necessary for an acceptable standard of living considering 2350 calories
of the food items provided by the government of Pakistan, we estimate
the three important indicator of poverty:
Head Count Ratio: This estimate of poverty is worked out by
counting the persons below an exogenously defined cut-off level of
consumption expenditure, known as the poverty line from the distribution
of persons obtained from the consumer expenditure modules of survey of
the P1DE/UOS. The ratio between the person below the poverty line and
the total number of individual in the sample is called Head Count Ratio
(HCR). Mathematically it is defined as:
HCR = - H/N
HCR = Head Count Ratio
H = Number of person below the given poverty line
N = Total number of persons in the sample
Poverty Gap: This indicates the aggregate poverty depth of the poor
relative to the poverty line. This is a good indication of the depth of
poverty in that it depends on the distance of the poor below poverty
line i.e., the average consumption gap between the actual expenditure of
the poor and the poverty line. Potential for eliminating poverty by
targeting transfer to the poor is another implication of this indicator
[Ravallion (1992)]. Poverty gap also represents the total amount of
income necessary to raise every one, who is below the poverty line up to
that line. Estimating Procedure for this indicator as follow:
p = 1/n [n.summation (i=1)] [Z - [Y.sub.i]/Z]
Where
P = Poverty Gap (Distance of the poor below the poverty line).
Z = Poverty line determining expenditure
[Y.sub.i] = Consumption Expenditure of the /th poor household
Severity of Poverty: It is Foster-Greer-Thorbecke [P.sub.2] measure
representing severity of poverty. For this the poverty gaps of the poor
are weighted by those poverty gaps in assessing aggregate poverty. This
also shows variance in the poverty gap. It is estimated as:
[P.sub.2] = 1/2 [n.summation over (i=1)] [[Z - [Y.sub.i]/Z].sup.2]
Where
[P.sub.2] = Severity of poverty
Z = Poverty line determining expenditure
[Y.sub.i] = Consumption Expenditure of the /th poor household.
5. POVERTY PROFILE OF THE MEDIUM SIZED CITY
Extent, Gap and Severity of Poverty
To measure the extent of poverty i.e. poverty ratio or head count
ratio, three different poverty lines are used. The result shows that the
head count ratio in Sargodha city is 14.3 percent by using official
poverty line, 15.9 percent by using poverty line given by Anwar (2006)
and 21.0 percent by applying urban specific poverty line calculated by
Qureshi and Arif (2001). Poverty gap and severity of poverty are
aggregate measures of 'spread' of the poor below the poverty
line i.e. they aggregate the distance of all poor individuals from the
poverty line. Analysis shows that poverty gap is sufficiently large (4.4
percent) in 2008 as compared to the poverty gap (2.1 percent) measured
in 2005-06 for urban area of Pakistan [Pakistan (2008)]. As the
alleviation of poverty is the individual household phenomenon, the
income distribution pattern and individual household poverty gap would
lead towards the actual increase in income needed for the household to
be out of the poverty trap. A lower value indicates that most of the
poor are bunched around the poverty line. Higher value of poverty gap
indicates bad condition of the poor. The severity of the poverty is
shown by the squared of the poverty gap. So more the poverty gap, the
more would be the severity of the poverty. Severity of the poverty for
Sargodha city is 2.6 percent by using official poverty line, 2.8 percent
by applying Anwar (2006) definition of poverty and 3.3 percent by using
Qureshi and Arif (2001) estimated poverty line (Table 5).
Poverty Dynamics in Sargodha City
Poverty by Demographic Characteristics of Household
Various characteristics of the household have direct or indirect
bearings on the income generating activities or consumption pattern of
the households. These economic aspects of the individual household
determine the living standard of the household by which the poverty
status has been measured. The first demographic characteristic is the
age composition of the head of the household. Analysis indicates that
poverty level reduces with the increase of age of the head of the
household. Lowest incidence of poverty is found among the age group of
61 and above (Table 6). These households probably had some assets, more
experience and relatively more earners, so less poverty in the
household. The second demographic characteristic is family size.
Household size is positively related with the incidence of poverty.
Large household were more likely to be poor than small household because
larger households probably had more young children, that encounter
financial burden due to high cost of living, education, health and other
social as well as societal activities and vice versa. The incidence of
poverty for the largest households (9 + members) were more than three
times the incidence of poverty for the smallest group (1-4 members).
