Rural Roads PDF

ASIAN DEVELOPMENT BANK IES: REG 2002-15
IMPACT OF RURAL ROADS
ON
POVERTY REDUCTION:
A CASE STUDY-BASED ANALYSIS
October 2002
ABBREVIATIONS
ADB − Asian Development Bank

DFID − Department for International Development
FRIP − Fifth Road Improvement Project
INO − Indonesia
km − kilometer
ha − hectare
NGO − nongovernment organization
NWP-WRDP − North Western Province Water Resources Development Project
PHI − Philippines
PRA − participatory rural assessment
SIADP − Sorsogon Integrated Area Development Project
SPRIP − Southern Provincial Roads Improvement Project
SRI − Sri Lanka
TCSSP − Tree Crop Smallholder Sector Project
TLRP − Third Local Roads Project
VOC − vehicle operating cost
NOTE
In this report, “$” refers to US dollars.
Operations Evaluation Department, IE-68

CONTENTS
Page
EXECUTIVE SUMMARY ii
I. INTRODUCTION 1
A. Background 1
B. Study Objectives and Scope 2
C. Methodology 2
II. CASE STUDY PROJECTS AND SITES 4
A. Study Projects 4
B. Study Roads 5
C Study Sites 6
III. UNDERSTANDING POVERTY 8
A. Who Are the Poor? 8

B. Key Poverty Themes Common to All Study Locations 9
IV. CONSIDERATIONS FOR PROJECT DESIGN AND IMPLEMENTATION 11
A. Considerations for Project Design 11

B. Considerations for Project Implementation 14
V. KEY IMPACTS AND RELATED ISSUES 16
A. Changes in Transportation Services 16

B. Changes in Travel Patterns 20
C. Changes in Village Profile 22
D. Changes in Income and Welfare 27
VI. CONCLUSIONS AND RECOMMENDATIONS 33
A. Study Conclusions 33
B. Recommendations for Strengthening Poverty Reduction Impacts 35
APPENDIXES
1. Methodology Details 38
2. Sample Framework for Documenting Field Research 47
3. Poverty Details 50
4. Study Roads, Project Sites, and Control Sites 59
5. Case Study Details in the Philippines 65
6. Case Study Details in Sri Lanka 80
7. Case Study Details in Indonesia 96
8. Household Case Studies from Along the Project Roads 108
EXECUTIVE SUMMARY
Participatory poverty assessments have long identified remoteness and isolation as

critical components of poverty. Although it is widely assumed that investments in rural roads
reduce poverty, there is little evidence of the ways in which these impacts occur or what their
determinants are. Through the collection of empirical evidence from a cluster of case studies
drawn from past Asian Development Bank (ADB) operations, the study addresses this void. The
objective of the study is to improve the design of rural road components to achieve sustainable
benefits for the poor. Realizing that pragmatic recommendations need to capture the real-life
impediments that often plague project design assumptions, the study focuses narrowly and
deeply on selected case study villages within a project area. This enables an understanding of
the factors that influence impacts on poverty.
Rather than relying on income poverty line definitions, the study analyzes poverty as a
multidimensional concept. It adopts an assets-based approach to understanding poverty, where
poverty is defined as a deprivation in assets and entitlements essential to life, and a
susceptibility to periodic physical and economic shocks, and seasonal crises. The study
compares project sites with control sites and uses a variety of tools to gather relevant qualitative
and quantitative data and to validate findings.
Undoubtedly, in all case study projects, the poor and very poor benefited substantially
from social impacts of rural roads through access to state services in areas such as health,
education, agricultural extension, and provision of information. Improved rural roads create the
conditions for better access of people to services, and of services to the village. Roads allow
regular contact with the outside world and bring remote areas within the purview of the state and
other networks. Such improvements reduce the perception of isolation and remoteness among
the poor and very poor. These social impacts could not be quantified rigorously due to the lack
of baseline socioeconomic data and periodic monitoring.
Benefit monitoring and evaluation systems were part of the ADB design of most study
projects, but were exclusively the responsibility of the executing agencies, which were often
expected to meet the cost from counterpart funding. In a situation of scarcity of local resources,
such systems were often abandoned, or allowed to lapse for so long that information that was
eventually collected was largely useless. The importance of baseline surveys and data
monitoring needs to be realized by all stakeholders if rigorous impact evaluations are to be
carried out in the future to further improve project design.
The study shows that the context within which economic impacts take place was often
determined by conditions such as climate, agricultural potential, spatial position and proximity to
networks, and world market commodity prices, as well as social structure and concentration of
assets. Although these conditions cannot be affected by road development, their careful
consideration during project identification and design would enable a better assessment of
potential for poverty reduction through such projects, while possible complementary measures
could be considered to increase positive impacts.
In the case study areas, economic benefits achieved were clearly different for different
socioeconomic groups. Most of the journeys made by the rural poor are for subsistence tasks.
For them, access to local facilities and the primary transport network is critical during times of
need, especially for health and social reasons. Besides, the poor lack both time and energy; and
impacts that either reduce or exacerbate these deficiencies have a critical bearing on poverty.
The poor rely on the primary network of paths, tracks, culverts, and basic access routes in the
immediate village vicinity. Theirs is a village, walking world, and improvements to this primary
village network that reduce the burden in undertaking basic household and productive tasks, as
iii
well as the increased availability of intermediate modes of transport with larger carrying capacity
to collect water, firewood, etc., are likely to have a greater initial impact on their well-beings than
improved availability of motorized transport services, which they do not or cannot afford to use,
and which are not linked to their existing livelihood strategies. Therefore, improving transport
within a village is as important to the poor and very poor as providing access to markets outside
the village.
The rural roads studied have provided an important economic safety net allowing
alternative livelihood opportunities. Such an alternative income stream, even if temporary or
seasonal, is still important for household food security. A good road surface and the guarantee
of all-year accessibility are important prerequisites for the development of any kind of regular
enterprise. With increased opportunities for mobility, however, there are also increased risks of
exposure to negative influences from nearby urban centers such as drugs and sex trade.
In the study areas, the relationship between rural road investments, spontaneous private
sector transport provision, and availability of cheaper transport services relied heavily on the
competitiveness of the transport markets. Reductions in the vehicle operating costs were
evident, but passing these on to consumers depended on the level of competition and
contextual factors. Rural roads, by definition, often serve remote areas where population density
is low, and market information imperfect. In the absence of competition, there is no incentive for
transport providers to improve their service or pass on their cost savings to consumers.
Competition is determined by potential demand for travel, users’ proximity to a

commercial center, the government’s commitment to regular maintenance, and regulatory
barriers. Rural roads, particularly gravel roads, quickly deteriorate if not regularly maintained,
and benefits can be quickly lost if they are periodically impassable or the overall condition is
bad. Competition then is minimal as transport providers gravitate toward areas with better
roads. The poor are generally risk averse and will not engage in a new activity if they know that
the road on which it depends will be temporarily unusable or if its poorer condition in the
following year will mean that costs and time will increase.
Problems of maintenance are compounded when it is unclear where the institutional

responsibility for rural roads lies. Regular maintenance is often neglected both because of lack
of funds and because there is little political capital in carrying out minor maintenance; major
rehabilitations would be authorized once the road has deteriorated very much, making it difficult
to maintain a guaranteed level of access. Devolving responsibility for rural road maintenance to
local communities supported by the technical expertise of local authorities yields direct
employment benefits, builds ownership among local communities, and promotes sustainability.
One very important way in which the poor can feel the direct economic benefits of rural
roads is through labor-based road construction and maintenance. Unfortunately, only one
project out of the six studied had a labor-based road component. Some authorities in the study
countries appear to believe that labor-based construction methods are slow and their quality
doubtful. But for low-traffic rural roads, the time and efficiency argument in construction or
rehabilitation is surely not so important, as the opportunity cost of disruption to traffic is very low.
Experience from Asia and Africa shows that, given a sufficiently long period of employment on
the road, the poor can escape chronic debt cycles and thus move away from poverty.
Vehicle operators and traders are often winners with both good and bad rural roads.
They can monopolize transport routes and the buyers’ markets along poor roads where
competition is weak, and thus dictate prices. They also benefit from better roads through lower
vehicle operating costs, time savings, as well as easier transportation and higher volumes.
Often, indebtedness of the poor allows traders to dictate prices even in the midst of competition.
iv
They are chronically indebted to these traders, and so have little scope for maneuver in deciding
whom they sell to and at what price. Better access does not necessarily lead to better prices for
poor farmers.
A traditional assumption on the mobility of the poor is that their lack of agricultural
assets, particularly land, makes them more likely to seek employment outside the community
and that road access helps this substantially. However, labor markets in remote rural areas are
imperfect, and accessing opportunities is difficult, particularly where there is a lack of
information. This lack of information and inability to command rights over work opportunities are
themselves a function of poverty. Better-off households are much more likely to have access to
information on well-paid, or stable, outside employment, while the poor and very poor access
only temporary, seasonal, and unskilled work opportunities, which are usually poorly paid. Lack
of education and lack of support networks in more urban areas compound this trend.
Roads are clearly a critical enabling condition for improvement of living conditions in
rural areas. However, the distribution of economic benefits is a separate issue, and there are no
guarantees or inherent mechanisms to ensure that economic benefits will be distributed
equitably between the poor and nonpoor in communities. In the rural road projects studied, their
ability to affect the distribution of assets and the skills capacity of the poor was limited and
largely outside their scope. Nevertheless, recognizing how assets are distributed is important
both for understanding how benefits will accrue and for planning complementary measures to
enable those who lack assets also to benefit from the investment. Given the right
complementary activities, projects can broaden livelihood opportunities. The poor need support
to make use of the opportunities that rural roads may bring. This suggests that integrated
projects are needed to tackle poverty effectively.
The case studies covered projects that were both sector road investments and
integrated projects, where the road was one part of a larger program. In practice, the latter were
either not truly integrated or were focused largely on benefiting better-off farmer groups. The
poor require genuinely integrated programs of support right through the cycles of production,
transportation, and sale. In order for the poor to travel for productive purposes, the provision of
transport services must be linked to some livelihood and income diversification activity, which
builds on or supplements their existing subsistence activities. For such a scheme to be
sustainable, it must eventually be self-financing. Implementation of integrated rural roads
projects is difficult and the contextual situation differs from place to place. ADB may suggest to
borrowers to take on partners in program design and implementation such as other cooperating
external assistance agencies, or local or international nongovernment organizations that have a
proven track record in mobilizing and working closely with communities.
In addition to the recommendations above, the study provides a range of other

