IN THE NAME OF ALLAH, THE MOST BENEFICENTAND

THE MOST MERCIFUL

Rural Infrastructural Development and
its Role in Poverty Reduction: Evidence
from Pakistan
[Research Paper to be presented in International Conference 2016 Transforming Economic Development: Policies and
Strategies at AERC, Karachi (Pakistan)]

By: Ishtiaq Ahmad

(Corresponding & Presenting Author)

Ali Azam
Babar Aziz
 Presentation

Introduction
Survey of Literature
Data and Methodology
Results and Discussion
Conclusions
 Introduction

This study attempts to explore the role of infrastructure in

poverty reduction for rural areas of Pakistan and
hypothesizes that it has a key role for this purpose.
In literature this area of research remained under dispute
widely. Proponents argue strong positive effects and
opponents raise question for resources of poor countries
to finance such investments on account of reason that
causality actually runs from poverty to infrastructure.
However, World Bank favors proponents and emphasizes
important role of infrastructure for poverty reduction.
 Introduction

Poverty has substantially been reduced in Pakistan from 40

percent in 1970s to 22 percent in 2005-06.
For future policy formulation it will be reasonable to find
out causes for this reduction and this study hypothesizes
rural infrastructural development as a cause in this regard.
This study will prove helpful not only to present evidence
from Pakistan on the issue but also contributes in rural
development studies by providing empirical investigation.
The issue in question is macroeconomic in nature but data
limitations in Pakistan lead the analysis towards
microeconomic dataset as proxy.
 Introduction

Data on poverty suggests that overall poverty in Pakistan

move side by side with rural poverty. So, it is reasonable to
work on rural sector for the issue to be explored.
OBJECTIVES OF THE STUDY
A) To explore the state of rural infrastructural
development in Pakistan.
B) To find out impact of rural infrastructural
development on poverty reduction in rural sector of
the economy.
 Literature Survey

Survey of Literature:
After having gone through a brief but comprehensive survey of
literature following guidelines have been dug out and employed in
analysis phase of the study:
Overall the studies point out strong positive role of
infrastructure for growth and poverty reduction.
Rural sector is the area of concern in studies related to
developing countries:
[Evenson (1986), Binswanger et al (1987, 1989), Looney (1992, 1997), Wanmali
and Islam (1997), Khan and Sasaki (2001), , Khandker et al (2006)]
Where the issue is explored at macroeconomic level, most of the
studies utilize time-series data with simultaneous equation
model.
 Literature Survey

Whereas at household level, cross-sectional dataset with single

equation model is considered through logit estimation
technique. One of the study also utilizes panel dataset for this
purpose.
For measuring infrastructural services enjoyed by the
communities public spending & core infrastructural variables
(for secondary and time-series dataset) and accessibility of
households and distances from infrastructural services (for
primary and cross-sectional datasets) are used.
[Saith (1981), Ahluwalia (1985), Gaitha (1989), and Ravalion and Datt (1994), Fan
et al (2000, 2002, 2003, 2004, 2005, 2006)]
As earlier mentioned data limitations in this study lead the
analysis towards household level, therefore, on the basis of
these guidelines acceptable methodologies have been taken into
consideration for evaluation purpose.
 Data and Methodology

This study has two objectives and for each objective a

separate dataset is utilized in following two different
analyses
a) Pakistan 2008 MOUZA Statistics for status of
Rural Infrastructural Development in Pakistan
b) HIES, 2005-06 for impact of Rural Infrastructural
Development on Poverty Reduction in Pakistan
 Data and Methodology:
Status of Rural Infrastructural Development

For this analysis, dataset titled Pakistan MOUZA Statistics 2008 is the
source. This is secondary dataset provided by Agricultural Census
Organization (ACO), Lahore (A subsidiary of Statistics Division, Pakistan).
MOUZA or DEH is a sampling unit of the survey and is defined on the basis
of revenue record of Pakistan
This dataset is part of series, 1971-1979-1983-1988-1993-1998-2003 & 2008,
delivered by ACO for analyzing rural development in Pakistan.
Number of Kanungo Circles: 1485
Number of Patwar Circles: 11298
Number of Total MOUZAs: 47482
Rural Infrastructural Development is measured by the distance of public
infrastructure from MOUZAs because services are not located within the
communities. Information , however, describes that in how many MOUZAs
service is available within 1 KM, between 1-10 KM, 10-25 KM, 25 to 50 KM and
more than 50 KM. This study takes into consideration number of MOUZAS
within 10 KM range for constructing an index in this regard.
 Data and Methodology:
Status of Rural Infrastructural Development
This study endeavors to construct an index (I) for rural
infrastructural development using principal variables i.e.