This gave the direct implication of family size and incidence of poverty
so family size is positively related with existence of poverty.
Migration status also plays vital role in moving household out of
poverty because migration provides better opportunities to get more and
more resources. Incidence of poverty was lower among those heads of
households who moved in the past to their current place of residence
(Table 6).
Poverty among Occupational Groups
In order to have an idea about the living status of persons engaged
in different occupations, the incidence of poverty has been calculated
for major occupation groups. Results show that incidence of poverty is
highest among the daily wage worker and lowest among the government
employees (Table 7). This indicates that secure job and proper flow of
income has direct implication for poverty status. People are more secure
in government sector, so they are less poor, while people working on
daily basis are not secure with their earnings. People with secure job
have more capacity to absorb economic shocks.
Poverty among Sectoral Groups
Sectoral composition indicates that incidence of poverty is more
likely in construction sector (Table 8). In urban areas, the informal
sector particularly construction sector, most of labours work on daily
wage basis. Informal sector create uncertainty and increase the chances
of unemployment in the economy. In this sector, there is no proper flow
of income for the household. This probably increase the chances that
individual is most likely to be poor if works in this sector i.e.,
construction sector. Another important finding is that poverty in those
household works in public sector is negligible, this indicate that
public sector is more reliable to reduced poverty.
Poverty by Access to Amenities
Distributional implications of the household's indoor
amenities affect not only the quality of life of the households but also
have direct bearings on the economic activities of the labour force of
the households. It is argued that households having access to amenities
are likely to be less poor compared to those without such provisions.
Table 9 shows that only very few household are without electricity (0.3
percent only). So electricity in term of poverty of the household did
not contribute much because almost all household has the facility of
electricity in their house. In city 82.6 percent of the sample household
have gas connection while the remaining 17.4 percent were managing fuels
by some alternative sources. The incidence of poverty was 14.2 percent
among the households having gas connection and 14.9 percent in the
households having no gas. So the poverty incidence was relatively higher
in the households having no access to this utility when compared with
households having gas connection in their vicinity.
In case of telephone facility in the households, only 58.2 percent
availing this facility and remaining 41.8 percent don not have this
facility. The incidence of poverty was more in the household having no
connections of telephone as compared with households having connection
(Table 9). Moreover, the fast growing mobile phone industry has solved
the communication problem and people prefer mobile connection rather
than fixed-line connection. In case of piped water supply, 85.2 percent
households availing this facility while only 14.8 percent deprived from
it. Poverty level was high in those households where this facility is
not available and less in those having this facility. Availability of
sewerage facility has the similar pattern with poverty.
6. DETERMINANTS OF POVERTY
Poverty is a multi-dimensional phenomenon, so varieties of factor
determine the nature and direction of poverty. These factors could be
economic, social or political. Identification of these factors helps us
to formulate policy to combat poverty. To measure the effect of these
factors, binomial logistic regression model is used in which the
dependent variable is dichotomous: 0 when a household is above and 1
when below the poverty line. Predictor variables are demographic, human
capital and dwelling endowment. The results will not be interpreted
through the coefficients but we will use the odd ratios in logistic
regression to see that the occurrence of any particular event will
increase or decrease the probability being poor of individual and with
what proportion as compared to the reference category.
Model Specification
Let's assume the general equation
[Y.sub.i] = f([X.sub.1i] [X.sub.2i] ... [X.sub.ki]) (1)
[Y.sub.i] is the dependent variable representing the
Households' level of poverty and Xs are the various household level
of education and experience. Let's suppose that the response
variable y captures a true status of the household either as poor or
non-poor so we can estimate the regression equation as follows
[y.sup.*.sub.i] = [[summation].sup.k.sub.j=0] [X.sub.ij]
[[beta].sub.j] + [[epsilon].sub.i] (2)
[y.sup.*] is not observable and is a latent variable. We can
observe [Y.sub.i] as a dummy variable that takes the value 1 if
[y.sup.*] > 0 and takes the value 0 otherwise. [beta]is the vector of
parameters and error terms are denoted with [epsilon]. The error terms
entail the common assumption of zero mean and underlying distribution of
the error terms is logistic. Let Pi denotes the probability that the rth
household is below the poverty line. We assume that the P, is a
Bernoulli variable and its distribution depends on the vector of
predictors X, so that
[P.sub.i](X) = [e.sup.[alpha]+[beta]X/1 + [e.sup.[alpha]+[beta]X]
(3)
[beta] is a row vector and [alpha] is a scalar. The logistic
function to be estimated is then written as
ln [[P.sub.i]/1 - [P.sub.i]] = [alpha] +
[summation][[beta].sub.i][X.sub.ij] (4)
In [P.sub.i]/1 - [P.sub.i]] is the natural log of the odds in
favour of the household falling below the poverty line whereas P7 is the
measure of change in the logarithm of the odds ratio of the chance of
the poor to non-poor household. Equation (4) is estimated by maximum
likelihood method and the procedure does not require assumption of
normality or homoskedasticity of error in predictor variables. [X.sub.i]
is the vector of independent variables. These variables include size of
household size, electricity connection in the house, phone connection in
the house, gas connection in the house, water supply in the house,
sewerage facility and education.