recommendations regarding project design (on participatory design and planning, and poverty
assessment), as well as implementation (working with partners and project performance
monitoring). It also acknowledges that governments have a critical role in this endeavor such as
facilitating a regulatory environment for competitive transport services and participatory
selection of roads to be improved, and promoting understanding of the priority of poverty
reduction among its agencies.
I. INTRODUCTION
A. Background
1. Need. Although it is widely assumed that investments in rural roads reduce poverty,
there is little systematic analysis or evidence to date either of the ways in which rural roads
actually impact on the poor or of what the impact process is. Designs of projects financed by the
Asian Development Bank (ADB) have often tried to capture the direct and quantifiable costs and
benefits associated with rural road investments. Following its adoption of the poverty reduction
strategy, ADB is carrying out much research and analysis into the poverty dimensions of its
operations. Poverty reduction has become a much more explicit imperative, and this is reflected
to some extent in recent proposals for infrastructure investments.1 It is now increasingly
recognized, however, that there are also other indirect impacts from road investments that may
have either a positive or negative effect on poverty reduction. To date, it has been difficult
empirically to capture evidence of these indirect benefits in a systematic way. This study
attempts to address this void through the collection of empirical evidence from a cluster of case
studies drawn from past ADB operations. It is primarily intended to address two questions:
(i) How do rural roads help reduce poverty? And (ii) How can rural road projects be designed to
help reduce poverty more?
2. Focus. The study was initiated in response to requests from ADB staff to better
understand what the impact of rural road investments was on poverty reduction. It was
developed as a retrospective study that assessed the impacts of past road investment.2 It did
not have the luxury of identifying, monitoring, and comparing impacts before, during, and after
the road investment.3 However, it had the advantage of in-depth fieldwork in a variety of case
study villages and systematic analysis of both quantitative and qualitative data collected and
verified through an array of tools. It focused particularly on poverty and transport linkages that
stem from rural road investments, rather than on the overall impacts of these roads on the rural
economy and the distribution of impacts across different segments of society. It is expected that
lessons learned and discussed here will increase understanding and strengthen the
effectiveness of poverty reduction of future ADB operations, especially those concerned with
rural roads and transportation.
3. Structure. The case studies selected are sufficiently varied and the methodology
adopted is effective to give a range of useful lessons. The report is structured to allow future
designers and implementers of road development projects to grasp the key issues that are
relevant at different stages in the project cycle and to understand the extent to which
outcomes/impacts may be affected by contextual factors. Chapter I details the background for
the study and summarizes its methodological approach. More details on methodological
approach are given in Appendix 1 for the benefit of evaluators and other researchers. Chapter II
gives a brief introduction to the projects and case study contextual scenarios. Chapter III
presents the study’s approach to defining poverty and factors that influence poverty in the case
study areas. These two chapters give real-life considerations that may shape the assumptions
that have traditionally been made about the impact of rural roads on poverty reduction. Chapter
IV, based on study findings, indicates what is important to consider when designing and
1
Loan 1385-CAM: Rural Infrastructure Improvement Project, for $25.1 million, approved on 28 September 1995, and
still under implementation, has been an early attempt to incorporate poverty dimensions in infrastructure projects.
2
The study team comprised H. Hettige (Senior Evaluation Specialist and Study Manager), P. Chaudhry
(International Social Development Specialist), and three country-specific specialists on poverty and transport,
namely G. Bimbao (Philippines), N. Gunetilleke (Sri Lanka), and A. Rohandi (Indonesia), who were supported by
groups of enumerators in each of the study areas.
3
Studies of this nature are under way (e.g., Van de Walle. 1999. Choosing Rural Road Investments to Help Reduce
Poverty. World Bank. Washington, DC that describes the initial stages of the study).
2
implementing rural road projects. Chapter V is devoted to a discussion of those key impacts that
may occur immediately following a road improvement project and those that may take time.
Issues that emerge from these impacts then lead to the conclusions and recommendations in
chapter VI. Details on each of these chapters, individual household case studies, and selected
photographs are presented in the appendixes.
B. Study Objectives and Scope
4. Objectives and Scope. The study is limited in its objectives. It focuses only on rural
roads and on how they relate to poverty reduction. Urban roads and road networks are taken as
a given in the study. The study’s overall objective is to learn how to better design the rural road
components of projects in the future to achieve sustainable benefits for the poor. Within this
overall objective, the following questions guided the study direction: (i) What are the different
kinds of impacts that rural roads have on poverty? (ii) How can we effectively capture the factors
that lead to these impacts? And (iii) What types of complementary services and considerations
are essential to accelerate the beneficial impacts of rural roads to the poor? The study findings
are intended to complement and add value to existing knowledge and work being undertaken.4
5. Case Study Approach. The study uses past ADB road improvement project
components5 to understand and capture the impacts of rural roads on poverty reduction. Most of
these projects were approved before ADB adopted its poverty reduction strategy and were not
designed specifically with poverty reduction as their main objective. The study, therefore, does
not attempt to evaluate these ADB project investments. Instead, it consciously adopts a forward
looking, lesson learning approach. Realizing that lessons and pragmatic recommendations need
to capture the real-life impediments that often vex project design assumptions, the study team
chose to focus narrowly and deeply on selected case study villages within each project area.
This enabled the study to capture both direct and indirect impacts, allowed it to focus on the role
of road transport services and accessibility within the broader socioeconomic-cultural context of
a village, and permitted it to understand the process and factors that influence impacts of rural
roads on poverty. A wider coverage of the project area would have prevented the study from
grasping the factors that underlie the impacts. In order to better design future projects, the
opportunity to understand the underlying factors and thereby learn how to influence poverty
reduction was considered more useful than providing irrefutable statistical evidence that roads
may have many beneficial impacts on the poor. This deeper understanding will enable ADB and
its developing member countries, as well as other external assistance agencies, to design
interventions that maximize beneficial impacts.
C. Methodology
6. Assumptions and Tools. The study methodology was carefully designed to maximize
the use of both qualitative and quantitative information available for a retrospective impact
evaluation of this nature. The study was based on the several methodological assumptions that
broadened its sphere of inquiry.6 It did not assume an automatic link between rural roads and
poverty reduction, but considered the multifaceted impacts that determine how people respond
to improved rural roads and shape livelihood constraints and opportunities. It questioned
whether rural roads would automatically lead to better rural transport. It accepted that poverty is
4
In particular, the ongoing work under TA 5947-REG: Assessing the Impact of Transport and Energy Infrastructure
on Poverty Reduction, for $800,000, approved on 25 October 2000, which has a wider scope and covers the
impact of infrastructure investments (several types of transport infrastructure and energy) on poverty reduction.
5
The projects considered in the study were components of broader transport sector projects or integrated rural
development projects.
6
Appendix 1 describes in detail the evaluation approach used, methodological assumptions, and project selection
criteria.
3
a multidimensional condition, and that lack of income is only one component. It focused on both
physical and nonphysical accessibility constraints that the poor may have, and reviewed both
direct impacts that are often felt immediately and indirect impacts that take time to be felt. It
recognized that women and men have different productive and household responsibilities and,
therefore, different transport needs. To capture this broad sphere of influences, it focused on
key impact indicators using available secondary statistics, and relied on classical road impact
assessment tools such as traffic and passenger surveys and changes in vehicle operating costs
(VOCs). In addition, it collected data from household surveys, key informant interviews,
participatory rural assessments (PRAs), and feedback workshops. These tools were used
sequentially, each intending to inform the next phase and cumulatively to validate the data. The
qualitative data were fed into a framework structure that captures not only the outcomes of the
project intervention but also the factors that affected the impacts.7 The use of different tools
ensured effective cross-checking and validation through the triangulation of findings8 for study
robustness.
7. Case Study Selection.9 Based on discussion with ADB staff working on roads and rural
development, suitable projects were identified in selected countries. Three countries, with two
projects in each, were selected after consideration of a variety of conditions in the Asian region
such as coverage of other ongoing studies, type of ADB projects that finance rural roads, and
practical limitations driven by resources, security concerns, and time. The countries selected
were Indonesia, Philippines, and Sri Lanka. Cumulatively, the cluster of field sites selected from
these projects covered a broad range of both physical and nonphysical factors likely to condition
the context for rural road interventions across ADB operations.10 One of the projects selected in
each country was a sector-based road investment, and the other an integrated rural
development project with a rural roads component. In selecting these different kinds of projects,
the intention was to compare the impact of a road investment on its own or as part of a wider
program. From each project, a road segment was selected as a case study area. Road
segments in poor districts (where the incidence of poverty was high) were purposely selected
during this process, as the study focus was the impact of roads on poverty reduction.11
8. Attribution. The most difficult part of any impact evaluation is attribution of impacts to a
particular intervention. This is because the impacts (i.e., on poverty) may come not only from
rural roads but also from other types of development activities in a particular area or community.
To attempt to attribute particular impacts to rural roads, adequate control mechanisms are
needed. Since projects under consideration had already experienced the rural road intervention,
it was not possible to use an experimental design technique where control groups were
established randomly. Instead, the initial intention of the study was to use a quasi-experimental
design, which utilizes control groups that resemble the intervention groups. Then, the double
difference technique12 could be used to attribute the impacts of the intervention (in this case a
rural road).
7
A sample framework for documenting field research is given in Appendix 2.
8
Triangulation of findings refers to the validation of study results through three key methods: (i) household and traffic
surveys, (ii) PRAs and key informant interviews, and (iii) key impact indicators.
9
Details on country, project, and case study area selection are presented in Appendix 1.
10
Since different countries adopt different definitions for rural roads, the study selected basic roads funded by ADB in
predominantly rural areas.
11
Therefore, the findings from these case study locations may not necessarily be applicable across the entire area
covered by each of the projects considered. The study team discussed this at the methodology workshop, and it
was advised by the workshop participants that since the study focus was to understand how best rural roads
projects can contribute to reducing poverty and not to evaluate the projects, using relatively poor locations as case
study sites was in line with the study objectives. Appendix 3 analyzes the concept of poverty in the case study
locations.
12
The double difference technique is a tool that uses the differences in impacts before and after an intervention and
also between a project site and a control site (with and without project) to attribute impacts of the intervention.
4
9. The two crucial ingredients needed for the double difference technique were the
selection of suitable control sites and good baseline information. Control sites were to be similar
to the study locations in every way, except that they did not have a road.13 After surveying
several road segments, the team selected areas that proved to be suitable control sites with
minimum selection bias and that did not compromise the quality of findings. However, collection
of baseline data on the preproject situation proved to be difficult and data collected were
inadequate. Instead, recall techniques were tried in an attempt to compare before and after
conditions in a qualitative sense. Even the recall techniques proved to be quite difficult to use in
the case study areas. The road appeared not to be one of the most important determinants in
the villagers’ routines and, therefore, road-related information could not be easily recalled. In the
end, given the absence of useful baseline information or detailed recalled information, it was
difficult to use even a modified double difference technique precisely, thus making rigorous
attribution of impacts impossible.14 Instead, the control sites were used to understand and
compare the difficulties that the poor face without accessibility, and the extent to which easing
this constraint would improve their status. This enabled the comparisons of “with” and “without
project” situations to a certain extent.
II. CASE STUDY PROJECTS AND SITES
A. Study Projects
10. The objectives of the case study projects and the components that were designed to
achieve these objectives are set out in Table 1.
Table 1: Project Objective and Components

Project Objectives
Approval/Completiona Road Components Other Components
Sorsogon Integrated Area Development Project (SIADP)—Philippines (PHI)
To reduce rural poverty in Sorsogon by Improvement and Rehabilitation of 15 communal
increasing incomes, generating employment, rehabilitation of 156 irrigation schemes
and improving the standard of living of kilometer of national Provision of health and water supply
subsistence farmers and fisherfolk. secondary and rural roads facilities to control schistosomiasis
Provision of plant nurseries, an abaca
Approval: 1988 Amount: $24.1 million tissue culture laboratory, artificial reefs,
Completion: 1997; study road was and strengthening of extension
rehabilitated and graveled in 1995 services
Study Area: Sorsogon Project management support
Fifth Road Improvement Project (FRIP)—PHI
To reduce transport constraints, increase Improvement of national
efficiency in transport services, and stimulate roads and rural roads in four
economic development. islands
Periodic maintenance of
Approval: 1990 Amount: $150 million national roads and rural
Completion: 1997; study road was roads
rehabilitated and asphalted in 1996 Procurement of road
Study Area: Negros maintenance equipment
Consulting services
a
Completion indicates the year in which the entire project operations were completed. The study road may have
been rehabilitated prior to that.
13
Initially, the team had difficulties in identifying suitable control locations due to the high density of rural road
networks, but subsequently this problem was solved. To avoid inherent selection bias of control sites, every
attempt was made to find areas that had the same agricultural potential as that of a project site.
14
The importance of implementing baseline surveys and collection of socioeconomic monitoring data needs to be
realized by all stakeholders if rigorous impact evaluations are to be carried out in the future.
5
Project Objectives
Approval/Completion Road Components Other Components
North Western Province Water Resources Development Project (NWP-WRDP)—Sri Lanka (SRI)
To improve the economic, social, and Rehabilitation and Rehabilitation, improvement, and
nutritional well-being of the people living in improvement of rural roads restoration of minor irrigation systems
the rural areas of the province through Credit lines for on-farm irrigation
increased agricultural production and crop systems and to rural women for
diversification and employment opportunities. establishing income-generating
activities
Approval: 1992 Amount: $30 million Institutional support through
Completion: 1999; study road was equipment, training, and consulting
rehabilitated and graveled in 1998 services
Study Area: Kurunegala
Southern Provincial Roads Improvement Project (SPRIP)—SRI
To raise the income and quality of life in the Civil works for road and
project area through creation of income- bridge rehabilitation
generating opportunities and strengthening of Capacity building, training,
the economic and social infrastructure. and provision of mobile road
maintenance equipment,
Approval: 1997 Amount: $30 million etc.
Completion: Still ongoing; study road was Consulting services for
rehabilitated and asphalted in 2001 project implementation
Study Area: Matara
Tree Crop Smallholder Sector Project (TCSSP)—Indonesia (INO)
To reduce poverty by improving the income Construction of plantation Plantation establishment and
and employment prospects of rubber and tea roads maintenance
smallholders and landless rubber tappers, of Support services including project
whom over 50% lived below the poverty line. management
Equipment, training, and consulting
Approval: 1991 Amount: $135 million services
Completion: 2000; study (farm) road was
constructed in 1993 and the district road was
asphalted in 1996
Study Area: Bengkulu
Third Local Roads Project (TLRP)—INO
To improve the road network in selected Rehabilitation of paved and
provinces and support periodic maintenance. unpaved roads
Replacement of bridges
Approval: 1993 Amount: $200 million Periodic maintenance
Completion: 1997; study road was asphalted Provision of road
in 1997 maintenance equipment and
Study Area: Yogyakarta consulting services
Source: Asian Development Bank.
B. Study Roads
11. From each project, a study road was selected as a case study.15 These roads are
broadly similar in that all are basic roads with seasonal and market day fluctuations in traffic
volume, primarily serving localized travel patterns and travel needs. The six roads can be
classified into two distinct groups. The three roads that were completed as part of integrated
projects are all gravel surfaced and located in clearly rural areas. They primarily serve a farm-to-
market purpose. Two of the roads (in Kurunegala, Sri Lanka and Bengkulu, Indonesia) can be
considered basic access roads, as they serve no strategic purpose on the localized network.
The three asphalted road segments were all completed as part of a road sector investment
program and traversed rural areas selected for the study. The volumes of traffic on these roads
15
Appendix 4 briefly describes the road segments studied and compares the project sites and control sites.
6
are higher, and there is some evidence that two of them (in Negros, Philippines and Matara, Sri
Lanka) are serving as bypasses for major traffic on the road network. All of the study road
segments demonstrate, to different degrees, a mix of “traditional” nonmotorized and pedestrian
traffic, and “modern” vehicular traffic. Table 2 summarizes the characteristics of these road
segments and their transport conditions.
Table 2: Description of Case Study Roads and Their Impacts
Integrated Projects Transport Sector Projects

Condition Sorsogon Kurunegala Bengkulua Negros Matara Yogyakarta
Country Philippines Sri Lanka Indonesia Philippines Sri Lanka Indonesia
Rehabilitated 1995 1998 1993/1996 1996 2001 1997
Length (km) 21.0 16.5 4.0/16.0 13.5 10.5 10.5
From Bulan Ambale Talang Padang Moises Padilla Urubokka Semanu
To Magallanes Inginimitiya Tanah Abang Guinpana-an Katuwana Giri Panggung
Bypass No No No Yes Yes No
Surface Gravel Gravel Gravel/Asphalt Asphalt Asphalt Asphalt
Condition Good Good Poor/Good Good Good Good
Increase in Moderate Moderate Little/Moderate Heavy Heavy Heavy
Transport
Volume
Modes of 3-Wheelers Bicycles Wheelbarrow/ Motorcycles Motorcycles Minibus
b
Transport Jeepneys Bus Trucks 3-Wheelers Trucks 3-Wheelers
Bicycles Trucks Bus Bus/Trucks Bus Trucks
Competition in Not much Not much Not much Competitive Competitive Competitive
Transport
Services
Transport Price Small Nonec Small Moderate Good Moderate
Reduction
Travel Time Reduced Reduced Reduced Reduced Reduced Reduced
km = kilometer.
a
Since the project road was a graveled farm road with little traffic, for the purposes of the study, this and the link
district road subsequently asphalted by the Government were considered together. Their details are shown
separately in one column.
b
Jeepneys are locally-made vehicles which are like minibuses.
c
Subsidized bus available before and after the road improvement.
Source: Study team.
C. Study Sites
12. Six project site locations were selected from along the project roads. Primary research
activities were focused in these communities. Six control locations (similar to the project sites)
were also selected for research; these were close to the study sites, but far from a motorable
road. Table 3 describes the similarities in the project site and its corresponding control site and
also contrasts their accessibility differences.16 Therefore, the control sites were well suited to
compare the with and without project situations.
16
Brief descriptions of the project sites and control sites, as well as photographs of accessibility constraints to control
sites, are in Appendix 4.
7
Table 3: Similarities and Differences Between Project Sites and Control Sites
Study Area/Project Project Site Control Site
Sorsogon/SIADP (PHI) Barangay Palale Barangay Bical

