Measurement of living standards deprivation in Punjab using AF method (periodical comparison approach).
Afzal, Muhammad ; Rafique, Shamim ; Hameed, Farhan 等
The purpose of this study is to assess multidimensional poverty
using Alkire and Foster (AF) method in the periods 2007 and 2011 in
province Punjab-Pakistan, using primary data from Multiple Indicator
Cluster Survey (MICS). The results are bifurcated for geographical
split-ups of the Punjab to further explore over time status of poverty
and monitor the disparities among different areas and regions of the
Punjab. As a whole the Punjab is exhibiting a decline of 5.7 percent in
poverty, whereas, the rural and urban areas are showing decline of 6
percent and 0.2 percent, respectively in the incidence of poverty in
2011 as compared to 2007 at poverty cutoff of 33 percent. However, the
rural areas have 37.4 and 31.3 percent more poverty as compared to urban
areas in the years 2007 and 2011, respectively. The highest decline in
poverty at division level is found to be 11 percent for Bahawalpur
division and at district level, the highest decline in poverty is found
in Vehari district of 30 percent. All the regional split-ups of the
Punjab are found not having similar status, so similar policies all over
the province will not prove useful. Allocation of resources should be
made on the basis of different bands of poverty. For instance D.G. Khan,
Bhawalpur and Sargodha divisions need more attention as compared to
Gujranwala, Lahore and Rawalpindi divisions. Regarding living standards
deprivation, it is high time the Punjab government and poverty reduction
policy makers focused on Rajanpur, Muzaffargarh and R.Y. Khan Districts
as the first priority.
JEL Classification: 132, P46, H53, H75,138
Keywords: Living Standards Deprivation, Poverty, MPI, MICS, Punjab
I. INTRODUCTION
In spite of taking and implementing various special measures by the
government of Punjab and the Pakistan to alleviate poverty in Punjab,
poverty is still there and has become a constraint in the way of
economic progress and prosperity of the people of the Punjab-Pakistan.
Poverty is pronounced deprivation in well-being. The conventional view
links well-being primarily to command over commodities, so the poor are
those who do not have enough income or consumption to put them above
some adequate minimum threshold.
The broadest approach to well-being and hence poverty focuses on
the capability of the individual to properly function in the society.
The poor lack key capabilities, and may have inadequate income or
education, and last but not the least living standards. How we measure
poverty can importantly influence how we come to understand it, how we
analyse it, and how we create policies to influence it. In recent years,
the literature on multidimensional poverty measurement has blossomed in
a number of different directions. The 1997 Human Development Report
vividly introduced poverty as a multidimensional phenomenon, and the
Millennium Declaration and Millennium Development Goals (MDGs) have
highlighted multiple dimensions of poverty since 2000.
Salahuddin and Zaman (2012) in the article entitled
"Multidimensional Poverty Measurement in Pakistan: Time Series
Trends and Breakdown" applied Alkire-Foster Multidimensional (AFM)
poverty measure given in 2007 for building time-series trends of poverty
in Pakistan for the period 1998-2006. Their study results show that
multidimensional poverty measures provide more elaborate and precise
picture of poverty in Pakistan. The authors found that people of
Pakistan were highly deprived in education and health.
Naveed and Tanweer-ul-Islam (2012) in their paper entitled "A
New Methodological Framework for Measuring Poverty in Pakistan"
presented a critical analysis of poverty measurement in Pakistan and
argues for adopting a multidimensional methodological framework.
Utilising AF methodology over the RECOUP Household Survey data (2006-07)
the paper provides multidimensional poverty estimates at the aggregate,
provincial and district level and identifies the major drivers of
poverty. Their paper seems helpful in elaborating how policy makers can
prioritise the development budget among districts and allocation within
each district based upon the level and nature of deprivation. The
authors found that consumption level as a single measure of poverty
alone was a poor measure of poverty in Pakistan. In another paper
entitled "Estimating Multidimensional Poverty and Identifying the
Poor in Pakistan: An Alternative Approach" Naveed and
Tanweer-ul-Islam (2012) critically examined the Poverty Scorecard, which
was recently introduced by the Government of Pakistan for the
identification of poor households under the Benazir Income Support
Programme. By employing the AF measure to analyse household data from
two provinces, Khyber Pakhtunkhwa and Punjab, their paper recommends an
alternative method to estimate multidimensional poverty and identify
poor households. This paper also investigates the relationship between
household consumption and multidimensional poverty. This paper contrasts
the results obtained by using a multidimensional measurement of poverty
with those of the official poverty line. The limitations of the official
poverty line were also identified and the role of household consumption
in explaining deprivations was discussed in this paper.
Contemporary methods of measuring poverty and wellbeing commonly
generate a statistic for the percentage of the population who are poor-a
Head Count Index (H). A practical aim of Alkire and Foster (2007, 2011)
was to construct poverty measurement methods that could be used with
discrete and qualitative data. It includes identifying 'who is
poor' by considering the range of deprivations they suffer, and
aggregating that information to reflect societal poverty in a way that
is robust and decomposable.
Pakistan, being the 6th highest populous and 9th largest (with
respect to size of its labour force) country of the world, have a
population of about 177 million in 2011. Punjab is the biggest province
of Pakistan with a population of 96.55 million (55 percent of total
Pakistan's population) in 2011. The labour force participation rate
remains low (32.98 percent) in Pakistan as compared to other countries
of the world, reflecting the large chunk of children and old ages (67.2
percent) in the population. The civilian labor force in Pakistan is
58.41 million in 2011.The crude birth rate, death rate and infant
mortality rate per 1000 persons has been found 27.5, 7.3 and 70.5
respectively, in 2011. The male (10 year and above) labour force
participation rate is only 68.83 percent as against only 21.5 percent
for female that remains very low in 2009-10. Some social, cultural and
religious factors that prevent female workforce to participate in paid
jobs are the main reasons for this low female participation rate.
Agriculture sector is considered as back bone and the major sector of
the Punjab and Pakistan's economy accounting for 44.75 percent and
45.27 percent, respectively of the total employment. The officially
Labour Force Survey reported unemployment rate in Pakistan stood at 5.6
percent in 2009-10. Pakistan's literacy rate for male, female and
both stood at 69.5 percent, 45.2 percent and 57.7 percent, respectively
as against Punjab's literacy rate for male, female and both stood
at 69.1 percent, 49.8 percent and 59.6 percent, and, respectively in
2009-10. The above literacy rate figures reveal that the overall
Pakistan's literacy rate is determined by overall Punjab's
literacy rate because of the size of literate population in Punjab.