Generalised functional form of the model is as under:
P = a + [b.sub.1] HHS + [b.sub.2] Sew + [b.sub.3]WS + [b.sub.4]Tel
+ [b.sub.5]Gas + [b.sub.6] Pr of + [b.sub.7] Bach + banter + [b.sub.8]
[b.sub.9] Mat + [b.sub.10]Midl + [b.sub.11]Exp + e
P = Poor Household [1= if poor, 0= otherwise].
HHS = Household Size [in numbers].
Sew = Sewerage Facility [1= Yes].
WS = Water Supply Facility [1= Yes].
Tel = Land Line Telephone Facility [1= Yes].
Gas = Gas Connection [1= Yes].
Prof = Professional [1= Yes].
Bach = Bachelor [1= Yes].
Inter = Intermediate [1= Yes].
Mat = Matriculation [1= Yes].
Midle = Middle [1= Yes].
Exp = Experience [in years].
e = Error Term.
Dependant variable is defined by using official poverty line.
Eleven explanatory variables are used in this model. Human capital
variables are dummy variables and defined in term of educational level
and experience. One of them will get the value one in response to the
individual's highest educational attainment. It means the
educational level of the individual will either fall in middle,
matriculation, intermediate, bachelors or professional (masters and
above) category. Here 'primary education' is used as reference
category. In past research, it is found that human capital variables are
negatively related with the poverty level. Other variables include
experience, public services utilised by the individuals and their family
sizes. The experience variable is attained through subtracting the years
of schooling and school starting age from the age of a person. It is not
the actual but the potential experience. To make potential experience
more meaningful we have included the individuals with age above 14
years. The services include the Gas, Land-line Telephone, Sewerage and
Water Supply. All these services variables are dummy in nature if the
individual is availing the particular facility the respective variable
will get the value one otherwise zero. Household size variable is
continuous. The household size is taken because it directly linked with
the distribution of resources within the family members and is
positively related with poverty level.
Results and Discussions
It is observed that the attainment of middle, matriculation,
intermediate, bachelors and professional (masters or above
qualification) will decrease the likelihood of being poor by 38 percent,
70 percent, 79 percent, 92 percent and 96 percent respectively as
compared to their reference category of primary education (Table 10).
All the educational variables are negatively affecting the poverty
status of individuals. Moreover, as we increase the educational
qualification of individuals their chances of being non-poor increases
or we can say that the probability of being poor declines vigorously. If
an individual succeeds in getting matriculation education after middle
than actually the increment in the probability decline being poor will
be of 30 percent (70 percent-40 percent). Also such inter-educational
level comparison shows little improvement between bachelors and
professional categories but improvement is visible. With the increment
of one year in potential experience will reduce the likelihood of being
poor by 0.02 percent, although it is a minor effect but expertise is
effective in reducing poverty. Provisions of public services are
altogether negatively related with the poverty status. The decline in
the chances being poor with the availability of gas, telephone, water
supply and sewerage is 28 percent, 87 percent, 66 percent and 67 percent
respectively (Table 10).
Family size is important because as we increase the family size the
burden upon the pool of resources of any family will increase and
practically we have lesser and lesser resources for the welfare of
individuals. Large families are more prone to poverty. Therefore, we
observe positive sign for the household size as expected so with the
increase of one individual in family the rise in probability being poor
of individual is 49 percent (Table 10). Provisions of public amenities
are negatively related with status of the poor. All variable are
significant and have expected sign. These results indicate that access
to these facilities play an important role in explaining the difference
in poverty levels.