Main Production Coconut with little intercropping, Coconut with little intercropping,
highly dependent on copra price highly dependent on copra price
Land Ownership/Employment Mostly tenant farmers; few Mostly tenant farmers; few
landowners; others are agricultural landowners; others are agricultural
laborers laborers
Accessibility All-year access on study road Inaccessible during rainy season;
takes 30–40-minute walk on track
over two rivers in dry season
Facilities Elementary school in the barangay Elementary school in the barangay
Total Households 96 96
Survey Respondents 31 40
Negros/FRIP (PHI) Barangay Magallon Cadre Barangay Macagahay

Main Production Sugarcane Corn for subsistence; sugarcane not
grown due to lack of road network
Land Ownership/Employment Landless seasonal labor and Smallholders of land; employment
absentee landowners of large opportunities scarce for landless
plantations
Accessibility On project road with high volume of Across footbridge on river plus 45
through traffic minutes difficult walk; inaccessible
during rainy season
Facilities Elementary school and health center Elementary school; no useful health
center
Kurunegala NWP-WRDP (SRI) Nugannoruwa Walahinikalla and Wembuwa

Main Production Rice with some dry land cultivation Rice with some dry land cultivation
Land Ownership/Employment Mostly small farmers and some wage Small farmers
laborers; a few large farmers
Accessibility Adjacent to study road 45-minute walk to bus route on
tracks across fields
Facilities Insufficient water for cultivation, Water available in village tank; have
community hall, preschool to access project site facilities
Matara/SPRIP (SRI) Heegoda Makiliyathanne

Main Production Tea and some spices Tea and some spices
Land Ownership/Employment Mostly smallholders and laborers Mostly smallholders and laborers
Accessibility Adjacent to study road Upper slopes accessible by winding
stone paths; very difficult
Facilities No community organization; school No community organization; school
on main road on main road
Bengkulu/TCSSP (INO) Talang Kabu-Talaatan Talang Padang

Main Production Rubber, coffee, paddy, and fruits Rubber, coffee, paddy, and fruits
Land Ownership/Employment Smallholders and wage laborers Smallholders and wage laborers
Accessibility Next to district road and farm road Track across river on small
providing access to fields suspension bridge
Facilities Elementary and high school on Elementary school in village
district road
FRIP = Fifth Road Improvement Project, INO = Indonesia, NWP-WRDP = North Western Province Water Resources
Development Project, PHI = Philippines, SIADP = Sorsogon Integrated Area Development Project, SPRIP = Southern
Provincial Roads Improvement Project, SRI = Sri Lanka, TCSSP = Tree Crop Smallholder Sector Project.
8
Study Area/Project Project Site Control Site
Yogyakarta/TLRP (INO) Candirejo Village Sumber Wungu-Kelayu

Main Production Rice, cassava, groundnuts, banana Rice, cassava, groundnuts, banana
Land Ownership/Employment Smallholders and laborers Smallholders and laborers
Accessibility Study road runs through site 6 kilometers from asphalt road; walk
on rocky paths to reach site
TLRP = Third Local Roads Project.

Source: Study team.
III. UNDERSTANDING POVERTY
13. This chapter briefly describes the concept of poverty used in the study. A detailed
description is provided in Appendix 3. Case study findings were used to understand what
poverty means across countries, within a country, across communities, and within a community.
Despite this relative concept of poverty in the community, there are key poverty themes that
were common to all locations and these are presented later in the chapter.
A. Who Are the Poor? Figure 1: Relationship of Livelihood Assets to Structural

Conditions, Institutions, and Processes at Work
14. Definition. Poverty is here defined as a
deprivation in assets and entitlements essential to STRUCTURAL CONDITIONS
life, and a susceptibility to periodic physical and CAPITAL ASSETS
economic shocks, and seasonal crises. It also Natural
describes a state of voicelessness and an inability to Social

Human
influence the structures, institutions, and processes

that shape rural livelihoods. Fieldwork revolved PROCESSES PROCESSES
Physical Financial
around the analysis of livelihoods through using a
capital assets-based approach.17 It was assumed Livelihood Livelihood
that people had five principal types of capital assets, Outcomes Strategies
and the relative strengths and weaknesses of these

assets determined vulnerability, susceptibility to INSTITUTIONS
shocks, and ability to participate in wider institutions

and processes. The relationship of the assets Source: Carney, Diana, ed. 1998. Sustainable Rural
Livelihoods: What Contribution Can We Make?
(physical, financial, natural, human, and social) to Department for International Development, United
the institutions, structures, and processes at work is Kingdom.
represented in Figure 1.
15. Poverty Groups. The study does not rely on income poverty line definitions but
analyzes poverty as a multidimensional concept.18 Poverty was defined by the communities
themselves. Even within a country across communities, material designation of poverty can be
very different.19 Therefore, the study sought to understand how communities themselves define
and understand relative socioeconomic status. This relational approach was also intended to
capture better some of the complexities and dynamisms of the broader aspects of vulnerability
present in each community, which are of critical interest to the study in seeing what the
necessary conditions are for the poor to make use of rural roads. Using PRA techniques in each
location, the study captured how the villagers grouped themselves according to perceived
17
Carney, Diana, ed. 1998. Sustainable Rural Livelihoods: What Contribution Can We Make? Department for
International Development, United Kingdom; and Hanmer, Lucia, Elizabeth Lovell, Robert Chapman, and Tom
Slaymaker. 2000. Poverty and Transport: A Toolkit. Overseas Development Institute.
18
Therefore, the concept is not directly comparable with the income and headcount definition of poverty.
19
See examples from the Kurunegala and Matara sites in Appendix 3, para. 6.
9
poverty status. These classifications were compared with self-perception inquiries in the
household survey and a poverty index built using information within the survey. All three
sources (PRA classification, self-perception, poverty index) classified the socioeconomic groups
into three distinct groups, and their approximate size was validated by all three exercises.
16. The socioeconomic groups were classified into the following three distinct groups (for
details, see Table A3.2 in Appendix 3):
(i) The “structural” poor or very poor. The members of this group face structural,
long-term factors of lack of access and entitlements to land and resources, which
severely restrict their ability to produce beyond subsistence levels. They are most
susceptible to crises resulting from natural disasters, as they have a small asset
base and few networks on which they can rely for support. Often, families with a
high care factor (where members are elderly, sick, or disabled) are heavily
represented in this category. In the six project sites and six control sites, the
proportion of the very poor ranges from 10% to 40%.
(ii) The “transient” poor or poor. This is a broad categorization covering a range
of people who are extremely vulnerable to shocks and fall into “severe”20 poverty
at times, but who also have some asset base or entitlements, which gives them,
potentially at least, the opportunity to move out of poverty. Many may slip into
severe poverty through seasonal vulnerability. Others in the top end of this
bracket are progressively securing themselves against poverty by building up
their asset base. Depending on the site, the proportion of the poor is between
45% and 70%.
(iii) The better off.21 These households have an income and assets beyond their
subsistence needs and can be categorized as having diverse livelihood options,
wide networks often extending beyond the immediate community, and a strong
and diverse asset base. Their proportion in the case study locations varies
between 10% and 40%.
B. Key Poverty Themes Common to All Study Locations
17. Research activities in all study locations revealed a number of key common themes,
which condition household susceptibility to poverty. These are summarized in the following
sections.
1. Resource Endowments
18. Land and Labor. Across all communities, the ability to have rights over natural
resources was critical in determining whether or not a household is poor. Without any form of
land, it is clearly difficult for households to graduate from poverty, as they have no capital other
than their labor nor any skills or education. Without land, most of the productive time is spent in
wage labor to meet the subsistence needs of the household, and there are few opportunities to
accumulate savings as rates for wage labor are at subsistence levels throughout the study
locations. A lack of land also means that the poor and very poor are unlikely to have much to
trade or sell outside the community and are, therefore, less likely to use a road, no matter what
its condition. Among all survey respondents22 in all case study areas (project and control sites),
20
A state where households are forced to use what little assets they have to borrow in order to meet their immediate
consumption needs.
21
The better off in the study communities are also not poor in a national sense (Appendix 3, para. 4).
22
The terminology “all survey respondents” is used to reflect all project and control site groups in all three countries.
10
26% of respondents do not farm any land whatsoever, either as owners or tenants, and most of
those who do farm use extremely small plots of land that are barely sufficient to meet their
needs.23
19. Livestock and Forest Products. Animals are an important source of livelihood, as well
as a source of security during emergencies in all of the project locations. Ownership of a few
heads of small livestock is, therefore, a critical poverty indicator. For the very poor, forest
products are also key assets because the very poor usually lack land on which to grow cash
crops. These are gathered from vacant lands and forests and often require considerable time
and effort for collection.
2. Debt Cycles
20. The level of household indebtedness is also a key poverty indicator. Many poorer
farmers and wage laborers in the study communities are heavily reliant on credit extended by
agricultural intermediaries and moneylenders in the community. During emergencies, tenants
Box 1: Gradual Emergence from Poverty and poor farmers borrow money against
the next season’s crop, which enables
Mr. Gripon’s case study in Sorsogon illustrates that it takes the intermediaries to dictate the price at
a long-term secure livelihood opportunity and which this debt is relieved. Often, debt
commercialization of the economy for families to move out
of poverty. A rural road is a necessary condition for this
cannot be fully paid from one crop. In
process, but not a sufficient one (Appendix 8, Box 12). this way, tenants and poor farmers are
locked into relationships of debt. Those
dependent on wage labor also get credit from large landowners to overcome periodic food
shortfalls. This ensures a labor obligation at predetermined rates when landlords most need it
during peak harvest times. Formal credit sources have numerous requirements usually too
onerous for the poor to meet and are often located outside the village, making it difficult for the
poor and very poor to access.
3. Family and Community
21. Household Size and Composition. In all of the study locations, there appears to be a
high correlation between large family size and poverty. Whether having a large number of
children is a symptom or cause of poverty is a moot point, but is probably both. Children offer a
potential source of future livelihood security for poor families, but also require a high level of
care in terms of income, resources, and time. When the care burden is heavy, productive
members of the household are forced to stay at home to engage in care tasks, and the potential
labor base of the household is, therefore, reduced.
22. Social Networks and Length of Residence. Close relationships of family and kinship
groups are features of rural societies throughout Asia, and all of the study locations clearly show
how important, complex, and dense these networks can be. These connections offer security in
times of hardship and are an important social safety net for the poor, with those better off in the
group obliged by tradition to look after the more vulnerable.24 Length of residence enables
family groups to build up relationships of trust and mutuality with neighbors. Length of residence
across a number of generations, combined with available family labor, would appear to be a
successful remedy in enabling the family to move out of poverty. Using multidimensional
poverty, which captures access to social networks, becomes important in this context.
23
See Appendix 3, para. 13 for charts and a description of farming activities and size of landholdings.
24
In many of the study locations, extended families would live in a “compound” arrangement, with a variety of
standards of housing reflecting the relative affluence or impoverishment of different members of the extended
family group.
11
IV. CONSIDERATIONS FOR PROJECT DESIGN AND IMPLEMENTATION
A. Considerations for Project Design
23. Impacts of rural road investments do not occur in a vacuum. Geographic, climatic, social,
and economic conditions determine the context within which impacts take place. This chapter
discusses different conditions that are conducive or not conducive to poverty reduction through
rural road improvement. A careful consideration of these conditions could yield a checklist that
could be used during project identification and design to assess the potential for poverty
reduction through road development projects. The conditions that are not conducive can be
categorized into two groups: external conditions and structural conditions. The influences of
these two types of conditions on the case study areas are described below using site-specific
examples. Although the conditions described here are not applicable to all sites, by using these
specific examples the study tries to illustrate the key factors that may determine the type and
magnitude of impacts that improved rural roads may bring to the poor. Awareness of these
factors and periodic monitoring of a sample of project areas may enable the provision of other
supporting measures or complementary efforts to increase the beneficial impacts on the poor.
1. External Conditions that Cannot be Affected by Road Development
24. Climate and Natural Resources. All of the roads covered in the study serve agricultural
communities, and the local people rely overwhelmingly on agriculture for their livelihoods. The
potential to make use of the road to improve livelihoods is, therefore, heavily reliant on the
fertility of the land and the potential for greater production and diversification of agriculture. A
number of factors are important here, including access to land (discussed below) and the
availability of extension services, for example. But a core necessity is a climate and natural
resource base that has the potential to support increased and intensified agricultural activity.
25. The Kurunegala project site shows how villagers’ ability to make use of a better road is
conditioned by wider climatic conditions such as the 3-year drought that has crippled the
village’s rice production (Appendix 6, para. 7). This has affected both landowners and wage
laborers who work in the fields. In another project site in Yogyakarta, the project road passes
through an area where water retention is difficult, the soil is rocky, and it is time consuming to
prepare fields for cultivation. Therefore, crop diversification into higher value cash crops would
be difficult (Appendix 7, para. 11). Consequently, despite the incentives provided by better
roads, these climatic and natural resource conditions limit the ability of some areas to respond.
26. Remoteness and Access to Marketing Networks. Locational factors have an