Education expenditure as a percentage of Gross National Product (GNP)
remained around 2 percent throughout the history of Pakistan [Pakistan
(2010-11)].
Considering the scope and subject matter of the study, the key
objective of this study is to measure Multidimensional Poverty Index
(MPI) for the considered periodical segments 2007 and 2011 in the
province Punjab and, in turn, going deep into different areas, divisions
and districts to have neck to neck evaluations of the poverty status in
the Punjab province of Pakistan.
Since the MPI is founded upon seven different indictors of living
standards so the overall results can also be decomposed to have the
absolute and relative contribution of each indicator towards the overall
MPI. Using this property of the MPI, we can go deep into each division
and district with the intention to observe the poverty status with
regard to MPI value of each indicator. The two period comparisons i.e.,
the years 2007 and 2011 will prove helpful to track the changes in
poverty over time in different areas, divisions and districts of the
Punjab. It will also prove helpful in auditory analysis of the allocated
funds to specific regions worthwhile along with political regime of
military and democracy.
Since the results of this study are bifurcated for geographical
split-ups of the province, this study aids the policy makers in Punjab
to eradicate poverty in the respective areas, regions, divisions and
districts. This study has its own significance to every reader and
specifically for government institutions because it also provides a
picture of the poverty status and helps to monitor the disparities among
different regions of the Punjab. The study is of a unique nature in the
respect that it is perhaps the first study assessing Living Standards
Deprivation in Punjab using MICS data and AF Method. This study would
also be helpful for policy makers for enhancing the living standards of
deprived segments of the society, especially the households. The finding
of this study could offer a base for formulation of sound policies for
deprived regions of the Punjab, exclusively to public and private
organisations for the betterment of rural households through increased
their living standards. This study may catch the interest of democracy
lovers regarding living standards deprivation when compared to guided
democracy of General Musharif as the MICS data for the period 2007
reflect the impact of policies of the government guided by General
Musharif and the MICS data for the period 2011 depicts the impact of
policies of the government guided by President Asif Ali Zardari.
This study is delimited to two period comparisons i.e., for the
periods 2007 and 2011 because of the non-availability of MICS data for
current periods i.e., after 2011. This study is also geographically
delimited to Punjab province of Pakistan as the MICS data for other
provinces is not available. Further, this study is delimited to only one
aspect of deprivation that is of the living standard deprivation as
sound and reliable data on the other aspects of deprivations are not
available in MICS data.
II. REVIEW OF LITERATURE
Keeping in vision the different dimension of the study, the review
of literature has been fulfilled. The Human Development Report, 1997
presented the most realistic approach by not only high lighting the
poverty of income, but also on poverty from human development outlook-
poverty as a contradiction of choices and opportunities to live
comfortable lifespan.
Salzman (2003) terms in her paper "Centre for the Study of
Living Standards" the methodological adoptions in the construction
of composite, economic and social welfare indices. The author derived
out with the result that "in current years a bulk of composite and
social welfare indices have been developed, but the development is made
inefficiently and methodologies are ignored". This paper suggested
a list of recommendations for best-practice methodologies founded upon
the recent paper by Booysen (2002) and the United National Development
programme [e.g., Anand and Sen (1994)].
Jamal (2003) uses the Index of Multiple Deprivation (IMD) based
upon the 1998 Population and Housing Census Pakistan data. This paper
focuses the poverty alleviation concerns in Pakistan. It presents the
practicable ways to deal income for poverty improvement in developing
countries. Furthermore, the study discussed about identification of
areas of concern, building up conclusions on local and sectorial main
concerns, smooth the programs for poverty lessening in the targeted
community and understanding the association between poverty and its
foundation.
Ashraf and Usman (2012) presented a new measure of Multidimensional
Poverty Index (MPI) for the province of Punjab using a method proposed
by Alkire and Foster (2007, 2009). The authors estimated MPI by applying
SPSS and MS-Excel on MICS data for the period 2007-08. This paper
integrates many aspects of poverty related to the MDGs into a single
measure. MPI also examines the most common deprivations related to
different districts of Punjab. According to this study, the less
multidimensional deprived districts were: Lahore, Multan, Rawalpindi,
Sialkot, Jhelum, Gujranwala, Sahiwal and Faisalabad are included. The
districts with moderate multidimensional deprivations according to MPI
were: Attock, Mandi-Bahauddin, Mianwali, Gujrat, Chakwal, T.T. Singh,
Vehari, Khushab, Nankana Sahib, Narowal, Bhakkar, Sargodha and
Sheikhupura. The districts Hafizabad, Kasur, Okara, Lodhran, Pakpattan,
Khanewal, Bahwalnagar, Jhang, Bahawalpur, Layyah, Rajanpur, R. Y. Khan,
D. G. Khan and Muzaffargarh were the most deprived in all dimensions.
A compact among nations to end human poverty-HDR (2003), and the
innovative century opened with an exceptional accentuation of
commonality and fortitude to eradicate the poverty from the world. In
2000, UN Millennium Declaration was made in the "largest ever"
meeting of the head of the States of committed countries - "Rich
and Poor" for doing all that can be done in order to eliminate the
poverty. The main apprehensions of this declaration are to promote human
decorum, maintain social equality, impartiality and achieving peace and
ecological sustainability by 2015 or earlier.
Originated from the Millennium Declarations, the MDGs are
associated to perceive poverty in the multidimensional way. Insufficient
income prevalence of hunger, gender inequality, deficient in education
and living standards are addressed for the reflection of the poverty
picture in the respective countries. This task was also accepted by
Pakistan being the signatory and various steps were taken in this
concern. MICS linked MDGs to have most of the data on the proposed
indicators to track changes over time. Various rounds of provincially
MICS are being conducted in Pakistan. In Punjab, MICS 2007 and 2011 is
the second and third round of MICS in the series.