7. CONCLUSION AND POLICY OPTIONS
Where poverty is concentrated, who is affected and to what extent,
are relevant questions in poverty analysis. The analysis of poverty
presented in this study uses the data from survey conducted in Sargodha
city during May 2008. It is first time that this type of analysis has
been carried out in Sargodha. The survey was conducted for 330
households.
The analysis, based on official poverty line, shows that the head
count ratio is 14.3 percent while this ratio increases to 15.9 percent
by using latest poverty line given by Anwar (2006) and 21 percent by
using urban specific poverty line. Poverty gap for Sargodha city is
sufficiently high (4.4 percent) as compare to the aggregate poverty gap
(2.1 percent) measured in 2005-06 for urban area. By using other two
poverty lines, poverty gap become very large. Severity of the poverty
for Sargodha city is 2.6 percent by using official poverty line and 2.8
and 3.3 percent by using Anwar (2006) and urban specific poverty lines
respectively. Socio-economics analysis shows that education, family
size, nature of occupation and public amenities play important role in
poverty alleviation. Incidence of poverty is highest among the daily
wage worker and lowest among the government employees. The results also
show that education, experience and public services are negatively
related with the poverty status of individuals. Moreover, results show
that public services availability is also very essential for poverty
reduction. It is actually beyond doubt that proper service utilisation
symbolises the improved living standard of the people.
Following policy options can be used to reduce the urban poverty in
general and particularly for Sargodha city: There is need to focus on
the education of the poor because human capital plays vital role in
breaking the vicious circle of poverty. Poverty incidence is positively
related to family size. This highlights the importance of population
policies. Problem of poverty in Pakistan cannot be solved without
addressing the problem of rapid population growth. Government should
devote more resources for provision of reproductive health services.
Female education is another very powerful tool to contain population
growth and at the same time improve human capital of the country. Public
sector and private sector along with community participation should
manage and create human capital in the shape of better technical
education that will increase the productivity of the urban poor.
There is need to formulate programmes which help poor people to
manage risk. Micro-insurance programmes, public works programmes, and
food transfer programs may be mixed with other mechanisms to deliver
effective risk management. There is need to develop programme which can
prevent and respond to financial and natural shocks. There is need to
increase local organisations' capacity which will help in promotion
of community development which eventually enhance the control that poor
people and their communities have over the services to which they are
entitled. But strong monitoring mechanisms are suggested in this regard.
There is also need to support poor people's social capital by
assisting networks of poor people to engage with market and nonmarket
institutions to strengthen their influence over policy.
Informal sector plays critical role in poverty alleviation. On one
hand these is needed to enhance the productivity of participants of the
informal sector through provision of microcredit, skills training and
advisory services. Sargodha city is surrounded by very fertile citrus
orchids and a network of citrus processing factories. Citrus research
centres and vocational training for citrus processing workers can be
used to enhance the productivity level of these workers that ultimately
strengthen the economic well being of the masses of the city.
On the other hand, there is also needed to formalise the informal
sector especially the construction sector. Steps should be taken to
bring the informal sector into formal fold for better earnings. Less
stringent rules and regulations for the formal sector can encourage
informal sector enterprises to join the formal sector. Steps should also
be taken by government to minimise the wage differentials between public
and private sector by increasing minimum wage to reduce poverty.
REFERENCES
Ali, M. S. (1995) Poverty Assessment: Pakistan's Case. The
Pakistan Development Review 34:1, 43-54.
Amjad, R. and A. R. Kemal (1997) Macroeconomic Policies and Their
Impact on Poverty Alleviation in Pakistan. The Pakistan Development
Review 36:1, 39-68.
Anwar, T. (2006) Trends in Absolute Poverty and Governance in
Pakistan: 1998-99 and 2004-05. The Pakistan Development Review 45:4,
777-793.
Anwar, T. (2008) Growth, Poverty and Inequality. Published in
Poverty Assessment Update. Asian Development Bank.
Anwar, T. and S. K. Qureshi (2003) Trends in Absolute Poverty in
Pakistan: 1990-2001.
The Pakistan Development Review 42:4, 859-978.
Anwar, T., S. K. Qureshi, and H. Ali (2004) Landless and Rural
Poverty in Pakistan. The Pakistan Development Review 43:4, 855-874.