important bearing on the potential for development. A comparison of project sites indicates that
there is a close correlation between the intensity and severity of poverty and the distance from
major population centers and marketing networks. In more remote locations (Sorsogon), it is
more difficult to attain a critical mass of demand and availability of transport services. In those
locations closer to major towns or centers (Matara), distances are shorter and, therefore, the
intensity of communications and of information exchange is much greater. Rural inhabitants
here are more likely to engage in alternatives to subsistence agriculture, at least seasonally,
and to be able to achieve better prices for the goods that they sell, where competition and
demand are higher. More remote areas, therefore, simply have a greater isolation barrier to
overcome before the anticipated economic benefits of improved rural roads can make
themselves felt.
27. Macroeconomic Context and Terms of Trade. Many of the study locations are heavily
reliant on a single cash crop. In Sorsogon, it is copra, in Negros sugarcane, in Matara tea, and
in Bengkulu rubber. Fluctuations in the world market prices of these commodities have an
12
enormous impact on the well-being of all socioeconomic groups in the study communities, as all
groups are tied into the commodity economy through sale or labor. Where crop prices are good
or increasing, significant benefits accrue from road improvement and resulting marketing
opportunities (tea in Matara [Appendix 6, para. 38]). But a slump in world market prices creates
a significant downturn in a village economy (copra in Sorsogon [Appendix 5, para. 9]). Then, the
poor and very poor, being risk averse, are more likely to concentrate on subsistence food
production rather than cash crops (rubber in Bengkulu [Appendix 7, para. 6]). The susceptibility
of poorer farmer groups to debt also means that they are forced to sell their crop immediately
after the harvest when prices are lowest; they cannot wait for prices to increase in the world
market.
28. The world market drives many commodity prices. Persistent difficulties experienced by
the study communities producing cash crops in the Philippines and Indonesia appear to support
a contention that developing countries are increasingly subject to fluctuating terms of trade in
agricultural products. This has led to what has been described as a process of “de-
agrarianization,” a fundamental change in the nature of rural economies in the face of a slump in
the terms of trade, with nonagrarian, nonfarm employment becoming increasingly important in
sustaining rural livelihoods.25 Some of these issues are returned to in the following sections with
a detailed discussion of the evidence from the case studies of increased outmigration (para. 70)
and employment. Mobility, particularly for the young, appears to be increasingly important in this
context.
2. Structural Conditions that Cannot be Affected by Road Development
29. Social Structure and Concentration of Assets. The prevailing social structure and
concentration of productive assets have an enormous bearing on determining how impacts
occur in each of the study locations. The concentration and distribution of land is particularly
important, and largely outside the area of influence of a road project.26 Nevertheless,
recognizing how assets are distributed is important in understanding how benefits will accrue,
and for planning, mitigation, and complementary measures to enable those who lack assets to
also benefit from any planned investment.
Box 2: The Very Poor Unable to Use the Better Road
30. In the Sorsogon case to Overcome Their Poverty Status
study area, land ownership
is heavily concentrated Mr. Grita’s family in Sorsogon is landless and lives on encroached land on
among a small number of the study road. He works about 4 days a month as a laborer when he can
find work. The family experiences acute food deficits. They worry about
owners (Appendix 5, para. sickness and damage to their small house. They recognize that more
9). Many are absentee transport services are available but only use them intermittently when they
owners residing outside the have money to buy essentials (Appendix 8, Box 1).
village, with longstanding Mr. Olarte’s family in Negros lives in a hut on the study road on a land
relationships with tenants owned by a plantation (hacienda). The family has work during the planting
and harvesting seasons and at other times experiences a food scarcity.
who farm the land. Often, They depend on remittances from two children who have migrated and
tenants are required to sell store credit during slack times. They usually walk to work and only use the
the crop (copra) to the road to buy provisions when they have money (Appendix 8, Box 2).
landlord, at a predetermined Mrs. Nandawatie and her family in Kurunegala are very poor and rely on
labor opportunities that her husband can find and the garden that she
price. In exchange, they cultivates. She sells the vegetables in the village and her husband uses
receive credit facilities the bicycle to search for work. They acknowledge that some goods are
throughout the year against cheaper since the road improvement. She uses the road to buy essentials
the next crop. Similarly in or to go to the hospital (Appendix 8, Box 4).
Bengkulu, intermediaries
25
Bryceson, D., C. Kay, and J. Mooij. 2000. Disappearing Peasantries? Rural Labor in Africa, Asia and Latin
America. IT Publications.
26
The study roads were mainly road improvement and, therefore, no land acquisition or resettlement was involved.
13
(tokeh) and traders effectively control market exchange in paddy and rubber with relationships
of debt. It is increasingly difficult to escape from these relationships as indebtedness rises
(Appendix 7, para. 17). In Negros, the concentration of land ownership is particularly heavy in
the haciendas27 where a majority of project site inhabitants live. Most villagers have no land at
all, and the residents are required to work in the hacienda when seasonal labor is needed
(Appendix 5, para. 39). The situation in Matara is slightly different because the level of land
ownership is less concentrated, but landholdings are small and the poor particularly are required
to work in the tea plantations throughout the year, limiting their opportunities for exploring more
lucrative livelihood alternatives (Appendix 6, para. 32). Many are, therefore, locked into these
long-term working relationships, which result from the prevailing norms in social relations,
economic exchange, and ownership of productive assets.
31. In these overwhelmingly rural, agricultural project settings, land ownership is clearly a
key parameter in determining if, and how, people can take advantage of opportunities that the
road may bring.
3. Conditions that Can be Affected by Road Development
32. The improvement of rural roads is broadly recognized as a fundamental precondition for
the development of rural areas, and remoteness and lack of mobility are widely identified by the
poor themselves as factors in
Box 3: The Very Poor Using the Road to Improve Their
Livelihood
heightening vulnerability and
perpetuating their poverty.
Mrs. Tumilah in Yogyakarta is a very poor widow with children. Investment in transport can create
She works 2–3 days a week as a laborer when she can find economic opportunities for the poor
work. She sometimes collects firewood that she can find from directly through employment in
vacant lands to sell. She has benefited somewhat from the road
as she does not find it difficult to sell the firewood because infrastructure construction and
more intermediaries come along the road to buy it. During road maintenance, and the provision of
rehabilitation, she was refused work. She uses the road to do rural transport services; and
her weekly marketing (Appendix 8, Box 3). indirectly through improving the
conditions and opportunities for
marketing goods and services, reducing input prices, opening opportunities in new markets, and
offering seasonal migration opportunities for work; it can also improve opportunities for
household travel for social purposes, such as visiting family or accessing health facilities.
33. Access to health facilities and other basic needs is critical for all. A rural road provides
the opportunity for the poor and very poor to access these services based on their needs and
gives government an opportunity to reach the poor. Although it may not be reflected in an
income sense, this impact reduces the vulnerability of the poor and can be reflected as a means
of multidimensional poverty reduction. Beyond this, basic access transport needs may be
different for different socioeconomic groups. The poor lack time and energy. Impacts, which
either reduce or exacerbate this time and energy deficiency, have a critical bearing on poverty.
Better-off groups who have time, energy, and capital may have a better opportunity to expand
their livelihood opportunities and welfare. Women and men may have different transport needs,
and an analysis of intra-household dynamics is also important in evaluating the impact of rural
road investments. Women are more likely to lack access to means of transport, or income for
travel, but have a heavy transport burden in meeting household productive, social, and
community obligations. They may also face significant nonphysical barriers to using some types
of transport (i.e., bicycle) or to travel alone outside the household or community.
27
Large estates or plantations.
14
34. During the project design, these different socioeconomic groups and possible outcomes,
as well as gender aspects, need to be recognized, with a particular focus on how they will
impact on the poor. The possible outcomes can be broadly classified according to direct
impacts, which can be expected to occur soon after road improvements, and other impacts that
may take considerable time. How these outcomes were different or similar in the various project
sites and control sites, as well as among different socioeconomic groups, is discussed in
chapter V under Key Impacts and Issues and, therefore, the outcomes are only listed here:
(i) changes in transportation services: transport modes, volume, cost, and

competition;
(ii) changes in travel patterns: number of trips, purposes, and distances;
(iii) changes in village profile: agriculture commercialization; social services like
health, education, information, credit and extension; and migration and
remittances; and
(iv) changes in income and welfare: food security, household income sources, and
welfare.
B. Considerations for Project Implementation
35. This section discusses implementation issues relevant for road development and how
they relate to poverty reduction. The issues range from labor-based construction and
maintenance options, to responsibilities for road sustainability and community participation.
1. Labor-Based Techniques and Poverty Reduction Option
36. One aspect of rural road development, which benefits the poor directly, is the opportunity
that roads offer for employment in construction when labor-based methods are used. Of the six
projects evaluated under this study, only one (Bengkulu) had some limited form of labor-based
road improvement component. Consequently, the level of employment of local people on the
projects was very low. Where limited opportunities did arise, it was seldom the poorest that were
explicitly selected for the works. Women were often excluded, even though female single-
headed households were among the most vulnerable in all of the study locations. Of the
2,271 people covered under the survey, 97% were not employed on the road in any way and
2% were engaged for about a month each. Most of those who were employed were from the
Bengkulu and Yogyakarta project sites. Twice as many men as women were employed, and
women cited that even when the opportunity to work on the road did exist, they were often
excluded in favor of men. About 8% of all household respondents knew of work opportunities on
the road, and of these, about one third were refused a job by implementation staff. Of those
employed, a majority were casual laborers during rehabilitation work. Because of the short
nature of the work opportunity, the wages were primarily used to meet household food deficits
(70% of responses). Others paid off debts or bought household items.
37. Experience from Asia and Africa over the past 40 years shows that a principal means for
the poor to gain direct benefit from rural road programs is through their paid employment in
carrying out physical works.28 Although employment on road reconstruction is temporary, the
short-term injection of cash can often provide the necessary start-up capital for the poor to
diversify livelihoods. Labor-based maintenance also offers an opportunity for the poor to earn
money on a regular basis through the road. In developing Asian countries where
underemployment is a major problem in rural areas and the state suffers from a lack of financial
capital, it may be more cost effective to use labor-intensive construction methods. Authorities in
28
Loan 1385-CAM: Rural Infrastructure Improvement Project (footnote 1) is successfully using a labor-based
construction method.
15
the study countries, when questioned on this point, indicated that labor-based construction
methods are slow and that the quality of works can be variable. But for low-volume rural roads,
the time and efficiency argument in construction or rehabilitation is surely not so important, as
the opportunity cost of disruption to traffic is very low. Arguments about the poor quality of labor-
based works are refuted by experience from elsewhere in Asia, particularly Cambodia and the
People’s Republic of China, which have traditions of labor-based works and where studies have
shown that the quality of works can be at least as good as that of contractors using machinery.29
A strategy using paid labor for most tasks, and machinery for compacting, might be a suitable
compromise for gravel road reconstruction in rural locations.
2. Governance Issues Surrounding Rural Roads
38. Roads as Political Capital. Roads play a critical role as political capital in all of the
study locations. Decisions about where roads should be built, which roads should be upgraded,
and which roads should be maintained are heavily influenced by the prevailing political context.
This is not a recent phenomenon, as the example of the Negros study location shows. Negros
has always been a key sugar producing area of the Philippines and many of the decisions
regarding its development have been made with sugar production, the plantation owners’
interests, and the interests of the industry in mind. In the Matara and Bengkulu study areas, the
villagers lobbied authorities and politicians for the inclusion of their roads in road rehabilitation
schemes. The ability to do this successfully appears to depend heavily on the level of political
influence communities can exert. As a result, decisions about road rehabilitation, maintenance,
and the prioritization of work are often far from transparent. Poorer areas are likely to suffer
under this regime, as they are least likely to have the connections and authority to lobby
effectively for better roads.
39. Neglect of Maintenance. Regular maintenance of rural roads is a critical precondition