The Human Development Index (HDI) is one of the most extensively
used measures of human development, developed and published by
UNDP's first annual Human Development Report (HDR), 1990. The HDI
is structured in the order of Amartya Sen's competency approach
which emphasises the consequences of standards of living, health and
education [Stanton (2007)]. Before HDI, many indices like GDP per
capita, GNP per capita , life expectancy, literacy and enrolment are
being used but none of these has not got much as gratitude as Mahbub ul
Haq's HDI [HDR (1990)]. In spite of all its significance, HDI is
being criticised for choice of variables, predetermined weighting
methodology and redundancy. Another imperative apprehension regarding
HDI is its equally weighting method. Ghaus, Pasha and Ghaus (1996) and
Noorbakhsh (1998) have provided the other ways of giving weights to the
dimension and variables.
Jamal (2009), constructed District Human Development Indices for
the Punjab for the periods 2004 and 2008 by using HDI that integrates
three dissimilar factors (a) a long and healthy life (life expectancy)
(b) education as a combination of adult literacy and school enrolment
and (c) a decent level of livings. The research utilises the district
based MICS 2004 and 2007-08 data.
While constructing Punjab Indices of Multiple Deprivations for the
periods 2003-04 and 2007-08, Jamal (2011) presented the income poverty
results using MICS data. However the authors ignore the multidimensional
aspect of poverty. These indices of multiple deprivations are intended
to evaluate the poorest or socially excluded segment of the society.
Niazi and Khan (2011) in the paper" The Impact of Education on
Multidimensional Poverty across the regions in Punjab" assessed the
educational deprivation and estimated the incidence of multidimensional
poverty in Punjab using AF Method. The study estimated the contribution
of lack of education in the incidence of multidimensional poverty in
urban and rural areas of province Punjab, Pakistan. The overall
educational deprivation of the multidimensional poor segment during
1998-99 was found to be 60.8 percent, which significantly increased to
83.4 percent in 2001-02 but decreased as 72.4 percent in 2004-05 and
again increased to 79.8 percent during 2005-06 along with little decline
as 78.0 percent in 2007-08, whereas the incidence of multidimensional
poverty during the same period was 48.6, 49.99, 40.80, 45.72 and 42.38
percent, respectively over the time. This study also found lowest
educational deprivation as well as the incidence of multidimensional
poverty in urban area as compared to the rural areas of the Punjab
throughout the period under consideration.
On 14 July, 2010, UNDP and Oxford Poverty and Human Development
Initiative (OPHI) presented a new index of measuring poverty level in a
multidimensional way. Alkire and Santos (2010) presented a paper on this
new Multidimensional Poverty Index (MPI) for 104 countries.
The Punjab provincial Reports of MICS, 2007 (vol-I) and MICS, 2011
(vol-I), are the outcome of continual efforts of Bureau of Statistics,
Planning and Development Department, government of the Punjab to provide
reliable data for monitoring the effectiveness of interventions to
eradicate poverty in the province. The indicators of MDGs for education,
health, water and sanitation and poverty are accessible in both reports
to track the changes in poverty over time and areas of distressing
concerns being highlighted.
Pakistan Economic Survey, 2010-11 reviews the development of
Pakistan's economy over the years; the reported source uses the
absolute poverty line method based upon the calorie method. The poverty
line was used for cutoff at 1.25 $ a day.
The above literature review indicates that poverty and its
dimensions remained the interest of social scientists since 1990. A
number of studies were also carried out in the recent past to assess the
scope of poverty in Pakistan both at micro and sectorial levels, but
very few studies have put emphasis on the fundamentals of poverty.
Poverty is a sign of many disorders in the configuration of Nations, so,
it is an effect of many causes. MPI is the very adequate alternative for
the measure of acute, absolute and relative poverty. Instead of using
direct income or consumption approaches, which have their own data
constrains and are very probable to be influence with the annexation of
random disturbance terms, due to fact that data on these variables is
attached to the human verbal and behavioural outcomes and by nature
these numerical facts and figures are tensional or intentional over
reported or under reported at the sweet will of the plaintiffs.
The idea of using multiple variables for the identification of
deprivation and in turns going for the poverty index measures through
the filters of dual cutoff is justified in manifold reasons. Just having
the sole identification process as most of the unidimensional measures
does, may include the certain number of individuals who are deprived in
particular indicator, but they may be at higher level of satisfaction in
having the sagacity that they have achieved such glassy.
Measuring social problems in a truthful way is an essential element
of modern and democratic governments and measuring it in a
multidimensional way helps government to do better in terms of policy
making as poverty is the multidimensional phenomenon and it must be
tracked over time for changes in the multidimensional way. This study
opens the new horizon and many innovations are in line to be considered
by having the series of the MPI measures with regular time lags. In this
connection the two different rounds of MICS are considered to have MPI
measures and changes over time are tracked. This will reflect and
provide the guide lines to design social polices strategically with
desired objectives for public sectors. The results can serve as
practical instruments for monitoring policies and are useful alerts for
decision making at a short and long term time spans.
III. DATA SOURCES, SAMPLING PROCEDURE AND METHODOLOGY
Data Sources
MICS (Multiple Indicator Cluster Survey) Punjab, 2007 and 2011
provide representative household survey estimates regarding more than
100 indicators vis-a-vis province, area of residence (major cities,
other urban and rural), 9 divisions, 36 districts and 150 tehsils/towns.
It was one of the largest surveys in the history of Pakistan with a
sample size of 102,545 households for MICS 2011 and 91280 for MICS 2007
with an exceptional response rate of 97 percent. The survey was planned,
designed and implemented by Punjab Bureau of Statistics under the
supervision of second author. The sample design of both MICS was
provided by Pakistan Bureau of Statistics. Technical input was obtained
from Regional Office for South Asia-UNICEF (ROSA) and Global Desk on
MICS4. Fieldwork was carried out from July to December in both surveys
for their respective rounds. Report and data of MICS Punjab, 2011 is
also available at one of the UN web domain Child info.
Sample Design
The sample has been selected in two stages. In urban areas, the
first-stage selection unit is the Enumeration Block. In the rural areas,
the first-stage selection unit is the Village. From each first-stage
sample unit, a sample of households has been selected: 16 in the rural
areas and 12 in the urban areas. The second stage units are selected
with equal probability. This gives a sample that is more or less
self-weighing within each selection stratum.
Multidimensional Poverty Index (MPI)
The MPI measure is very smooth and robust and the advantage of
using MPI is that it is sensitive to the changes as compared to simple
Head Count Ratio (H), the H remnants unbothered if a person who is
censored as poor after the poverty cutoff becomes more deprived or less
deprived, the H only changes when the person become non-poor or become
poor. On the contrary, the MPI being the product of H and Average
Intensity of Poverty (A) grosses the changes according to the
deprivation rank of the censored poor.