Arif, G. M. and N. Iqbal (2009) Infrastructure and Poverty Nexus:
the Case of Rural Pakistan in Socio-Economic Challenges Faces by
Pakistan. Proceedings of National Conference, Islamic International
University, Islamabad.
Becker, C., A. Hammer, and A. Morrison (1994) Beyond Urban Bias in
Africa. Portsmouth, N. H: Henemann.
Chaudhry, I. S. (2009) Poverty Alleviation in Southern Punjab
(Pakistan): An Empirical Evidence from the Project Area of Asian
Development Bank. International Research Journal of Finance and
Economics 23, 24-32.
Chaudhry, I. S., S. Malik, and S. Imran (2006) Urban Poverty and
Governance: The Case of Multan City. The Pakistan Development Review
45:4, 819-830.
Cheema, A. A. and M. H. Malik (1984) Changes in Consumption
Patterns and Employment under Alternative Income Distribution. The
Pakistan Development Review 24:1, 1-22.
Egziabher, T. G. (2000) An Overview of Urban Poverty in Addis
Ababa. Paper Presented at City Consultation Addis Ababa, Ministry of
Works and Urban Development Addis Ababa and UN-HABITAT, Nairobi.
Ercelawn, A. A. (1990) Absolute Poverty in Pakistan: Poverty Lines,
Incidence, Intensity. Applied Economics Research Centre, University of
Karachi, Karachi: (Draft Paper).
Haq, R. (2005) An Analysis of Poverty at the Local Level. The
Pakistan Development Review 44:4, 1093-1109.
Irfan, M. and R. Amjad (1984) Poverty in Rural Pakistan in Poverty
in Rural Asia. In A. R.
Kemal and Eddy Lee (eds.) ILO, Asian Employment Programme.
Jafri, S. M. Y. and A. Khattak (1995) Income Inequality and Poverty
in Pakistan. The Pakistan Economic and Social Review 33:2, 95-113.
Jamal, H. (2005) In Search of Poverty Predictors: The Case of Urban
and Rural Pakistan. The Pakistan Development Review 44:1, 37--55.
Jamal, H. (2007) Updating Poverty and Inequality Estimates: 2005
Panorama. Social Policy and Development Centre, Karachi. (SPDC Research
Report No. 69.)
Kurosaki, T. and H. Khan (2006) Human Capital, Productivity, and
Stratification in Rural Pakistan. Review of Development Economics 10:1,
116-134.
Malik, M. H. (1988) Some New Evidence on the Incidence of Poverty
in Pakistan. The Pakistan Development Review 27:4, 509-515.
Manda, K. D. and W. Odhiambo (2003) Urban Poverty and Labour Force
Participation in Kenya. Kenya Institute for Public Policy Research and
Analysis (KIPPRA).
McKay, A. and D. Lawson (2002) Chronic Poverty: A Review of Current
Quantitative Evidence. (CPRC Working Paper No 15.)
Naseem, S. M. (1977) Rural Poverty and Landlessness in Pakistan. In
ILO Report on Poverty and Landlessness in Asia, Geneva.
Nasir, Z. M. and H. Nazli (2000) Education and Earnings in
Pakistan. Pakistan Institute of Development Economics, Islamabad.
(Research Report No. 177.)
Pakistan, Government of (2008) Pakistan Economic Survey 2007-08.
Islamabad: Ministry of Finance.
Punjab, Government of (2007) Punjab Development Statistics.
Government of Punjab.
Qureshi, S. K. and G. M. Arif (2001) Profile of Poverty in Pakistan
1998-99. Pakistan Institute of Development Economics, Islamabad. (MIMAP
Technical Paper Series No. 5.)
Ravallion, M. (2007) Finance and Development. A Quarterly Magazine
of IMF44:3.
Schultz, T. (1961) Education and Economic Growth. Chicago: Chicago
University Press.
Shirazi, N. S. (1993) An Analysis of Pakistan's Poverty
Problem and its Alleviation Through INFAQ. Unpublished Ph.D.
dissertation in Economics, International Islamic University, Islamabad.
Siddiqui, R. (2001) Gender Dimensions of Poverty in Pakistan. Paper
presented at the 'Asia and Pacific Forum on Poverty: Reforming
Policies and Institutions for Poverty Reduction' Asian Development
Bank: Manila.
UN (2005) Urban Agglomerations 2005. New York: United Nations.