for sustaining the positive impacts that roads bring to rural communities. Minor maintenance is
often neglected because of lack of funds,30 but it is also neglected because there is little political
capital or mileage in maintaining roads regularly as the results of minor maintenance are not
highly visible. Instead, politicians prefer to authorize major rehabilitation or reconstruction to
take place after the road has deteriorated very much. The promise of improved roads from
politicians is often sufficient to ensure their election, and, if works do take place, politicians are
quick to claim credit. This practice prevails in all of the study countries to a greater or lesser
extent. And in this context, establishing regular and transparent maintenance regimes and
criteria for rehabilitation is very difficult. Roads consequently get reconstructed, are left to
deteriorate, and then are reconstructed again in 10 or so year cycles, so that villagers
experience peaks and troughs of accessibility, rather than having a constant and guaranteed
level of access.
40. Institutional Responsibility for Roads. Problems of maintenance, which stem from
scarcity of funds, are exacerbated when the institutional responsibility for rural roads is not clear
or where the budget source for maintenance has not been properly established. Often, there
appeared to be a lack of clarity (in practice, if not in theory) over who was responsible for
maintaining the project roads and where the funds would come from. This was particularly the
case with the integrated projects, which were rehabilitating gravel roads. The asphalt roads
rehabilitated under transport sector projects were usually the responsibility of the public works
department or department of roads/highways. With the integrated projects, roads were often
only a subcomponent of the works, and the executing agency for the project was the agriculture
29
It may be useful to increase the awareness of success of labor-based road construction and maintenance through
workshops or study tours.
30
See Operations Evaluation Department. ADB. 1998. Special Evaluation Study on the Operation and Maintenance
of Road Facilities and their Impact on Project Sustainability. Manila.
16
or irrigation department, with no direct funding line for road maintenance. Also, problems of
unclear institutional responsibility are compounded when there are frequent shifts in personnel
at responsible implementing agencies due to the high level of politicization of the bureaucracy.
Institutional continuity and ownership of the roads suffer as a consequence, and roads fall into
disrepair.
3. Community Participation as a Key to Long-Term Sustainability
41. Full community participation and management are increasingly recognized as important
for the long-term sustainability of infrastructure investments. However, when community
management means cash or labor contributions, the burden can fall disproportionately on the
poor. They may be forced to contribute free labor time which leaves them less time to engage in
their primary productive tasks, and their subsistence needs and food security suffer as a result.
They may also be forced to contribute toward the provision or maintenance of infrastructure,
which they hardly use, or are unable to make use of. Community involvement should, therefore,
mean more than labor contributions for the poor. They should be involved in the assessment of
needs31 and decisions regarding maintenance and management.32 This will ensure that
community participation is genuinely pro-poor and inclusive, and will contribute to the long-term
sustainability of investments, as the poor too will have a stake in using and maintaining
appropriate roads, which serve their needs.
42. Villagers throughout the control areas in the study recognize the importance of roads in
bringing benefits to their communities. In the Yogyakarta control site, the villagers have
maintained and managed resources for their own intra-village roads for the past 30 years
(Appendix 7, para. 14). In the Bengkulu project site, the project road was originally constructed
through the active involvement of community members, and the farmers’ association maintained
it for a few years and later neglected it, as the villagers felt that they were receiving no benefit
from it (Appendix 4, para. 9). The road has consequently reverted to a footpath/track that still
appears to meet their needs. Similarly, in Sorsogon, the level of access on a road (nonproject
road) was more than required. The villagers felt that they were neglecting other livelihood
activities by maintaining the road regularly and stopped maintenance (Appendix 5, para. 9).
Among the survey respondents who answered, 61% said that they would be prepared to
contribute labor for maintenance of rural roads as against 27% who said that they would not be
prepared to contribute. These examples confirm that the level of access required by the poor is
basic, but that where this is lacking the potential to engage the poor themselves in the works
and maintenance is high.
V. KEY IMPACTS AND RELATED ISSUES
A. Changes in Transportation Services
43. This chapter describes key impacts that can be expected soon after road improvement
and other impacts that may take time. The evidence from case study sites on each type of
outcome gives important clues on how the project interventions can influence poverty-related
conditions that can be affected by road development alone, as well as on other conditions that
need additional measures to reduce poverty.
31
The needs of the poor need not be confined to roads and can include tracks, culverts, crossings, and
improvements to intermediary modes of transport.
32
Decisions regarding engineering specifications will have to be sought from technical experts in any case.
17
1. Transport Modes, Volume, Cost, and Competition33
44. A widespread traditional assumption governing the development of rural roads is that
investment in roads will spontaneously lead to private sector transport service provision and
this, in turn, will lead fairly quickly to provision of cheaper and better transport services to
everybody through competition. The case studies suggest that in all project sites, a variety of
transport modes have emerged and travel time has substantially fallen. However, Table 2
shows that increases in transport volume and decreases in fares have occurred only when there
is competition among transport providers. Competition is clearly the critical precondition for the
development of better transport services. But the relationship between rural road investments,
transport service development, and competition is not straightforward and an automatic link
cannot be assumed. There are intervening factors and contextual barriers to this. Rural roads,
by definition, often serve remote areas, where distances to more urbanized centers may be
long, population density low, and market information and demand imperfect. Both the
development of available transport and lower transport fares rely heavily on the competitiveness
of prevailing transport markets. In the absence of competition, there is no incentive for transport
providers to improve their service or pass on their cost savings to consumers.
2. Preconditions Necessary for the Emergence of Competition
45. All of the case studies witnessed an upturn in the number of vehicles using the road.
There was also an increase in the availability of transport services. However, it is not clear
whether there has been a reduction in transport costs to consumers. Rehabilitated roads
significantly reduced the VOCs of existing transport service providers, largely through a
reduction in maintenance costs. Across all case study roads, operators now spend less money
on maintenance of their vehicles, and are losing fewer days a month when their vehicles are
being repaired. Whether these benefits are passed on to service users depends heavily on the
level of competition that develops along the roads. For competition to emerge, certain
preconditions must be present. This section discusses these preconditions.
46. Demand. Potential demand must be there from areas served by the rural road, both for
commercial and personal travel. Potential demand is closely linked to population density and the
agricultural potential of the area. Little competition among transport providers has emerged in
the Kurunegala project site due to lack of demand caused by low production which, in turn,
stems from a severe drought that has lasted 3 years. In contrast, in the other project site,
Matara, competition is emerging rapidly, fueled by the better demand for green-leaf tea
(Appendix 6).
47. Distance to Markets. In the case of the Yogyakarta project road, which is relatively
close to market centers, the benefits are clear. The improved road surface has reduced the
VOCs for vehicle providers, more providers have been attracted to the route, and more
competition and a variety of available transport services have resulted. Transport operators
have had to reduce fares and no longer run full vehicles, but spend much less on vehicle
maintenance, and are able to make more trips a day because of the better surface, thus earn a
higher income. In contrast, the longer distances to markets and networks have kept competition
weak from Bengkulu. Although the availability and frequency of transport services have
increased in both the Bengkulu and Yogyakarta project sites, the transport prices have not fallen
for the district asphalt road in Bengkulu.
48. Maintenance. Evidence from the control area in Yogyakarta also suggests the
importance of road maintenance. It shows that high demand for services driven by the
33
More information on these aspects is provided in Appendixes 5–7 on case study details in each country.
18
agricultural potential of an area is not sufficient for competition to emerge; transport providers
gravitate toward areas with better roads. The control area in Yogyakarta has agricultural
marketing potential, but the road surface is very bad and, therefore, lacks transport. In such
areas, operators and traders can largely dictate the transport or crop prices, as farmers have no
option of readily selling in a wider market outside the community.
49. Regulatory Barriers. The Kurunegala road shows another important factor. The existing
regulatory environment for transport services in Sri Lanka makes it more difficult to operate a
pick-up truck or van as a person carrier along the road. There is also a long tradition of provision
of subsidized bus services by the state, making it more difficult for private bus carriers to
compete. It is unlikely that the Kurunegala road would be an attractive routing for private
operators, as the existing subsidized state bus service more than meets the apparent demand
for transport. There does not appear to be the critical mass of repressed demand and available
services for a major shift in service provision to take place. In the case of the Sorsogon project
road, there may be other types of barriers. In the Sorsogon study area, there is no automatic
direct link between improved roads, more transport services, and lower fares. The number of 3-
wheelers plying the route from the study community to the nearby markets has increased
somewhat since road rehabilitation. Many villagers now prefer 3-wheelers to jeepneys34 which
take longer to fill up. But the fare for 3-wheelers has not decreased markedly since road
rehabilitation. The demand has, therefore, not been sufficient to force down transport prices or
overcome price-fixing arrangements between operators.
50. Competition, and thus benefits for consumers (poor and nonpoor), appears to occur only
once a threshold has been reached.35 There must be a commercial center of sufficient size
within easy reach of the rural area, and there must be a commitment to regular road
maintenance. Without maintenance, roads will quickly revert to their previous condition and
transport entrepreneurs will have no incentive to provide services. Regulatory barriers also play
a key role in determining competitive mechanisms. Informal barriers, such as price-fixing
monopolies and “mafia” restrictions on the open entry of operators to the market, can be
significant barriers. Government actions too can repress competition. The creation of subsidized
services in areas with high demand acts as a deterrent to private operators who cannot
compete with subsidized prices. Case study experiences suggest that governments have an
important role to play in creating a suitable competitive environment for transport provision in
rural areas. In some places, existing demand is low and monopoly practices are simply too
prevalent to be spontaneously overcome. Here, governments can play a role in stimulating rural
areas and in regulating transport markets effectively to ensure both that benefits from rural
roads accrue more equitably and, particularly, that the poor have access to cheaper, more
reliable, and more frequent transport services.
3. Who Provides Transport Services?
51. Having established the conditions for competition to emerge, the transport service
providers and their socioeconomic status are discussed in this section, considering the
relationship between improved roads, better services, lower prices, and the impacts on different
groups within and outside the community. A critical question is whether the poor and very poor
are able to make use of the better availability of transport services, and whether transport prices
have fallen sufficiently to make the services affordable to them. Do the current livelihood
strategies of the poor require them to access transport services, and if not, what barriers do they
face in exploring new opportunities through wider and more frequent travel?
34
Jeepneys are locally-made vehicles which are like minibuses (Appendix 5).
35
See summary traffic data in Appendix 1, Table A1.1.
19
52. Transport Providers. Vehicle operators are often winners with both good and bad rural
roads: (i) they can monopolize transport routes along poor roads where competition is weak and
dictate prices and the level of
Box 4: Owner/Operators—Transport Providers
service; (ii) they also benefit
from better roads through Mr. Supardi has been a vehicle operator on the district study road in
lower VOCs and reduced Bengkulu since 1993. Since road rehabilitation, he has saved
considerably on vehicle operating costs and time spent on the job. His
maintenance; and (iii) they can income has increased by 300% (Appendix 8, Box 14).
save time on better roads Mr. Gnanatilake has been the sole truck operator on the Matara study
leaving them more opportunity road since 1994. Once, he spent his savings to repair sections of the
to engage in other activities. road. Since the 2001 study road rehabilitation, he has invested in a
minibus and has improved his income (Appendix 8, Box 15).
They appeared to be the Mr. Lalith is from a better-off family in the Kurunegala project site.
primary beneficiaries from road Faced with a water crisis, he gave up paddy farming to provide
improvements in almost all of transport services and used his savings to buy two vehicles that ply
the case study areas. In most the study road. His income improved (Appendix 8, Box 16).
sites, there are a few large
vehicle operators on improved roads, but they are based in larger towns at either end of the
road. Operators of smaller motorized vehicles, however, particularly 3-wheelers, are often
based in the study communities. Owning and operating a 3-wheeler is a significant income-
generating activity for better-off members who have the initial capital and sense of security to
invest in a vehicle.
53. Traders and Intermediaries. Traders and intermediaries are a feature of all of the
agricultural areas covered by the study. They were not typically resident in the study villages,
but were usually based in nearby town centers. They are also often transporters in their own
Box 5: Traders and Intermediaries
right, owning and operating a vehicle to
conduct their business. Like transport
Mr. Anwar is an intermediary in Bengkulu who depends providers, traders benefit both from poor
on large outside traders for capital. With the road
rehabilitation, he increased the volume of trading and his and better rural roads. When the road is
income increased, but he still finds it difficult to get bad, they can often establish a monopoly
capital at reasonable rates (Appendix 8, Box 18). position in buying primary products at the
Mr. Niran had left the study village in Bengkulu in 1993 prices they dictate, as theirs is often the
but came back in 1996 after road rehabilitation. He
rented a truck and became a trader and has now bought
only truck visiting the village. With
two vehicles and a rice miller for his growing business improved roads, they can transport more
(Appendix 8, Box 19). produce and increase their income
although they may be subject to a more
competitive environment with more traders attracted to using the road to buy products from the
agricultural hinterland. Farmers themselves can also take advantage of better road access,
faster travel times, and more available transport services to explore wider markets in urbanized
areas, where the demand for commodities is likely to be higher, and so the price. However,
evidence from the Indonesian sites suggests that, despite the improved transportation links,
traders and agricultural intermediaries are able to maintain their monopoly position in the
commodities trade because farmers are locked into relationships of debt with them (para. 20
and Appendix 7). Box 6: A Very Poor Family of Porters
Mrs. Kustina and her family live in the Bengkulu project
54. Headloading. Shifting patterns site. They are very poor and landless, and work as porters.
in transportation services may affect the With the plantation road improvements, their portering has
poor either beneficially or adversely become easier, more income generating, and allows them
depending on the contextual situation. more time for other activities. They do not prefer the road
The example in Box 6 shows how reliant to be improved further as that would take away their main
source of livelihood (Appendix 8, Box 5).
the poor are on intra-village travel over
external travel, and nonmotorized over motorized transport means. In Bengkulu prior to the
project, poor households were engaged in porterage to bring in farm produce. The headloading
was extremely hard. The work was exhausting and did not allow porters to carry out other
20
livelihood activities. Since the establishment of farm roads, however, the work has become
easier with the ability to use handcarts. Very poor porters have increased their income and also
engage in other activities such as basket weaving. The volume of produce moved has increased
mainly because of the road. However, the poor in the village are ambivalent about improving the
road further as it would enable vehicles to use it, thereby competing with their portering. These
thoughts are echoed in control areas throughout the study countries where porterage of produce
is an important livelihood opportunity for the poor. With better roads and access come vehicles,
which can transport more produce. The poor are vulnerable to loss of this livelihood activity and
are unclear what alternative benefits would come to them from the roads (Appendix 7).
B. Changes in Travel Patterns
1. Modes of Transport of the Poor
55. Despite the difference between the case study sites, the travel patterns of the poor and
very poor across the study areas appear to be remarkably similar. Most appear to mainly restrict
their travel to the village area and occasionally travel outside the village. Even when the village
is close to a nearby marketing center (as in the case of Negros), the poor travel only on market
days and not regularly. The fundamental reason for this appears to be lack of capital; they have
little to sell and little money to buy anything but essential items. They also lack time as their
days are taken up with subsistence tasks and laboring, and have little time for speculative
activities to diversify their livelihoods. The primary mode of travel is walking, though the poor
also use bicycles when they can gain access to them. Poverty is compounded for those in
control areas where remoteness and isolation are greater.
56. Access to means of transport among those who are in the better-off category of survey
respondents is, unsurprisingly, far better than among those who are poor or very poor. About
65% of the better off have a bicycle, against 35% of the poor and very poor. Twenty-five percent
of better-off households own one or more scooters/motorcycles, against 4% of the very poor
and 9% of the poor. Comparisons between the very poor, poor, and better off indicate that for
crop processing, the percentages of each group traveling are 9%, 17%, and 24%, respectively.
To sell crops, the percentages traveling are 20%, 16%, and 32%, respectively. In traveling for
business or employment, the better off travel more often, generally more than once a week (i.e.,
for regular employment). In fact, PRAs in all of the study communities showed that those with a
regular government or salaried position were by far the largest subgroup of the better off in all
locations.
57. The responses to the question of whether the condition of the road had changed after
rehabilitation were generally very positive (Table 4). Among all project respondents, 59% said
that it was better than 5 years ago, although 24% disagreed. However, consciousness of the
road and its importance seems to be much higher among the better off than the very poor and
poor. Table 4 also shows that the better off are less likely to give a “don’t know” response, and
more likely to give a positive response.
Table 4: Change in the Condition of the Road Over 5 Years (%)

Response Socioeconomic Group Total
Better Off Very Poor Poor
Don’t know 3 9 9 8
Better now 71 53 58 59
Worse now 21 23 26 24
No change 5 15 7 9
100 100 100 100
Source: Household survey data.
21
58. In looking at access to modes of transport, it is useful to consider the buying of