The MPI can be used to imitate the clear depiction of the
individuals, households or communities and even countries living in
poverty. With the decomposition property of MPI it is also potential to
perceive shallow into each of the dimension and bifurcating some certain
geographical split-ups. Additionally, we can have the pattern of the
poverty by taking array of poverty cutoffs to expedite the policy maker
with poverty index rendering to different bands of poverty namely low,
medium and high.
The AF Method generates Head Counts and also a unique class of
poverty measures ([M.sub.[alpha]]). [M.sub.0] (for [alpha] = 0) is an
adjusted Head Counts. This Mo reflects both the incidence (the
percentage of the population who are poor) and intensity of poverty (the
number of deprivations suffered by each household, A). [M.sub.0] is
calculated by multiplying the proportion of people who are poor by the
percentage of dimensions in which they are deprived ([M.sub.0] = H x A).
For the measurement of the MPI, seven indicators from the Household
Characteristics Module of MICS 2007 and 2011 are considered with the
total weight evenly distributed among them. The reason for the inclusion
of these indicators is that most of the data obtained in this module are
the results of the observational and visual retorts of the enumerators.
So, the chances of false information are very low.
To obtain the Achievement Matrix (X): which shows the achievement
of each household in each of the seven indicators, for M1CS 2011 of
order (95238 X 7) and of order (91280 X 7) for MICS 2007, the responses
for each indicator in the MICS: 2011 and 2007 Standards of Living
Modules responses are re-coded according to the definition provided by
UNICEF, Joint Monitoring Program (JMP) of improved and unimproved
sources for each indicator. The definition for improved and unimproved
sources for each indicator with their relative weights and deprivation
cutoff are presented in Table 1. Equal weights to different living
standard indicators are assigned in Table 1. Applying scientific methods
to assign weights may mislead the preferences of the household to each
living standard indicator as each indicator yield different importance
to different households.
Achievement Matrix (X)
The X is the one which represents the outcome of the indicators for
each household; it is of the order n x d, in this particular case of
MICS 2011, the X will be of the form.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].
For MICS 2007, the X will be of the form.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].
Deprivation Cutoff Vector and Matrix
A vector Zj = [Improved, Improved, Improved, Improved, Improved,
Improved, 50 percent of Assets] for 7 deprivation cutoffs (one for each
dimension) is used to determine whether a person is deprived. If the
person's achievement level in a given dimension "j" falls
short of the respective deprivation cutoff Zj, the person is said to be
deprived in that dimension; if the person's level is at least as
great as the deprivation cutoff, the person is not deprived in that
dimension.
According to the cited criteria the entries in the achievement
matrices are substituted into dichotomy i.e., [go.sub.ij] = 1, if
[X.sub.ij] < [Z.sub.j] (Deprived) and, [go.sub.i j] = 0 if [X.sub.ij]
[greater than or equal to] [Z.sub.j] (Non-Deprived). In this way the
Deprivation Matrices [g.sup.o]'s are obtained for both of MICS 2011
and 2007.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].
Weighted Deprivation Matrix (WDM)
The relative weights W = [1/7, 1/7, 1/7, 1/7, 1/7, 1/7, 1/7]
indicators are applied to the deprivation matrices. Such that
[go.sub.ij] = [W.sub.j] = 1/7, if [X.sub.ij] < [Z.sub.j] (Deprived)
and [g.sub.ij] = 0, if [X.sub.ij] [greater than or equal to] [Z.sub.j]
(Non-Deprived) so that this study obtaineds the WDM as given below:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].
Deprivation Count Vector (DCV)
These vectors are the count or score of each person in all the
indicators. It is the sum of weighted deprivations, i.e., [C.sub.i] =
[g.sub.i1] + [g.sub.i2] - + [g.sub.i7]. The DCVs for MICS 2011 and 2007
are given below:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Poverty Cutoff
Given the poverty cutoff K, This study compares the deprivation
count with the K cutoff and then censor the deprivation of those who
were not identified as poor.
If [[rho].sub.k](x, ;Z) = 1, if [C.sub.i] [greater than or equal
to] K
If [[rho].sub.k](x, ;Z) = 0, if [C.sub.i] < K
Censored Weighted Deprivation Matrix
It is the key matrix over which we will perform the aggregation and
find the set of AF measurements for [M.sub.o] (MPI). Here [go.sub.ij]
(k) = [W.sub.j] = 1/7 if [C.sub.i] [greater than or equal to] k
(Deprived and poor) [go.sub.ij] (k) = 0, if [C.sub.i] < k ( Deprived
or not, but non-poor).
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Censored Weighted Deprivation Count Vector
After the implementation of dual cutoffs, this vector counts the
score of each person from the Censored Weighted Deprivation Matrix. Here
[C.sub.i](k) = [C.sub.i], if [C.sub.i] [greater than or equal to] k and
[C.sub.i](k) = 0, if [C.sub.i] < k.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Head Count Ratio of MD Poor
It is the proportion of people who have been identify as poor. It
is called incidence of poverty, or poverty rate and is calculated as:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Intensity (Breadth) of MD Poverty
It is average proportion of deprivation in which the poor are
deprived and is calculated as:
A (2011) = [[summation].sup.95238.sub.i=1]
[c.sub.i](k)/7[q.sub.2011], A(2007) = [[summation].sup.91280.sub.i=1]
[c.sub.i](k)/7[q.sub.2007]
[M.sub.o](MPI)
This is the final step for the calculation of MPI. It is the
adjusted Head Counts and is the product of H and A, i.e., [M.sub.o] = H
x A
IV. RESULTS AND THEIR INTERPRETATION
Poverty Identification
With the poverty K-Cutoff, this study is considering the range of
cutoffs to observe the pattern of each of the AF measurement. Table 2
shows the results for the periods 2011 and 2007 and corresponding
graphical representation are shown in Figure 1 and Figure 2.
It is substantiation from Table 2 that the Head Count Ratio (H) is
very high for both time periods, when we have established the poverty
cutoff at 10 percent deprivations. As one move from 10 percent to 100
percent poverty cutoff, H keeps on deceasing, but still one got some
percentage of multidimensional (MD) poor people even at 100 percent
poverty cutoff.