World Bank (2006) Summary of Key Findings and Recommendation for
Pakistan: Poverty, 2004-05. Islamabad
World Bank (2007) World Development Indicators 2007. Washington,
DC: The World Bank.
Comments
This paper is an attempt on analysing poverty in Sarghoda city.
Since the focus is only one city, therefore it can be more effective in
understanding the dynamics of poverty in Sarghoda city. Following are
few observations, which need to be addressed seriously:
(i) Analysis on almost six years old data has a little significance
in current scenario, when latest data is available.
(ii) Sampling criteria is questionable as half of the union
councils (11 out of 22) are dropped from the analysis and remaining half
are given equal representation (30 households from each union council)
irrespective of population proportion.
(iii) The significance of using three different poverty lines is
not given. Authors used three poverty lines (Official: 1140, Anwar: 1211
and Qureshi and Arif 1476). Why own poverty line is not estimated when
data was also available.
(iv) On page 12 Author write that Severity of Poverty index is the
Foster-Greer-Thorbecke index, but Po, Pi and P2 all are special cases of
FGT index
(v) Authors took household size as one of the variable to analyse
poverty. Mere use of household size ignores the number of earners in a
household. The best option could be to take dependency ratio.
(vi) In modelling the determinants of poverty relation with
experience is not linear. Authors wrote "with the increment of one
year in potential experience will reduce the likelihood of being poor is
0.02 percent" which is not true, the effect of initial years is
usually less as compare to mid years. Therefore the square of experience
should also be incorporated.
(vii) In defining the categories of education the lowest (and
reference) category is "primary education". It seems to be
that it is primary and below.
(viii) Provision of public amenities (Gas, Electricity, Water
Supply and Sewerage) is a bit confusing. As according to results (Table
10) the likelihood of being poor for households with gas connection is
28 percent less than households without gas connection. Should it not be
user or not, rather than available or not. Then why electricity is
dropped from the regression analysis.
(ix) One of the policy implications (page. 21) is "There is
need to formulate programs which help poor people to manage risk. From
which result this policy is drawn? Similarly it is concluded that there
is need to enhance the productivity of participants of informal sector
through provision of microcredit etc. Again from which result this
policy is recommended.
Muhammad Idrees
Quaid-i-Azam University, Islamabad.
(1) Although sample size was small due to resource limitation, yet
an utmost effort was made in sample selection process to make the sample
highly representative by using expert opinion and guidance from Bureau
of Statistics Sargodha.
(2) While adjusting household consumption expenditure in order to
get per adult equivalent expenditure, this study has used an equivalent
scale that gives a weight of 0.8 to individuals younger than 15 years
and 1 for all other individuals.
Nasir Iqbal <nasir@pide.org.pk> is Assistant Professor at the
Pakistan Institute of Development Economics, Islamabad. Masood Sarwar
Awan is Associate Professor at University of Sargodha, Sargodha.
Table 1
Type of Education Facility Availed
Type Percent
Government (Public) 73.4
Private 25.2
Deeni Madaris 0.6
NGO, Foundation 0.3
Elementary Educational School * 0.1
Others 0.4
Source: Computed from the survey of 'Assessment of Poverty in Sargodha
City'. * Major certificates and degrees of these schools include: PTC,
CT, B.Ed and M.Ed.
Table 2
Sectoral Composition of Labour Force
Sectors Percent
Agriculture 5.1
Mining and Quarrying 2.2
Manufacturing 6.2
Construction 6.7
Electricity and Gas Distribution 2.2
Transport, Storage and Communication 3.0
Wholesale and Retail Trade 15.8
Finance and insurance 6.4
Ownership of Dwellings 0.9
Public Administration and Defense 5.0
Social Services 21.0
Others 25.6
Source: Computed from the survey of 'Assessment of Poverty in Sargodha
City'.
Table 3
Presence of Number of Rooms in a House
No. of Rooms Percent No. of Rooms Percent
1 5.2 8 3.6
2 16.4 9 1.2
3 24.5 10 0.6
4 14.5 11 0.3
5 14.5 12 0.6
6 11.2 14 0.6
7 6.4 15 0.3
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Table 4
Houses with Availability of Infrastructure (Percent)
Services With Service Without Service
Electricity 99.7 0.3
Gas 82.6 17.4
Telephone 58.2 41.8
Water 85.8 14.2
Sewerage 95.1 4.9
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Table 5
Extent of Poverty, Poverty Gap and Severity of Poverty (Percent)
Poverty Line
Official Anwar Qureshi and Arif
Indicators (2006) (2001)
Head Count Ratio 14.3 15.9 21.0
Poverty Gap 4.4 5.2 6.0
Severity of Poverty 2.6 2.8 3.3
Source: Computed from the survey of 'Assessment of Poverty in Sargodha
City'.