Where Wm are weights and Zm are principal variables. Let Xi

represent ith component of a principal variable Zm and Yi is measured
as:

and , where m is number of principal variables

Weights Wm are measured such that:

0< < 1 and = 1, with methodology:

and {ensures inverse relationship of Wm and variance}

 Data and Methodology:
Status of Rural Infrastructural Development
Principal Transport And
Variables Education Health Energy Market Banking
Communication

Commercial
Primary School Hospital/Dispensary Metaled Road Electricity Livestock Market
Bank
Diesel / Petrol On-Line
Middle School Rural Health Center Transport Grains Market
Pump Banking
High/Higher Fixed Line
Basic Health Unit CNG / LPG Fruits Market -
Secondary School Telephone
Components

Child & Mother Computer/ Vegetables

College - -
Care Centre Internet Market
Govt.
Population Welfare
Vocational Center P.C.O. - Procurement -
Centre
Center
- N.G.O. Dispensary Post Office - Seeds Shop -
Private Doctor
- - - Fertilizers Shop -
[MBBS]
- Midwife Facility - - Pesticides Shop -
- Veterinary Facility - - - -
 Data and Methodology:
Impact of Rural Infrastructure Development on Poverty Reduction

Data employed for this analysis is Household Integrated

Economic Survey (HIES), 2005-06 which is sub-sample of
Pakistan Social & Living Standards Measurement Survey (PSLM),
2005-06. Secondary sampling unit (SSU) in HIES is household
and primary sampling unit (PSU) is rural community of
household.
Each PSU includes nearly 16 SSU and PSU is representative of
revenue record having a specific Revenue Unit No and known
as MOUZA/DEH.
Three separate questionnaires (male, female and rural
community) are used for data collection purpose.
There are 1109 PSUs of the survey out of which 531 belong to
urban and 578 belong to rural areas. During this survey in total
15453 households were interviewed and this study employed
observations related to rural areas only.
 Data and Methodology:
Impact of Rural Infrastructure Development on Poverty Reduction

The dataset is latest in the sense that it provides information on

rural communities which is related to major theme of this study
i.e. rural infrastructural development.
Keeping in view that poverty of a household is a binary response
qualitative dependent variable, Logit model technique is
befittingly employed for empirical analysis.
MODEL
POVERTY: Official poverty line of Pakistan for the year 2005-06
equal to Rs. 947.47 is used as threshold against adult equivalent
consumption of households. This poverty line could be termed
suitable for this analysis on account of two reasons:
 Data and Methodology:
Impact of Rural Infrastructure Development on Poverty Reduction

a) Validation by World Bank

b) Computed for the same time period as HIES 2005-06
For independent variables two vectors are included in the
study:
a) Household level renowned socio-economic determinants
of Poverty.
b) Community level infrastructural variables on the basis of
hypothesis of the study.
Variables, their definitions and relationships with
dependent variable are defined as under:
 Data and Methodology:
Impact of Rural Infrastructure Development on Poverty Reduction

Variable Definition Expected Sign

DEPENDENT VARIABLE
POVERTY Discrete Response Variable i.e. 1=Poor & 0=Non-Poor Dependent