provisions, which is a common Figure 2: Means of Transport Used for
task. Figure 2 shows the Buying Provisions (project sites)
y g (p j )
proportionate modes of transport other
100
for each group in the project site. 9 19
bus
90 16
The very poor are much more 18
reliant on walking than the better 80
10
minibus/route van
33
off. The latter are more likely to
percentage of respondents
70 truck/lorry
23
have access to private motorized 60
34
means (motorcycle or 3-wheeler) car/van
50
or to a car or van. Interestingly, 3-wheeler
the poor are more likely to use a 40 16
41
bicycle, while bicycle usage 30
motorcycle
9
among the very poor is 20 bicycle
23
negligible. The very poor’s heavy 18
10
reliance on walking is reflected in animal drawn/cart
0
other tasks too, such as in on foot
better off very poor poor
accessing health services, going
to school, and selling products.
2. The Importance of Roads to the Poor
59. Through the improvement of road surface and provision of all-year access, the six
projects studied have allowed better modes of transportation, reduced travel time, and at times,
even the travel costs, thereby bringing benefits to those who travel outside the community and
its vicinity. PRAs indicate that the poor and very poor inhabit a localized village, walking world,
and as such make little use of medium- or long-distance transportation links. Of more
importance to them are the network of paths, tracks, culverts, and access routes in the
immediate village vicinity, on which they rely to access water, firewood, fields, and local
employment opportunities. Saving time in their within-community travel is important to them.
Intermediate modes of transport that help them increase their carrying capacity are also useful
to save time for more productive work. Most things critical to their lives can usually be found
within the village locale, and travel outside is occasional and for a special purpose. Incremental
benefits to them are more likely to come from accessing nonmotorized transport and ability to
cross waterways, etc., to help in their daily routine tasks. Often, they cannot afford to use
motorized vehicles, and these vehicles travel to destinations beyond their sphere of livelihood.
Therefore, increasing the mobility within the village is as important for poverty reduction as
providing access to markets outside the village. The time savings within the village will allow the
poor and very poor to be more productive and generate small savings to explore opportunities
outside the village.
60. Men and women in all of the study locations have different household responsibilities. As
such, they also had different transport and travel patterns and needs. Responses from PRAs
confirm that women across the study countries are deemed to be responsible primarily for
household tasks, and are likely to have to spend more time on these tasks. In both their travel
within the community and outside, men and women have different travel patterns, tasks, and
responsibilities. In local travel, responses indicate that both men and women share
responsibility to undertake crop production. For water collection, men overall have a slightly
higher responsibility (30%) than women (26%). In collecting firewood, another major household
task requiring travel within the immediate village area, men and women have similar
responsibilities. However, in looking at the total time spent to undertake this task, women spend
nearly twice as long in firewood collection tasks as men (42% to 22%).
22
61. Inside the community, the survey shows that women are much more likely to travel for
health purposes (55% as opposed to 5% for men), either for themselves or, more frequently, to
accompany children. They are also much more likely to travel for provisions within the
community, with 46% of responses against 17% for men, and 18% shared by both. Men are
more likely to travel for crop processing (53% to women’s 14%, with 17% shared) and social
travel within the village is largely shared. In travel outside the community, these patterns are
broadly replicated. Outside the village as well, women are more responsible for buying
provisions and traveling for health reasons, while men travel outside for employment, crop
processing, and selling their produce. Survey returns also show that, for various tasks, men are
broadly more likely to have access to private means of transport like a bicycle, 3-wheeler, or
motorcycle. Women are more likely to travel on foot to fulfill tasks, or use a public form of
transport, like a bus or truck. The opportunities for men to travel outside the village and to take
up outside opportunities are reinforced and perpetuated by traditional gender roles in the study
sites, with women responsible for household tasks and men for productive or economic tasks.
C. Changes in Village Profile
1. Community Well-Being and Development
62. The study looked both at project and control sites to understand what the impact has
been of road rehabilitation on the general socioeconomic well-being and development of the
respective communities. In general terms, the benefits of better roads (to all socioeconomic
groups as a whole) are highly evident when project villages are compared to control villages.36
Average travel time taken is often half or less for project households than control households for
all types of activities. For a variety of tasks, project households in the survey were more likely to
travel on a weekly basis, and control households on a monthly basis. Due to difficulties of
access, control site households often have to wait and combine a number of important tasks
into one trip to avoid spending long periods of travel for one purpose only. In response to the
questions on what the primary purposes of travel are and how often they travel outside their
village, respondents in project sites and control sites had different priorities (Table 5). Control
households travel more frequently for crop processing and for selling their produce than project
households. This suggests that (i) primary agricultural activities are more important in the
control areas, which may lack alternative livelihood opportunities; and (ii) due to better access,
many of the services that come directly to the project site are not available in the control site.
Evidence from PRAs in all locations suggests that, as a broad trend across all three countries,
both of these factors are important. Project sites have a wider variety of services (e.g., crop
processing) available within the village, and are also accessed more regularly by buyers,
meaning that primary producers are less likely to have to travel outside the community to sell
their produce.
36
The selected project and control areas are similar in their socioeconomic characteristics except for the road and, in
broad terms, the level of development in the study and control communities prior to the project was similar
(Appendix 1).
23
Table 5: Travel Patterns and Time Outside the Village for Project and Control Sites
Activity Project Site Control Site

Number of Average Number of Average
Trips per Travel Time Trips per Travel Time
Month (minutes) Month (minutes)
Buying Provisions 2 20 1 30
Selling Crops/Products 1 20 2 60
Going to School 2 30 1 50
Processing Agriculture Products 2 30 2 60
Visiting Family and Friends 1 90 1 180
Accessing Medical Care 1 30 1 90
Going for Employment or Business 1 60 1 240
Obtaining Official Documentation 1 30 0.3 110
Other 1 90 0.6 360
Note: Person trips are number of persons times the number of (two-way) trips made.
63. In the project site, livelihood opportunities outside the village are better as a result of the
road. Twice as many study community members have outside employment, business interests,
or a secondary livelihood income stream than control community members. Project households
are more likely to travel to visit friends and family or to fulfill state documentation or
administrative tasks than control households. In terms of the time spent by households in
traveling, the project communities demonstrate considerable savings over control communities.
64. Evidence from the survey also shows that control community households are much more
likely to travel on foot for all tasks. Project households are much more likely to use either a
bicycle, motorcycle, or 3-wheeler. In general then, project households have the opportunity to
travel more for all tasks, have greater access to and make greater use of transport services, use
private modes of transport like bicycles and motorcycles more, and have greater choice in
deciding whether or not to travel for tasks like crop processing and selling as they have a
greater variety of services available in the community. Project communities also appear to have
better access to safe sources of drinking water and to have better sanitary and toilet conditions
(Table 6). This may be a function of the general increased level of development of project over
control sites (itself a function of better access to roads, communications, and opportunities). It
also reflects the better accessibility of state services and nongovernment organizations (NGOs)
to communities; roadside communities are more likely to have services provided under these
schemes. Overall, 19% of all respondents have functioning electricity. The proportion in project
locations is far higher (28%) than in control areas (11%). This too is a direct consequence of the
better location and accessibility of project areas being beside the road. The level of general
education of project households is better than that of control households. Average years of
schooling for the head of household and spouse were 6.4 years in project sites compared to
5.2 in control sites. Roads are clearly a critical enabling condition for development of living
conditions in rural areas. However, the distribution of the benefits of these roads within
communities is a separate issue, and there are no guarantees or inherent mechanisms to
ensure that benefits will be distributed equitably between the poor and nonpoor in communities.
24
Table 6: Access to Water, Sanitation, and Electricity

Item Share of Households (%)
Project Site Control Site
Drinking Water
Protected Well or Tap 37 35
Private Well or Tap 28 14
Piped to House 16 0
Unprotected Well 11 24
Private Rainwater Tank 4 17
Other 4 10
Sanitation
Private Flush 5 7
Private Latrine 53 31
Public Latrine 4 1
Open Pit 20 22
Field 15 29
Other 3 10
Electricity
Yes 28 11
No 72 89
2. Community Commercialization and the Flow of Goods and Services
65. Chapter III shows that there are gradations of the poor from the completely landless, to
farmers on small plots producing for subsistence and others producing sufficient surplus to sell
at certain times of the year. The majority of those classified as very poor or poor are engaged
both in smallholder farming and wage labor. They too are, therefore, heavily reliant on the
commercialization of agriculture that the improved road may bring. Although average
landholdings are small and the potential for taking over more land is slim, roads can potentially
bring new demand for commercial cash crops and a resulting increase in price. In addition,
extension authorities encourage crop diversifications (pili in Sorsogon) and intensification
(rubber in Bengkulu). A majority of all the respondents (61%) who farm land say that they
produce the same type of crops as 5 years ago. However, there is some difference between
project sites (55% no change) and control sites (67% no change).
66. Figures for the sale of agricultural produce are broadly similar between the project and
control sites, with 53% of project and 55% of control respondents selling some of their crop.
Responding to whether they sell more quantity now than 5 years ago, 67% answered that they
do not. Among those who do sell more, about 8% attribute this to better production techniques,
8% to more buyers visiting the community, and 5% to better opportunities to visit outside
markets. Among project respondents only, the percentage who are selling more and who
attribute this to better outside marketing opportunities is higher, at 12%. Comparing
socioeconomic groups, 71% of the very poor and 70% of the poor say they do not sell more
quantity now than before road rehabilitation; only 48% of the better off make the same
response. This suggests again that the opportunities for commercialization are much greater for
the better off than for the poor and very poor.
67. Improved roads clearly improve general opportunities and the environment for buying
and selling, particularly to visiting buyers. In the project sites, 54% of households say that more
buyers visit the community now than 5 years ago compared to 36% in the control households.
With respect to transporting produce for sale, the poor and very poor are still much more likely
to lack access to transport services than the better off. Among all respondents, 71% of the very
poor say that their primary transport means for crop sale is on foot, against 63% of the poor and
25
46% of the better off. The importance of better roads to crop commercialization generally can be
seen from Table 7. For control communities, the primary condition necessary for the better
marketing of crops is seen to be better roads as well as better transport services. Project
households feel that better prices and more buyers in the community are important, but
otherwise would appear to feel that the conditions for selling their crops are broadly present.
Table 7: Conditions Necessary for Better Selling of Agricultural Produce (%)

Condition Project Site Control Site Total
Not Selling 39 27 33
More Buyers in the Community 15 11 13
Better Prices for Goods 15 9 13
Better Roads to Travel to Market 10 28 19
Better Transport Services to Market Centers 5 8 6
More Frequent and Reliable Transport 2 3 2
Cheaper Transport 2 0 1
Other 12 14 13
100 100 100
68. Where there is no state or project support for creating the infrastructure and enabling
environment for diversification of production to take place, the degree of integration of the local
area into the cash crop economy is crucial. The comparison of the Kurunegala and Matara
project sites in Sri Lanka (Appendix 5) demonstrates clearly again the importance of the
prevailing context and traditions in determining peoples’ response to the potential opportunities
that better roads may bring. In the Sorsogon survey, respondents said that there were no
discernible differences in the pattern of sale and commercialization of the crop compared to
before road rehabilitation and no cash crop alternatives have developed. This is probably a
result of the structure of land ownership in the area, with most farmers operating as smallholder
tenants, and often committed to selling their crop back to their landlords as part of the tenancy
agreement. Collection and sale of forest products is an important livelihood resource that has
developed in the Indonesia and Philippines case study areas. The Yogyakarta and Matara study
areas have witnessed an increase in seasonal transit markets along the road for the sale of
crops. These transit markets act as collection points for visiting intermediaries to buy produce at
peak agricultural times and suggest increased agricultural intermediary activity in the study area
following road rehabilitation. However, many poorer farmers in the area are chronically indebted
to these same intermediaries, and so have little scope for maneuver in deciding whom they sell
to and at what price. Better access does not necessarily lead to better prices for poor farmers.
69. Small Business Development. Road investments have had significant indirect impacts
on the general level of economic development in each of the study locations. This is clear from
the development of small businesses
Box 7: Small Business Development
in the project communities. Improved
Mr. Piyadasa’s family in Kurunegala has given up farming roads and the better ability to transport
(despite owning land) to start a confectionary business which goods provide opportunities for those
has benefited substantially from the road improvement
(Appendix 8, Box 7). who can afford the investment to start
Mrs. Sukarta in Yogyakarta does a small part-time business a small store in the village or buy
dressing brides because she now saves travel time in her village produce or make their own and
regular teaching job (Appendix 8, Box 9). sell it in the nearby market centers.
Mr. Nimal in Matara has moved his tea plant nursery near the
road and has managed to establish a thriving small business
They also save people time in their
(Appendix 8, Box 13). previous occupations, allowing those
who have the skills and/or savings to
invest in other small businesses. Among project case respondents, 64% observed that the
number of small businesses in the community had increased since the road was built or
26
rehabilitated. Of those who had a business prior to the rehabilitation of the road, 55% felt that
the project had a positive effect on these enterprises. Of the 17% of project respondents who
had started a business since road rehabilitation, 69% said that the road was a factor in their
deciding to start the business. Of these businesses, the majority in all locations were small
provision stores supplying the local community. There is little difference between those in
project and control locations who are planning a small business (about 51% for each); but
among social groups, the better off are much more likely to have such a plan (63%), against
52% of the poor, and 43% of the very poor.
3. Migration and Remittances
70. Better roads offer an expanded scope of opportunity outside the village. But there
appears to be no significant link in the study areas between migration for employment and
poverty status—if employment is short term or seasonal and badly paid, then it is unlikely to be
secure and regular enough to replace existing income and subsistence activities. Seeking
outside employment opportunities, particularly in urban areas, can be a high-risk strategy for the
poor and very poor. A traditional assumption on the mobility of the poor is that their lack of
agricultural assets, particularly land, makes them more likely to seek employment outside the
community. However, labor markets in remote rural areas are imperfect, and accessing
opportunities is difficult, particularly where there is a lack of information. This lack of information
and inability to command rights over work opportunities are themselves a function of poverty.
Better-off households are much more likely to have access to information on well-paid, or
stable, outside employment, with the poor and very poor accessing only temporary, seasonal,
and unskilled work opportunities, which are usually poorly paid (Appendix 7). Lack of education
and lack of support networks in more urban areas compound this trend.
71. Evidence from the study communities suggests that the more severe the degree of
poverty, the less likely household members are to travel outside the community. The exception
to this appears to be the Sri Lanka case studies. Here, there appears to be a higher level of
mobility among all socioeconomic groups. Sri Lanka’s small size, high density of population and
settlement, and high levels of literacy and human development may account for this. In the other
case study areas, the mobility of the poor and very poor stems largely from the search for
seasonal wage-labor opportunities that are close to existing locations and usually for a short
period (up to 3 months) only. Among all survey households that provided valid responses,
68% do not work away from the community at any time of the year. Analysis of the results from
the remaining 32% shows that there are clearly two kinds of migration taking place: (i) seasonal,
relatively short-term migration to other rural areas, usually in the same district, in search of
agricultural wage-labor opportunities; and (ii) relatively longer-term migration to urban areas, in
search of nonagricultural employment.
72. In looking at the activities in which the migrant laborers are engaged when working
away, it is clear that project households have access to better opportunities for employment
outside agriculture. More are employed either in regular jobs or in other activities such as urban,
relatively longer-term, employment. Among all survey responses, temporary and seasonal work
accounts for 62% of all outside employment and activities. In terms of the time spent away from
home, 25% of all outside activity is weekly, i.e., for regular employment that may require staying
in town during the week and returning to the village at the weekend. A further 20% is accounted
for by temporary employment of a month or less. Information collected during the PRAs
indicated, however, that employment, as an alternative activity to farming or wage laboring, is
not a guaranteed means of accumulating income. Alternative employment strategies need to be
sufficiently secure and well paid to replace the income and security of the activity that has been
neglected or abandoned.
27
73. The differences between the project areas and the control communities are shown in
Figure 3. The comparison
shows that more members of Figure 3: Household Members Over
Other
15 Years Old Who Work Away
the project site work away
than in the control sites. They 160
Capital city
are more likely to work in a City in another

140 9
rural zone in another district, province
33
in the provincial capital, and in 120 Rural zone in
the national capital city. In 9 another province
gender terms, more men than
number of cases
100
19 17
women are likely to work Capital city of
80 10 province
away, but of the women who
do work away, a larger 60
13 City in another
17
proportion are employed in the district
17
capital city or other cities, and 40 13 Rural zone in
less in rural zones, implying 35 32 another district
20
that they are more likely to
Town/city in
move for skilled or semiskilled 0
same district
nonagricultural labor than for Control Project
seasonal agricultural work in Source: Household survey data.