The average intensity (A) has the increasing pattern, it is due to
the fact that in the Censored Weighted Deprivation Matrix as the
percentage of poverty cutoff increases the household with more
deprivations are censored as poor, and the Average Intensity of the
poverty is the average of the MD poor people. At the initial poverty
cutoffs, the A is low and with the increase in poverty cutoff the
percentage of A keeps on increasing and becomes 100 percent for both
time periods.
The [M.sub.o] is the product of H and A and it is the percentage of
people who are MD poor and facing deprivations at the same time, with
the increase in the poverty cutoff, the value of Mo decreases, but even
at 100 percent poverty cutoff, this study still got some percentage of
the MD poor.
Overall Comparison of Mo (2011) and Mo (2007)
There is difference of approximately 6-10 percent in the value of
Mo (2011) and Mo (2007) at each of the poverty cutoff level. The Figure
3 shows the prominent decrease in the poverty for the period 2011 as
compared to the period 2007.
In conclusion, this study observed that each of the AF measure has
shown decrease in poverty in 2011 as compared to 2007 at all cutoffs.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Poverty Identification (K-Cutoff at 33 percent)
To converse about MPIs at a specific poverty cutoff, this study set
the K-cutoff at 33 percent. Having AF measures at this cutoff this study
will drill down into Regions/Divisions/Districts for independent MPIs
and their contribution to the provincial MPI.
The poverty identification for poverty cutoff K=33 percent for both
the considered time periods are presented in the Table 3 and Figure 4.
The overall results show a decrease in each of the measure for the year
2011 as compared to the year 2007. It is worthwhile to note that the H
and the A have decreased by 3 percent and 7.9 percent, respectively
whereas; the MPI (Mo) has decreased to 5.7 percent. Here, the advantage
of using AF method is that the H has shown just 3 percent (does not take
into account the phenomena that poor become more deprived or less
deprived), in contrast the Mo (MPI) reflect the real situation and has
shown the decrease of 5.7 percent.
[FIGURE 4 OMITTED]
Interpretation of the Results at K-Cutoff33 percent
(i) For the Period 2011
* The incidence of poverty H = 45.76 percent indicating the
percentage of the people who are multi-dimensionally poor.
* The Intensity of Poverty A = 61.01 percent which shows that, on
average, the poor people are facing 61.01 percent of the depravations.
* The value of MPI = M0 (2011) = 0.279 which is the product of H
and A. It is percentage of those people which are multidimensional poor
as well as they are deprived at the same time.
(ii) For the Period 2007
* The incidence of poverty H = 48.71 percent indicating the
percentage of the people who are multi-dimensionally poor.
* The Intensity of Poverty A = 68.94 percent which shows that on
average the poor people are facing 68.94 percent of the depravations.
* The Value of MPI = Mo (2007) = 0.336 which is the product of H
and A. It is percentage of those people which are multidimensional as
well as they are deprived at the same time.
The results for both time periods can be summed up that the overall
Punjab has shown the decline in the poverty measured by MPI of 5.72
percent in 2011 as compared to 2007.
Urban and Rural Bifurcation of Mo
The region-wise comparison of MPI results is presented in Table 4.
In the regionwise comparison, the AF-measures have fallen in period 2011
for both the urban and rural regions. The decrease in the poverty is
found to be 6 percent for the rural areas, whereas the urban areas have
shown the fall of just 0.2 percent.
The region-wise comparison of the MPIs results for both of the time
periods is also presented in Figure 5. The results in Figure 5 reveal
the clear difference between the poverty status of urban and rural
regions and highlight the disparities faced by the rural region of the
Punjab.
In conclusion, the poverty in the rural areas of the Punjab for the
period 2011 is found to be 31.8 percent more than that of the urban
areas, whereas the poverty in the rural areas of the Punjab for the
period 2007 was found to be 37.8 percent more than that of the urban
areas. This also means that although the poverty has fallen in rural
areas of the Punjab in 2011 as compared to 2007, yet the poverty gap
between rural and urban regions of the Punjab is still evident.
[FIGURE 5 OMITTED]
Sorting by Divisions and Bands of Poverty
The Punjab province comprises of nine divisions namely Bahawalpur,
Rawalpindi, Gujranwala, Lahore, Multan, Faisalabad, Sahiwal, Sargodha,
and D.G. Khan. The results for both time periods are ranked from lowest
to the highest poverty level. On the basic of the poverty level the
divisions are classified into the low (up to 20 percent), medium (21
percent to 35 percent) and high (above 35 percent) poverty bands in this
study.
(i) For the Period 2011
The Table 5 presents MPI (2011) results for each division ordered
from lowest to highest with the classification of poverty band for the
period 2011. The D.G. Khan division has the highest MPI of 0.489
followed by Bahawalpur at 0.369, Sargodha at 0.348 and Sahiwal at 0.322.
D.G. Khan and Bhawalpur divisions fall in the high poverty band.
Faisalabad, Multan, Sahiwal, Sargodha are ranked under medium poverty
band whereas, Gujranwala, Rawalpindi and Lahore having value of MPI up
to 20 percent, categorised in the low poverty band. The graphical
representations of divisional MPI results are also shown in Figure 6.
[FIGURE 6 OMITTED]
(ii) For the Period 2007
The Table 6 presents the MPI (2007) results for each division
ordered from lowest to highest with the classification of poverty band
for the period 2007. The D.G. Khan division has the highest MPI of
0.5299 followed by Bahawalpur at 0.4782, Sahiwal at 0.4013 and Sargodha
at 0.40. Multan, Sargodha, Sahiwal, Bahawalpur, and D.G. Khan Divisions
ranked in the high poverty band. Lahore and Faisalabad are found under
Medium poverty band, whereas Rawalpindi and Gujranwala divisions are
found under low poverty band. The graphical representations of
divisional MP1 (2007) results are also shown in Figure 7.
The above findings indicate that all the divisions of the Punjab
Province are not at the similar situation with regard to the poverty
status for periods 201 land 2007. In 2011, D.G. Khan division is at
least 30 percent poorer than Gujranwala, Lahore and Rawalpindi. Whereas,
Bahawalpur and Sargodha divisions are round about 14 to 18 percent
poorer than Gujranwala and Lahore similar prevalence of disparities
among the division for the period 2007.