HOUSEHOLD LEVEL CHARACTERISTICS

Households Head Gender (1=Male & 0=Female)
HHHG Negative
Hypothesis: If gender is male than less is the chance that household will be poor
Households Head Education (number of schooling)
HHHE Hypothesis: Higher the education of household head, less is the chance that Negative
household will be poor
Households Head Age (number of years)
HHHA Hypothesis: Higher the age of rural household, less is the chance that household Negative
will be poor
Household Size (number of members)
HHS Hypothesis: Higher the size of household, more is the chance that household will Positive
be poor
Dependency Ratio (Dependents/Independents)
DR Positive
Hypothesis: Higher the DR, more is the chance that household will be poor
PR Participation Ratio (Workers/Household aged 10 and above) Negative
 Data and Methodology:
Impact of Rural Infrastructure Development on Poverty Reduction

COMMUNITY LEVEL CHARACTERISTICS

Higher the Rural Development, Less is the chance that household will be poor
Availability of Road to household. A Dummy variable with three
attributes i.e. metaled, paved, un-paved
ROAD ROAD1 (1=metaled road, & 0=otherwise) Negative
ROAD2 (1=paved road, & 0=otherwise)
ROAD3 (1=unpaved road & 0=otherwise)
ELECT Availability of Electricity to households (1=yes, 0=no) Negative
GAS Availability of Gas to households (1=yes, 0=no) Negative
Primary School (number of primary schools accessible to
PS Negative
household where he resides)
PHOND Phone Service Distance in KM from households residence. Positive
BHUD Basic Health Unit Distance in KM from households residence. Positive
 Results and Discussion
Status of Rural Infrastructural Development

COMPOSITE INDEX OF PRINCIPAL VARIABLES

Average of Components
Principal Variables Weights (Wm) Wm x Zm
of Principal Variables (Zm)
Education 0.163338 0.49923 0.0815431
Health 0.222619 0.72619 0.1616630
Transport &
0.181420 0.52456 0.0951652
Communication
Energy 0.153936 0.55372 0.0852374
Markets 0.150841 0.37695 0.0568592
Banks 0.127846 0.5 0.0639230
Sum 1 - 0.5443910
 Results and Discussion
Role of Rural Infrastructure Development on Poverty Reduction
Logit Regression Estimates of Rural Infrastructural Development
and Rural Poverty: Evidence from Pakistan
LR Statistic 1418.4160
McFadden R2 0.148451
Probability LR (Statistic) 0.0000
Number of Observations 8480

Independent
Coefficient z-Statistic Probability Odd
Variables
C 0.369631 1.905976 0.0567 1.447
HHS 0.047657 6.621234 0.0000 1.049
HHHA -0.024344 -11.186110 0.0000 0.976
HHHG -0.179198 -1.787296 0.0739 0.836
HHHE -0.126490 -16.794570 0.0000 0.881
PR -1.269385 -9.963761 0.0000 0.281
DR 0.468132 15.340570 0.0000 1.597
 Results and Discussion
Impact of Rural Infrastructure Development on Poverty Reduction

Logit Regression Estimates of Rural Infrastructural Development

and Rural Poverty: Evidence from Pakistan
Independent
Coefficient z-Statistic Probability Odd
Variables

ROAD2 -0.224418 -2.069476 0.0385 0.799

ROAD3 0.292320 3.552242 0.0004 1.340

ELECT -0.396249 -3.786653 0.0002 0.673

GAS -0.377397 -3.304925 0.0010 0.686

PS -0.010623 -1.407971 0.1591 0.989

BHUD 0.004883 2.818057 0.0048 1.005

PHOND 0.000085 0.031975 0.9745 1.000

 CONCLUSIONS
Overall status of rural development in Pakistan could be termed
as poor because most of the public services are not available to
rural inhabitants in their areas of residence.
The composite index also shows almost half of the MOUZAs are
located at a distance of more than 10 KM from all type of
infrastructural services.
It is also found out that the vectors of households
characteristics and community characteristics have shown
significant positive impact on poverty reduction in rural areas of
Pakistan (in line with international evidence).
So far as comparison of household level characteristics and
community level characteristics is concerned it could easily be
inferred that community level characteristics have secondary
role for poverty reduction while household level characteristics
have played primary role (contribution of this study).
Out of rural infrastructural public services, gas and electricity are
proved to be comparatively more important for poverty
reduction of rural areas in Pakistan.
Thanks for Patience