Rural zone,
the immediate locality. same district
D. Changes in Income and Welfare
74. The multidimensional poverty definition used in the study considers that poverty
reduction can come from changes in sources of income and better accessibility to social
services. Diversification of income can reduce the vulnerability to external shocks.37
Improvements in rural roads are expected to generate new opportunities to earn income and
enhance accessibility to other
Figure 4: Reported Number of Months
with Food Deficit number of essential services. This
months section presents evidence
200 from the case study areas
12
180 about changes in income
10 sources and welfare and how
160
14 such changes relate to
8
140 13 15 different socioeconomic
number of cases
120 25 20 7 groups.
100 26
44
6 1. Food Security
80 44 5
75. Food security is a
60
42 4 critical indicator of poverty
40
29
common to all of the study
3
20
locations. A large part of each
18 13 2 community is affected by lack
0
Control Project
of food at critical points of the
1
year.38 Overall, 74% of all
Source: Household survey data. survey respondents reported
37
The study assumes that the poor and very poor being risk averse will not diversify to other sources of income
unless it reduces their poverty status.
38
Though difficult, a common definition of food security was applied across the three countries. Food security was
defined as having sufficient food to meet the household’s daily needs through the normal productive activities of
household members.
28
facing some food shortages during the year. This broadly corresponds to the poverty
classifications shown earlier. It is only those who are considered better off or rich who are
completely free from any food difficulties. Of those who reported food difficulties, marginally
fewer live in the project sites than in the control sites (Figure 4). There is also little difference in
the reported number of months of food insecurity between the sites, suggesting that food
insecurity affects a broader group of people. However, more project than control respondents
felt that food security had improved, and fewer reported that it had worsened, suggesting that
there may be some connection between improved food security and the road improvements.
Among the very poor in the project sites, 49% felt that the situation had worsened, in
comparison with 47% of the poor, and 28% of the better off. This may be a result of several
other factors that influence poverty.
2. Household Income
76. Better-Off Groups.39 All groups in the study villages have benefited potentially from the
improved opportunity that better roads bring. There has been a better availability and mix of
transport services, and an increased flow
of goods and services. The project and Box 8: The Better Off Get Better
control sites studied are primarily Mr. Labarga from the Negros project site belongs to the
agricultural communities and, hence, better-off group. He has invested savings from his fish
vulnerable to natural shocks or vending business to improve his business as a result of
the better road (Appendix 8, Box 11).
fluctuations in commodity prices. Their
ability to diversify to other sources of
income is, therefore, important to reduce poverty. In practice, of course, it is those who are most
secure and with savings who are able to make the best use of the opportunities better roads
may bring. In fact, case study evidence suggests that better rural roads allow those with some
savings to diversify into activities with substantially better returns. Potential benefits from better
road access and transport services appear to increase in relation to the degree to which
households are nonpoor. The better off have surplus funds to invest in trading (even at very
modest levels), have an agricultural surplus to sell, or have the network of connections and
relationships outside the community enabling them to take advantage of trading or working
opportunities. They have the security to be able to explore outside the village and the immediate
locale for opportunities to diversify income and livelihood. People engaged as salaried workers
in nearby town centers rely on a regular and rapid link and so benefit substantially from the
efficiency and cost savings in commuting. They are heavily represented among the households
considered better off in the study sites.
77. Improvements in income were a key area of inquiry for the household survey. Table 8
below shows that a substantial proportion of villagers reported no change in income sources
reflecting their lack of prerequisites to diversify. Those reporting no change were higher among
control group (58%) than project (47%) respondents. Among all socioeconomic groups, shift
away from agriculture in the past 5 years was more apparent in the project sites than in the
control sites. Of all project site respondents, 23% reported receiving less income from
agriculture, more from other sources, against 14% of control site respondents. Of the different
socioeconomic groups across all study communities, the better off have both diversified and
increased their income more than the poor and very poor. About 22% of better-off households
report increasing both agriculture and other sources of income, whereas more than 50% of both
poor and very poor groups report no change at all in sources of income.
39
The better off in the study communities are not poor in a national sense (Appendix 3, para. 4).
29
Table 8: Change in the Source of Income Over the Past 5 Years (%)
Change Type of Site Socioeconomic Group
Project Control Total Better Off Very Poor Poor Total
No Change 47 58 53 36 54 55 52
Less Income from Agriculture, 23 14 19 24 15 20 19
More from Other Sources
More Income from Agriculture, 8 7 7 9 7 8 8
Less from Other Sources
More Income from Both 8 5 7 22 3 5 7
Agriculture and Other Sources
More Money Sent from Outside 2 2 2 2 2 2 2
Other 12 14 13 8 19 10 12
100 100 100 100 100 100 100
78. All socioeconomic groups spend an approximately similar proportion of household

expenditure on transport and travel, 5% on average. In response to what plans they have to
increase income in the future, control households showed themselves more likely to expand
agricultural production (29%) or raise small animals for income (22%), whereas project
households appeared more likely to arrange a job locally (7%) or start or expand a small
business (25%). Differences are also clear between socioeconomic groups in how they plan to
respond to the better opportunities that roads may bring. Of the very poor, 29% have no plan to
diversify or increase income, against 18% of the poor and 15% of the better off. The better off
are twice as likely (34%) to start or expand a small business as the very poor.
79. Poor and Very Poor Groups. Transport needs are clearly different for different
socioeconomic groups. The case studies show that most of the journeys made by the rural poor
are for subsistence and household tasks, rather than for activities that are directly productive in
Figure 5: Frequency of Use of Transport Services an economic sense. For
(project sites) the rural poor, access to
100 local facilities and the
21 47 30 primary transport network
90
is critical during times of
80 need. The poor lack both
76 time and energy; and
percentage of respondents
70
68 impacts that either reduce
60 or exacerbate these
50
deficiencies have a critical
51
Transport use bearing on poverty.
40 Survey responses among
30
Use transport only different socioeconomic
occasionally groups in the project
20
locations show clearly
Often use transport
10 how the use of transport
0 Never use transport services differs (Figure 5).
better off very poor poor Among the very poor,
Source: Household survey data. 47% say that they use
transport only occasionally
because they have little need for traveling outside the community more regularly, compared to
only 21% of the better off and 30% of the poor.
30
80. There is little evidence from the case studies of an increase in personal mobility among
the very poor following road rehabilitation in search of job opportunities outside the community
or for any other reason. Box 9: The Poor Able to Graduate if Conditions Permit
But where the economic Mr. Bon and his family belonged to the poor group in Sorsogon. Using his
conditions are right, better carpentry skills, he was sometimes able to save money, which his
basic road access can enterprising wife used for occasional banana supply business. The
impact on the local wage- improved road greatly helped their business and increased their income
(Appendix 8, Box 6).
laboring and trading
Mr. Valenzuela and his family belonged to the poor group in Negros project
prospects of the poor, and site. He used his carpentry skills and a network of contractors (made
thus enable them to benefit possible by the improved road) to find other opportunities outside. He is
from wider processes of also using the travel time savings for productive purposes (Appendix 8,
increased agricultural Box 8).
Mrs. Peti in Bengkulu who used to be poor started a small business as a
commercialization and vendor at the market before the road rehabilitation and used her savings to
trade, facilitated by better expand the business to a kiosk. She has now been able to buy a truck to
roads (Appendix 6). In the further her opportunities (Appendix 8, Box 17).
study areas, a few
households graduated from the poor to nonpoor status because of the opportunities that the
road provided.
3. Welfare Indicators
81. Previous sections have shown that direct benefits of rural road improvements appear to
go substantially to transport operators and better-off members of the communities along the
road who use the road regularly and rely on it for their business or employment. The poor and
very poor use the road irregularly for occasional but necessary travel and do not depend much
on fast and reliable transport services. Many of the perceivable benefits of rural roads to them
are indirect and nonquantifiable, but very important nevertheless. These indirect impacts are
discussed in detail in the following sections.40
82. Safety Net for a Crisis. In circumstances where conditions for agriculture are difficult,
the road has proved to be an important safety net of sorts in generating alternative livelihood
opportunities. The severe drought over the
Box 10: Complementary Activities Connected to the
Road—Helping the Poor Cope with Crisis past 3 years in Kurunegala has heavily
impacted on the landless who are
The Weeragama Milk Cooperative’s collection center employed on the fields of the landowners
allows even the poor to sell milk regularly as an in a variety of tasks throughout the year.
alternative livelihood. This operation relies heavily on
all-weather road access (Appendix 8, Box 10).
All have been forced to try and find
All-weather access has also allowed small fish traders alternative ways of generating income and
on bicycles and motorcycles to operate in the securing their basic needs, and the road
surrounding villages (Appendix 6). has been crucial in the development of
The villagers collect cow dung as a collection truck these alternative livelihood strategies such
began operating after road rehabilitation (Appendix 6).
as raising livestock. This alternative
income stream, though temporary or seasonal, is important for household food security. A good
road surface and the guarantee of all-year accessibility appear, therefore, to be important
prerequisites for the development of any kind of regular enterprise, and the road in Kurunegala
has proved to be an important enabling factor in offering villagers such security in difficult times
(Appendix 6).
40
Improvement of social activities to attend social and religious events and strengthen family social networks is an
important advantage of rural roads, though not elaborated below.
31
83. Social Safety Net. One of the major benefits to the poor of improved rural roads is their
ability to better access state, in particular health, services. In all study sites, key informant
discussions revealed that this was one of the most regularly cited positive impacts. Enabling
travel to a health center or hospital is one of the most important uses of the road to the poor.
Isolation and remoteness contribute to the perception of poverty among rural people. This
perception is reduced with the
Figure 6: How Transport to Health Services assurance that it is possible in an
Has Changed in the Last 5 Years emergency to easily access health
facilities and that the road is passable
Can now use bicycle
at any time of the year. In comparing
3.0% the medical services used by survey
Services nearby now respondents in the control and project
18.2% areas, control households are twice as
likely to use a traditional healer.
Project households clearly make better
Other Now use transport
use of medical facilities through the
road. Of control households, 14% stay
10.6% 68.2%
at home in response to poor health or
a health emergency against 7% of
project households. For basic needs,
53% of control households access a
hospital or dispensary against 76% of
Source: Household survey data from project sites.
project households. All socioeconomic
groups appear to access health
services broadly in equal measure. In traveling for treatment, over 50% of control households
walk in comparison to 32% of project households, and the latter are much more likely to use
either a bicycle, 3-wheeler, or van than control households. Among the very poor, 50% overall
still walk to access medical treatment, against 42% of the poor and 21% of the better off.
Answers to the question of how travel to health services has changed over the past 5 years in
the project sites since the improvement of the road are shown in Figure 6. Previously, they
would either walk or have no easy access to such services.
84. Improved Services. As well as enabling the poor and very poor to access medical
facilities, the road also helps teachers and medical staff working in remote locations to carry out
their jobs more effectively. The situation in both Box 11: Teacher/Nurse/Extension Officer
study locations in Indonesia demonstrates this
clearly. The improved asphalt roads in Mr. Sidik, a teacher in Yogyakarta, saves time
Bengkulu and Yogyakarta ensure all-season traveling for his regular job and uses spare time to
access for teachers to the village. Better roads work on his farm (Appendix 8, Box 20).
Mrs. Suryatini, a nurse in Yogyakarta, can do her
have shortened their travel times, and the wider job much more effectively, visit her clients year
variety of available transport services means round, and improve preventive care (Appendix 8,
that they can stay longer at school. Extension Box 21).
officers also can achieve more of their work Mr. Sumianto, an extension worker in Yogyakarta,
has been able to improve his achievement rate to
targets. They are able to cover the areas of 90% of target compared to 40% before road
their responsibility much more quickly and rehabilitation (Appendix 8, Box 22).
efficiently, and so deliver a more regular and
reliable service. Officials report higher motivation to do their jobs, and some have been able to
take up secondary occupations to supplement their incomes because of the time saving through
better transport. Education and health officials in Sorsogon report that it is easier to monitor the
provision of services in the field with better roads and the use of motorcycles or public transport.
85. Furthering the Reach of the State. Improved rural roads create the conditions for
better access of people to services, and of services to the village. They also facilitate the reach
32
of the state into new areas. At the Matara project site, villagers remarked how, prior to the road
rehabilitation, the traditional authority of the village head extended into all aspects of village life.
He would arbitrate in disputes over land issues, for example, and would resolve matters of crime
that arose in the community. Since road rehabilitation, police from the vicinity can visit the
community more regularly. Villagers now have access to a more neutral arbiter and consider
that they have some protection under the law as these matters are now outside the sphere of
everyday village relations. In the Sorsogon project site, a number of key informants commented
on the relationship between extending and improving the road network and tackling the issue of
insurgency in the area. Better roads mean that remote locations are no longer so isolated and
vulnerable, and the army has much better mobility in accessing interior areas. Better rural roads
promote the spread of information and ideas seen by the authorities as crucial in countering any
ideological hold the insurgents may have over more isolated communities. Roads bring regular
contact with the outside world and bring remote areas within the purview of the state and other
networks.
86. Broadening Opportunities. A major benefit shared across socioeconomic groups is the
greater availability of time, resulting from better access and improved transportation. Obviously,
those in the community who travel most frequently benefit most from this. But for the very poor
too, improved access even locally can mean significant time savings, and thus time to engage in
other productive activities (Appendix 8, Box 5). The road also broadens opportunities and
choice. There are more services and facilities now available. The road is an important enabling
condition in the decision on whether or not to develop a small business. Small and micro-
businesses are an important means of income creation and enhancement for the poor, provided
that they have some initial capital. Examples of small businesses along the roadside include the
vending of fish, vegetables, and cooked foods; stores selling small quantities of household items
such as soap, candies, salt, and rice; and small manufacturing and repair, like bicycle repair
stalls. The road, together with other enabling conditions, offers better opportunities for
broadening and diversifying livelihoods.
87. Exploitation. Roads are critical social arteries for the penetration and exchange of
ideas, culture, and information. This can be a positive force in exposing villagers to new and
dynamic flows of information and opportunities, and increasing the desire for mobility,
particularly among the young. With increased opportunities for mobility, however, there are also
increased risks of exposure to negative influences from towns and cities, such as drugs and
risks of exposure to the sex trade in nearby urban centers and trafficking, particularly for young,
poor rural women. Provincial health officials in the Philippines identified the risks of sexual
exploitation and drugs as key negative risks arising from the development of an improved rural
road network. During the PRAs in the Indonesian locations, older members of the focus groups
particularly associated increased access and mobility with an influx of new ideas and practices
and the erosion of older village values.
88. Other Impacts. An impact often associated with the development of rural roads is
increased land prices, resulting from better accessibility between rural and urban areas. This
process can often alienate the poor, whose rights to accessing land and resources may not be
formal, but traditional and based on long years of practice. As land values increase, land
ownership is formalized, and the poor are alienated from the resource base on which they have
traditionally relied. In the study areas, however, there is little evidence of such a shift in access
to land or of a steep increase in land prices. Neither is a high turnover in land ownership
following road rehabilitation apparent. The structure of land ownership and the market for land
have remained much as they were before project implementation. The survey inquired about the
main perceived negative effects of the road improvements. The most frequent response by far
was that no negative effects (74%) had resulted. Of those negative effects observed, outside
33
interests buying up land accounted for the second highest response (10%), followed by
increased traffic accidents (5%), and traffic noise (4%).
VI. CONCLUSIONS AND RECOMMENDATIONS
A. Study Conclusions
89. Improved roads guarantee rural access throughout the year, so that those villagers who
are able, can now invest time and resources in an outside endeavor. They know that
communications and links with the outside world are reasonably assured, and can make use of
a wider variety of transport services, which are more frequent, take less time to travel to nearby
centers, and are often cheaper than before road rehabilitation. Rural roads are, therefore, an
important enabling condition for livelihood development for people in the project sites.41
90. The poor and very poor assign a high priority to basic access. It reduces their
vulnerability, and they consider it a matter of dignity to be able to communicate with the outside
world and engage in social activities outside the village. This is confirmed by evidence from
PRAs in the project and control sites. In the absence of improved opportunities to use roads, the
poor rely on the primary network of paths, tracks, culverts, and basic access routes in the
immediate village vicinity. Theirs is a village, walking world, and improvements to this primary
village network of tracks, etc., that reduce the burden in undertaking basic household and
productive tasks are likely to have a significant poverty reduction impact by reducing their time
and energy impoverishment. In this context, the increased availability of intermediate modes of
transport with larger carrying capacity to collect water, firewood, etc., is likely to have a greater
initial impact on their well-being. They need to first accumulate surpluses even periodically to be
able to seize new opportunities that motorized transportation may bring.
91. Therefore, the study confirms that better rural roads are a necessary but not sufficient
condition for graduating from poverty. There is little evidence that roads have impacted directly
in terms of reducing poverty on those groups in each study community who were identified
explicitly as being very poor. The ability of the poor and very poor to make significant economic
use of the road depends on their asset base and the entitlements to resources and opportunities
that they can command, as well as the passage of time. In a few instances, the poor who have
invested savings in a small business or used their skills have graduated from poverty, using the
benefits from the road. However, the poor and very poor primarily benefit through the indirect
impacts of road improvements, of better access to state services and improved provision of
services to the village, and of opportunities in alternative livelihood income streams where the
preconditions for their development are right. The poor can also benefit broadly from
improvements to the rural economy through increased opportunities for agricultural wage labor,
but again these impacts are contingent on favorable preconditions being in place. The
preconditions shaping the potential for development of an area can be classified broadly into
two groups.
92. The first group consists of the external conditions around which projects must be
designed. These are the macro factors, which, in the context of an investment project design,
must be accepted as given. The spatial position appears to have a significant bearing on
development. Remoteness is an aspect of poverty, and where communities are far from existing
marketing centers the dynamism of development is lessened considerably. Impacts appear to
be of a higher order in locations closer to major centers, or where the density of population and
settlement is higher, than in areas that are much more remote. Climatic conditions and the
natural environment are also important; they must be able to support any potential increase in
41
Those along the roads are also subject to road safety issues, though this was not prominent on case study roads.
34
agricultural production. The macroeconomic context and the prevailing terms of trade for
agricultural produce are also critical to all of those area economies that are intimately linked to
primary commodity production. Again, the scope for intervention of a rural road investment here
is limited, except through livelihood diversification.
93. The second group covers the structural poverty problems. These are issues of resource
allocation and distribution, and prevailing social structure. The socioeconomic and cultural
landscape is a critical prism through which communities are impacted by rural road investments.
But again, the capacity of such investments to shape or influence this landscape is limited. The
nature and distribution of land ownership is a fundamental determinant, but is a political issue
that ADB can hope to influence only in broad terms over a period of time and at a macro level,
and not over the course of a project cycle. Similarly, the social and cultural structure of
communities is determined by broad historical processes and is unlikely to change immediately
through rural road investments. What the investment can do is provide a vehicle for change
through complementary activities, which broaden opportunities for the poor and, at the same
time, challenge some of the ingrained structural perceptions that restrict opportunities otherwise.
An example of this is the direct employment of single, female household heads in rural road
rehabilitation work. In most countries, they are one of the poorest groups, and employment on a
road can improve income, broaden choice, and enhance self-esteem and confidence. Although
rural road projects cannot directly challenge the structure of land ownership, they can, given the
right complementary activities, broaden livelihood opportunities outside agriculture for the
landless poor. This, in turn, can help the poor escape from chronic debt cycles, and so escape
from poverty.
94. The case studies selected in each country were either sector road investments or
integrated projects. The impact of improved transport services that enhanced competition and
reduced transport costs for the entire community (especially those using regular transport)
appeared to be greater in the transport sector projects. Under the integrated projects, it was
anticipated that the poverty reduction impact would be greater as roads were linked to other
investments, which could broaden livelihood opportunities for the poor. In practice, however, this
was not always the case.42 In Sorsogon, the parallel project activities were not effectively
integrated: investments in roads were made in one area, and in tree nurseries and irrigation
works in others. There is some evidence that the integrated projects in Indonesia and Sri Lanka
were more successful than those in the Philippines in improving livelihoods for the poor, as the
roads covered in these projects were basic gravel roads of which the poor can make more use
in accessing local services and destinations. However, the projects were not sufficiently focused
on the livelihoods of the poor. The Bengkulu project had a minimum landholding requirement for
farmers to be able to take part, and they had to have sufficient land and reserves of family labor
in order to qualify.43 In Kurunegala, the irrigation tank rehabilitation work favored landholding
farmers directly, with benefits assumed to trickle down to the poor in the form of increased
wage-labor opportunities.
95. It is recognized that poverty reduction was not an explicit and primary objective in the
design of most of these projects, a point frequently made during fieldwork by local executing
agency staff. They requested that project design in the future be much more explicit about the
objective of poverty reduction and that appropriate measures be included as the means of
achieving this. More pro-poor investments in the future require genuinely integrated project
components that offer the poor, too, some opportunity to diversify and broaden livelihoods, and
thus strengthen livelihood capital with which to make use of improved rural roads. For this,
linkages with complementary activities and services, which support the broadening of livelihood
42
Some anticipated impacts were diminished by shocks such as droughts and adverse world market prices.
43
These criteria were introduced by the executing agencies to ensure the viability of newly-established plantations
and to ensure high aggregate rubber yields.
35
opportunities for the poor, are needed. Simply improving a road is not enough; the poor also
require support in being able to make use of it.
B. Recommendations for Strengthening Poverty Reduction Impacts
1. Measures Directly Related to Transport
96. Broadening Access. All of the case study locations demonstrated, to a greater or lesser
degree, a high level of road density.44 Given the density of existing rural road networks in the
study locations, perhaps there is a need now also to look at how to make better use of existing
roads and infrastructure. It could mean designing interventions that concentrate on removing the
access and mobility constraints of the poor in their existing livelihoods, and thus making
investments in tracks, paths, culverts, and crossings, as well as improving transport modalities
and their carrying capacity, especially intermediate (nonmotorized) means of transport that
benefit the poor.
97. Periodic Maintenance. Long-term social and economic benefits from roads are often
threatened by a neglect of periodic maintenance. Rural roads, particularly gravel roads, quickly
deteriorate if not regularly maintained, and benefits can be quickly lost if they are periodically
impassable or the overall condition is bad. The poor are generally risk averse and will not
engage in a new activity if they know that the road on which it depends will be temporarily
unusable or if its poorer condition in the following year will mean that costs and time will
increase. Devolving responsibility for road maintenance to local communities, particularly for
basic rural roads, is a means of ensuring simultaneously that the poor can receive benefits
through direct employment and that local communities are stakeholders in the road serving their
area. This can be a powerful means of ensuring long-term sustainability, if supported by local
ownership and technical expertise of local authorities. The income generated from employment
on the road can also provide start-up capital for the poor to make use of the road in some future
enterprise or endeavor.
98. Regulatory Environment and Affordable Transport. The role of governments is