[FIGURE 7 OMITTED]
Division-wise Comparison of MPI
The division wise comparisons of the MPI results are presented in
Table 7. The results show decrease in poverty for all the divisions of
the Punjab except Rawalpindi division. The highest decrease is of 11
percent in the Bahawalpur division followed by 9 percent in Multan, 8
percent in Sahiwal, 6 percent in Lahore, Sargodha and Faisalabad. The
lowest decrease in poverty of just 4 percent and 1 percent is observed
in D.G. Khan and Gujranwala, respectively.
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
In conclusion, the corresponding decrease in the poverty has pushed
some divisions out of their ranked band of poverty. Particularising for
each, it is detected that Lahore division which was falling under the
medium poverty band during 2007 has decreased the poverty and now, under
the low poverty band for the year 2011. On the same lines Multan,
Sahiwal and Sargodha divisions have revealed progress and are in medium
band of poverty in 2011 as compared to 2007 when these were tumbling
under high poverty band.
The graphical demonstration of comparisons is given in Figure 8,
the corresponding increase or decrease in each division is given in
Figure 9.
District-wise Comparison of MPI
The side by side comparisons of district-wise results for MPI for
the periods 2007 and 2011 are given in Table 8. Bold figures in Table 8
show the rise in the poverty. The decrease in poverty is shown in
districts Vehari of 30 percent, Multan of 25 percent, T.T Singh of 24
percent, Pakpattan of 22 percent, Sailkot of 15 percent, Narowal of 16
percent, Khanewal of 15 percent and Rawalpindi of 14 percent. The
increase in the poverty has observed by 23 percent in R.Y.Khan, 12
percent in Rajanpur, 10 in percent Muzaffergarh, 8.5 percent in
Sheikhupura, 7.5 percent in Mianwali and 1 percent in Sargodha. The
district-wise comparisons of MPIs are shown in Figure 10, while
increases/decreases in poverty are shown in Figure 11.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
V. CONCLUSION AND RECOMMENDATIONS
Conclusion
The purpose of the study is to assess multidimensional poverty
using Alkire and Foster (AF) method for the periods 2007 and 2011 in
province Punjab-Pakistan, using primary data from Multiple Indicator
Cluster Survey (MICS). The results are bifurcated for geographical
splitups of the Punjab to further explore over time status of poverty
and monitor the disparities among different regions of the Punjab. The
calculated figures of MPI (multidimensional poverty index) for the
Punjab province at different k-cutoffs and detailed results for
particular poverty cutoff of 33 percent indicated that the overall
condition of Punjab province of Pakistan concerning to the deprivation
in the economic barometers of living standards is at the moderate level
of poverty. But the disparities and issues are evident when results are
bifurcated area, division and district wise. The rural area of the
Punjab has almost MPI at 0.40 in 2011 which means 40 percent of the
rural population is MD poor and having deprivation in the living
standards. Furthermore, the nine different divisions of the province are
found to be have isolated thresholds of MPI. D.G. Khan, Bahawalpur and
Sargodha divisions have been found to have the high values of MPI,
whereas Gujranwala, Rawalpindi and Lahore divisions are having
comparatively low values of MPI. Additionally, going shallow into
district level results the circumstances get inferior and inferior.
There are gigantic slits between different districts of the province
Punjab. In Rajanpur, D.G. Khan, Muzaffargarh, Layyah, Jhang and Bhakkar
more than 40 percent of the population is MD poor and having
deprivations. There is dissimilarity ranging from 20 to 35 percent shown
by the MPIs results of Gujranwala, Lahore, Gujrat, Faisalabad, and
Jhelum districts when paralleled with the MPIs of Rajanpur, D.G. Khan,
Muzaffargarh, Layyah, Jhang, and Bhakkar districts.
Recommendations
On the basis of the results of individual time periods and
chronological comparative findings of the study, the following
suggestion and recommendation is being depicted.
* It is clear that all the regional split-ups of the Punjab
province are not having similar standing, so the similar policies for
all over the province will not prove its worth. To allocate the
resources, there is dire need to focus on the different bands of poverty
and allocation should be made accordingly, for instance D.G. Khan,
Bhawalpur, Sargodha divisions need more care and attention as compared
to Gujranwala, Lahore and Rawalpindi divisions.
* As we have identify the divisions which are under different bands
of poverty, then utilising it as a base line we should carefully
observed the status of the poverty in the particular district of the
respective division to see which of the district should be focused first
e.g., considering D.G. Khan division having Mo (2007) = 0.5299 and Mo
(2011) = 0.4899, this division consists of four districts i.e., D.G.
Khan, Layyah, Muzaffargarh, and Rajanpur having MPI in the order at
0.50, 0.50, 0.36 and 0.46 for year 2007 and 0.47, 0.46, 0.46, 0.58 for
year 2011, respectively. From this comparative analysis of the MPI it is
perceived that the D.G. Khan and Layyah districts were having uppermost
MPI value in 2007 and 2011. They have lessened their poverty level by 4
percent each. Whereas, Muzaffargarh and Rajanpur districts were at 0.36
and 0.47, respectively in 2007 but in period 2011 they have flown up to
0.46 and 0.58, respectively. This deductive technique of identifying the
poorer of the poor with the periodic check provides guide lines to
introduce interventions in the right direction. As in the case of D.G.
Khan Division, there is a dire need to focus Rajanpur and Muzaffargarh
districts alarmingly.
* Consider Bahawalpur division having Mo (2007) = 0.48 and Mo
(2011) = 0.37, it shows 11 percent decline. This division consists of
districts Bahawalnagar, Bahawalpur, R. Y. Khan, having MPI values at
0.49, 0.47 and 0.138 for 2007 and 0.38, 0.37 and 0.365 for 2011,
respectively. Now it is evident that Bahawalpur and Bahawalnagar
districts have shown decline in poverty whereas, the R.Y. Khan District
has shown sharp rise in poverty. Here, policy makers need to focus R.Y.
Khan at the first priority.
* Decomposition of the result by indicators may also helpful for
having the particular direction for the allocation of resources.
* For the lovers of democracy, this paper may be used as evidence
that even poor democratic regime regarding living standards deprivations
is better in enhancing living standards in Punjab as compared to guided
democracy guided by General Mushraf and especially of the dictatorship.
Future Avenues
* As MICS 2014 data collection and data entry process have not been
yet completed and is in process. The findings of this study may be
generalised using data of MICS 2007, 2011 and 2014 in the measurement of
MPI.
* The sampling distribution of A and Mo can be classified and test
of goodness of fits can be performed in order to detect the underlying
distributions of each of the measures.