Table 6
Decomposition of Poverty by Demographic Characteristics (Percent)
Poverty Line
Official Anwar Qureshi and
Household Characteristics % Share (2006) Arif (2001)
Age (Head of Household)
14-40 22.4 15.9 18.7 22.7
41-60 61.3 15.8 16.5 22.2
61 and above 16.3 6.9 10.1 13.8
Sex (Head of the Household)
Male 93.5 14.2 15.6 20.0
Female 6.5 16.4 20.0 21.1
Household Size
1-4 Members 9.8 7.1 7.1 8.5
5-6 Members 32.9 7.4 8.2 11.2
7-8 Members 28.4 14.5 17.1 20.7
9 and above Members 29.0 24.4 26.3 36.7
Migration
Non-migrant 79.2 16.8 18.7 24.4
Migrant 20.8 4.9 4.9 8.2
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Table 7
Poverty among Occupational Group (Percent)
Poverty Line
Official Anwar Qureshi and
Occupation % Share (2006) Arif (2001)
Government Employees 18.9 3.3 3.4 5.7
Private Employees 23.8 11.0 11.0 13.6
Own Business/ Firms etc. 31.0 7.5 9.0 11.5
Daily Wage Workers 13.9 30.0 33.3 45.6
Overseas Employees 5.3 8.8 8.2 8.8
Pensioners 1.9 8.3 8.3 25.0
Others 5.2 20.9 20.9 20.9
Source: Computed from the survey of 'Assessment of Poverty in
Sargodha City'.
Table 8
Poverty among Sectoral Group (Percent)
Poverty Line
Official Anwar Qureshi and
Sectors % Share (2006) Arif (2001)
Agriculture 5.1 9.1 9.1 9.1
Manufacturing 6.2 7.5 10.0 15.0
Construction 6.7 27.9 32.6 37.2
Electricity and Gas 2.2 7.1 7.1 7.1
Distribution
Transport, Storage and 3.0 15.8 15.8 21.0
Communication
Wholesale and Retail Trade 15.9 10.8 12.8 16.7
Finance and Insurance 6.4 0.0 0.0 2.4
Public Administration and 5.0 0.0 0.0 3.1
Defense
Social Services 21.0 5.9 5.9 10.4
Other 34.0 18.7 19.2 23.1
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Table 9
Decomposition of Poverty across Availability of Amenities (Percent)
Poverty Line
Official Anwar Qureshi and
Amenities % Share (2006) Arif (2001)
Electricity
Yes 99.7 14.3 15.9 21.0
No 0.3 0.0 0.0 0.0
Gas
Yes 82.6 14.2 15.6 19.2
No 17.4 14.9 17.0 29.8
Telephone
Yes 58.2 5.2 6.1 7.3
No 41.8 27.1 29.4 40.2
Water Supply
Yes 85.8 2.9 4.2 8.5
No 14.2 18.9 20.5 26.0
Sewerage
Yes 95.1 12.5 14.1 18.5
No 4.9 47.1 47..0 65.8
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Table 10
Logistic Regression Model of Being Poor with Multiple Independent
Variables
Variables Coefficient Level of Significance Odd Ratios
Experience -0.011 0.01 0.98
Education
Middle -0.592 0.03 0.62
Matriculation -1.231 0.00 0.30
Intermediate -1.819 0.00 0.21
Bachelor -2.608 0.00 0.08
Professional -3.291 0.00 0.04
Gas -0.351 0.07 0.72
Telephone -2.252 0.00 0.13
Water Supply -1.200 0.00 0.34
Sewerage -1.192 0.00 0.33
Household Size 0.346 0.00 1.51
Constant 0.850 0.10 1.92
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Fig. 1. Occupational Status of the Household
Daily wag worker 14%
Overseas Employee 5%
Others 8%
Government Employee 19%
Private Employee 24%
Own business/firm etc. 30%
Source: Computed from the survey of 'Assessment of Poverty
in Sargodha City'.
Note: Table made from pie chart.