* Based upon the findings and evidence of the distributions, the
statistical inference and predictions can be made.
* A robust analysis of the MPI class of measures can be done. For
example, association among class of measures, Gap Analysis, Standard
Error (Precision and Accuracy) etc.
* Scientific method of assigning weights to different indicators
and dimensions may be used.
* Exiting data sets does not allow us to include more and more
indicators as the scopes of available data sets are either too narrow or
too broad. In order to include further dimensions and indicators it is
very necessary, to conduct a purpose based survey which includes all
indicators and dimensions which are dynamic and internationally
comparable in measuring MPI.
REFERENCES
Alkire, S. and J. E. Foster (2011) Counting and Multidimensional
Poverty Measurement. Journal of Public Economics 95:7, 476-487.
Alkire, S. and M. E. Santos (2010) Acute Multidimensional Poverty:
A New Index for Developing Countries. Human Development Initiative
(OPHI) Working Paper No.38, Oxford Department of International
Development Queen Elizabeth House (QEH), University of Oxford, pp.
1-133.
Alkire, S., A. Conconi, and J. M. Roche (2013) Multidimensional
Poverty Index 2013: Brief Methodology Note and Results. Oxford Poverty
and Human Development Initiative, the University of Oxford.
Alkire, S., J. M. Roche, M. E. Santos, and S. Seth (2011)
Multidimensional Poverty Index 2011: Brief Methodology Note. Oxford
Poverty and Human Development Initiative, the University of Oxford.
Ashraf, S. and M. Usman (2012) Acute Multidimensional Poverty: A
New Index for Punjab. International Journal of Scientific and
Engineering Research 3:9, 821-825.
Development Report (2003) A Compact Among Nations to End Human
Poverty. United Nations Development Programme, pp. 6-10.
Jamal, H. (2003) Mapping the Spatial Deprivations of Pakistan. The
Pakistan Development Review 42:2, 91-111.
Jamal, H. (2007) Indices of Multiple Deprivations 2005. Social
Policy Development Center, pp. 99-122.
Jamal, H. (2009) Estimation of Multidimensional Poverty in
Pakistan. Social Policy Development Center, Karachi (Research Report
No.79). pp. 3-16.
Jamal, H. (2011) Punjab Indices of Multiple Deprivations 2003-04
and 2007-08. Government of the Punjab, Planning and Development
Department Strengthening Poverty Reduction Strategy Monitoring, Punjab,
pp. 1-43.
Naveed, A. and Tanweer-ul-Islam (2010) Estimating Multidimensional
Poverty and Identifying the Poor in Pakistan: An Alternative Approach.
(RECOUP Working Paper No. 28).
Naveed, A. and Tanweer-ul-Islam (2012) A New Methodological
Framework for Measuring Poverty in Pakistan. In S. Khilji (ed.) Annual
Rrport 2012 (1st ed). Islamabad: Sustainable Development Policy
Institute (SDPI).
Niazi, M. 1. and A. U. Khan (2011) The Impact of Education on
Multidimensional Poverty Across the Regions in Punjab. Journal of
Elementary Education 21; 1, 77-89.
Oxford Poverty and Human Development Initiative and the United
Nations Development Programme (UNDP) (2010) Multidimensional Poverty
Index-Country Rankings. Human Development Report yip. 1-3.
Punjab Government of (2011) Multiple Indicator Cluster Survey
(M1CS), Punjab 2011. Provincial Report (Volume 1), Planning and
Development Department, Bureau of Statistics.
Punjab, Government of (2009) Multiple Indicator Cluster Survey
(MICS), Punjab 2007OS. Provincial Report (Volume 1). Planning and
Development Department, Bureau of Statistics.
Salahuddin, T. and A. Zaman (2012) Multidimensional Poverty
Measurement in Pakistan: Time Series Trends and Breakdown. The Pakistan
Development Review 51:4,493-504.
Salzman, J. (2003) Methodological Choices Encountered in the
Construction of Composite Indices of Economics and Social Well-being,
Centre for the Study of Living Standards, United Nations Development
Programme, pp. 4-27.
UNDP (1997) Human Development Report. United Nations Development
Programme.
Comments
This is a comprehensive research, based on huge data set. It
covered all regions of Punjab and had a broader base as considered seven
indicators of poverty. The paper is a good contribution literature on
multidimensional poverty in Pakistan. Followings observations are noted
in this paper:
(i) Last line of page 02 "Punjab the biggest province of
Pakistan, having the same poverty indicators as of Pakistan". This
statement seems to be wrong as many studies have reported that incidence
of poverty is least in Punjab or authors should give evidence in favour
of their statement.
(ii) In Review of Literature many important studies that worked out
MPI for Pakistan are not given such as, Salahuddin and Asad Zamad (PIDE,
2012), Arif (SDPI, 2012) and Niazi and Atta Ullah (PU, 2012)
(iii) What is rationale of considering these seven indicators (why
education, Health, nutrition etc are not considered)
(iv) What is rationale of giving same weight to each indicator,
when they are not of equal importance. For instance access to drinking
water is more important than Main material of roof.
(v) An excellent District-wise comparison of MPI is give (Table 8),
but reasons of differentials across districts and over time are not
given
(vi) Last point of recommendations "this paper may be used as
evidence that the worst type of democracy is even better than guided
democracy, especially of dictatorship". This is a big claim merely
on the basis of MPI, when key indicators education, health are not
considered.
Muhammad Idrees
Quaid-i-Azam University, Islamabad.
Muhammad Afzal <muhammad_afzalch@yahoo.com> is Professor of
Economics, Head of Department of Economics. Incharge Dean Faculty of
Management and Administrative Sciences, Lahore College for Women
University, Lahore. Shamim Rafique is Director General, Bureau of
Statistics, Planning and Development Department, Punjab, and Farhan
Hameed is Statistical Officer, Bureau of Statistics, Planning and
Development Department, Punjab.
Authors' Note: The authors would like to sincerely acknowledge
Bureau of Statistics, Planning and Development Department, Government of
the Punjab for ensuring timely data availability along with the supply
of questionnaire and underlying definitions of variables of each of the
indicators that make it possible to complete this research work
comprehensively on the sound technical grounds.
Table 1
Weights and Deprivation Cutoff for Each Indicator
Relative
Indicator Weight Deprivation Cutoff
Access to 1/7 A household is consider deprived if it
Drinking Water has unimproved source for "access to
drinking water" (unprotected well,
unprotected spring, pond,
tanker-truck, cart, surface, other)
Source of 1/7 A household is consider deprived if it
Sanitation has unimproved source of "sanitation
(Toilet Facility) (toilet facility)": (flush somewhere
else, flush to unknown place, pit
latrine without slab, composite
toilet, bucket, no facility/bush/
field, other).
Main Material of 1/7 A household is considered deprived if
Floor it has unimproved "floor material"
(earth/sand, dung plastered)
Main Material of 1/7 A household is considered deprived if
Roof it has unimproved "roof material" (no
roof, thatch/palm leaf, wood planks,
metal, wood)
Main Material of 1/7 A household is considered deprived if
Walls it has unimproved "walls material" (no
wall, cane/palm/trunks, dirt, bamboo
with mud, stone with mud, uncovered
adobe, plywood, cardboard/crate,
reused wood)
Cooking Fuel 1/7 A household is considered deprived if
it uses unimproved "cooking fuel"
(coal/lignite, charcoal, wood, straw/
shrubs/grass, animal dung, animal
dung, other
Assets 1/7 A household is considered deprived if
it has less than 50 percent assets of
(motorbike, computer, television, car/
van/tractor/trolly, washing machine,
air cooler/fan, motor/pump, bicycle,
fridge/ air-condition)
Table 2
H, A and Mo at Different K-Cutoffs for the Periods 2011 and 2007
2011
Head Count Average [M.sub.0]
K- Cutotf (percent) (H) Intensity (A) (MPI)
10 0.865 0.422 0.365
20 0.653 0.513 0.335
30 0.458 0.610 0.279
40 0.458 0.610 0.279
50 0.304 0.702 0.213
60 0.186 0.784 0.146
70 0.186 0.784 0.146
80 0.086 0.866 0.074
90 0.005 1.000 0.005
100 0.005 1.000 0.005
2007
Head Count Average [M.sub.0]
K- Cutotf (percent) (H) Intensity (A) (MPI)
10 0.872 0.478 0.417
20 0.667 0.581 0.388
30 0.488 0.689 0.336
40 0.488 0.689 0.336
50 0.409 0.740 0.303
60 0.303 0.799 0.242
70 0.303 0.799 0.242
80 0.169 0.865 0.147
90 0.009 1.000 0.009
100 0.009 1.000 0.009
Table 3
Comparing MPI 2007 vs. MPI 2011 at K = 33 percent
AF Measures MICS 2007 MICS 2011 Increase/Decrease
H 0.488 0.458 -0.030
A 0.689 0.610 -0.079
[M.sub.0] 0.336 0.279 -0.057
Table 4
Urban and Rural Bifurcation of MPl
2011 2007
Region H A [M.sub.0] H A [M.sub.0]
Urban 0.173 0.517 0.089 0.153 0.600 0.092
Rural 0.650 0.627 0.407 0.667 0.700 0.467
Punjab 0.458 0.610 0.279 0.488 0.689 0.336
Table 5
Sorting [M.sub.0] (2011) by Divisions
Division [M.sub.0] (2011) Bands of Poverty
Gujranwala 0.181399
Lahore 0.192033
Rawalpindi 0.206952 Low poverty
Faisalabad 0.257276
Multan 0.28914
Sahiwal 0.322424 Medium poverty
Sargodha 0.348195
Bahawalpur 0.369109
D.G. Khan 0.489913 High poverty
Table 6
Sorting [M.sub.0] (2007) by Divisions
Division [M.sub.0] (2007) Bands of Poverty
Rawalpindi 0.178248 Low poverty
Gujranwala 0.192727
Lahore 0.245671 Medium poverty
Faisalabad 0.316711
Multan 0.378095
Sargodha 0.40051
Sahiwal 0.401381 High poverty
Bahawalpur 0.478288
D.G. Khan 0.529922
Table 7
Division-wise Comparison of MPI 2007 vs. MPI 2011
Division M0(2007) M0(2011) Increase/Decrease
Bahawalpur 0.478 0.369 -0.109
D.G. Khan 0.530 0.490 -0.040
Faisalabad 0.317 0.257 -0.059
Gujranwala 0.193 0.181 -0.011
Lahore 0.246 0.192 -0.054
Multan 0.378 0.289 -0.089
Rawalpindi 0.178 0.207 0.029
Sahiwal 0.401 0.322 -0.079
Sargodha 0.401 0.348 -0.052
Table 8
MPIs 2007 vs. MPIs 2011by Districts
M0 M0
District (2007) (2011) Inc/Dec
Attack 0.222 0.206 -0.015
Bahawalnagar 0.494 0.376 -0.118
Bahawalpur 0.471 0.368 -0.103
Bhakkar 0.442 0.417 -0.025
Chakwal 0.212 0.208 -0.005
Chiniot 0.422 0.399 -0.023
D.G. Khan 0.510 0.470 -0.040
Faisalabad 0.225 0.155 -0.069
Gujranwala 0.138 0.142 0.004
Gujrat 0.121 0.105 -0.016
Hafizabad 0.366 0.305 -0.061
Jhang 0.497 0.433 -0.064
Jhelum 0.177 0.152 -0.025
Kasur 0.373 0.304 -0.069
Khanewal 0.435 0.288 -0.147
Khushab 0.446 0.369 -0.077
Lahore 0.056 0.055 -0.002
Layyah 0.507 0.461 -0.046
Lodhran 0.379 0.337 -0.042
Mandi Bahaudin 0.257 0.258 0.001
Mianwali 0.275 0.350 0.075
Multan 0.523 0.272 -0.251
Muzaffar Garh 0.361 0.465 0.104
Nankana Sahib 0.323 0.301 -0.022
Narowal 0.431 0.275 -0.156
Okara 0.383 0.338 -0.045
Pakpattan 0.573 0.354 -0.219
R.Y. Khan 0.138 0.365 0.227
Rajanpur 0.468 0.584 0.116
Rawalpindi 0.372 0.233 -0.140
Sahiwal 0.351 0.271 -0.080
Sargodha 0.288 0.298 0.010
Sheikhupura 0.135 0.220 0.085
Sialkot 0.299 0.147 -0.152
T.T. Singh 0.449 0.208 -0.241
Vehari 0.586 0.284 -0.302