critical in two ways. First, governments have a crucial role to play in facilitating a suitable
regulatory environment for the development of competition in the transport sector. If the
preconditions for the development of competition are not present (para. 50), governments may
need to provide transport services as a public service obligation. However, if the provision of
roads has been coupled with incentives for livelihood enhancement (para. 101), the poor may
be able to diversify income through selling craftwork, cash crops, or providing labor in nearby
markets. The state transport services could then be removed as income from these activities
gives rise to sustainable private transport services.
99. Criteria for Road Location and Type. Often, the decisions about what roads are
selected are not transparent. Participation is a key element in the selection of roads to be
rehabilitated or reconstructed under rural road programs, not only to decide which roads need
improvement but also to understand the type of road and ancillary investments that are
needed.45 If poverty reduction is to be a serious consideration, then criteria for road selection
should include a poverty component. This can be a weighting by poverty population within the
zone of influence of a road, in combination with other conventional criteria such as vehicular
traffic and population density. The roads serving poverty areas could be upgraded only to a
standard that is needed to connect them to the main road network.
44
This was confirmed by the difficulties in identifying suitable control areas (Appendix 1, para. 27).
45
However, the study acknowledges that detailed design of a particular road needs to be carried out by engineers
with appropriate competencies in the most cost-effective manner.
36
100. Labor-Based Technology. Rural roads are an enabling condition for poverty reduction,
and a very necessary one. Direct benefits of roads alone do not come very readily to the poor.
One very important way in which they do come, however, is through direct employment of the
poor in labor-based road construction and maintenance. Experience from Asia and Africa shows
that, given a sufficiently long period of employment on the road, the poor can accumulate capital
to invest in alternative livelihood opportunities and thus move away from poverty.46
2. Measures Not Directly Related to Transport
101. Integrated Projects. Experience from the case studies shows that rural roads alone are
not enough in tackling poverty. The poor face fundamental deficiencies in their assets to take
advantage of better opportunities that a rural road may bring and, therefore, need support to
capitalize. This suggests that integrated projects are needed to tackle poverty effectively. The
case studies covered both projects that were sector road investments, and integrated
investments where the road was one part of a larger program of support. In practice, these
integrated projects were either not truly integrated, or were focused largely on benefiting better-
off farmers’ groups. The poor require genuinely integrated programs of support right through the
cycles of production, transportation, and sale.47 Diversification into alternative livelihood
opportunities will also cushion them against the impact of adverse movements in commodity
prices.
102. Promoting Understanding. Another key role that governments have to play is in
fostering an understanding of the priority of poverty reduction, and disseminating this
understanding to lower levels of government and to implementing agencies. Where a
commitment to explicit poverty reduction objectives is missing at local implementation levels, the
poverty impact can be missed completely. Governments’ institutional understanding of, and
commitment to, poverty reduction must be equal to that of ADB, and shared objectives should
be agreed toward the common end of poverty reduction and addressing related structural
barriers.
103. Participatory Design and Planning. The preparation in the context of an integrated
program needs a proper period of mobilization and preparation in order to be effective and
sustainable. Mechanisms should be institutionalized to ensure that the poor themselves are
involved in many aspects of the investment design (but not engineering design),
implementation, and operation and maintenance. Including the poor in identifying livelihood
opportunities would ensure that programs are relevant to their needs and skills, that they are the
principal beneficiaries, and that the benefits are sustainable.
104. Poverty Assessment. If poverty reduction is an objective, a proper analysis at project

design stage is necessary of who are the poor and on what factors do their livelihood strategies
depend. This requires a stakeholder analysis and poverty benchmarking to establish a baseline
against which project impacts can be measured in the future. Experience from the study shows
that significant social benefits accrue to the poor from rural road investments, particularly in
accessing outside services. These are inherently unquantifiable, but can be assessed if
effective, structured baseline research is done at inception.
46
An example of the effective and sustainable use of labor-based methods for poverty reduction is the Nepal Rural
Access Program design, where poverty groups are targeted for employment (including poor female household
heads, and women employed traditionally as long-distance porters whose livelihood will suffer from road
construction) during agricultural off-peak times without compromising other livelihood activities. Workers should
contribute to parallel credit and savings schemes that allow them to move to livelihood alternatives once road
construction is finished.
47
Experience from the project site in Bengkulu, Indonesia shows that the formation of farmers’ groups and
cooperatives is not in itself enough. The groups must go beyond simply producing a particular commodity or crop
and be able to undertake credit and marketing and also be able to access affordable transportation.
37
105. Project Performance Monitoring. Regular monitoring and evaluation of livelihood

impacts against this baseline should be carried out through a clearly defined project
performance monitoring system. This will ensure that, in the future, impacts on the poor are
captured, rather than assumed.48 Benefit monitoring and evaluation systems49 were part of the
ADB design of most of the case studies covered by this evaluation, but were exclusively the
responsibility of the executing agencies, which were often expected to meet the cost from
counterpart funding. In a situation of scarcity of local resources, such systems were often
abandoned, or allowed to lapse for so long that information that was eventually collected was
largely useless. Monitoring indicators were also usually progress indicators, measuring fund
disbursement and physical works, and were not complemented by impact indicators for
identifying exactly what the outcome of works had been, particularly on the poor.
106. Partners. How best to implement rural roads with integrated livelihood support
components is a difficult question. ADB may require borrowers to take on partners with a proven
track record in aspects of the program design, such as bilateral agencies, or local or
international NGOs who have a proven track record in mobilizing and working closely with
communities. It should be recognized that complex, integrated programs may require a number
of partners, and ADB can play a key role in helping executing agencies manage and coordinate
this process effectively.
48
Often, impacts on the villagers and particularly the poor and very poor take long time to materialize and need to be
monitored periodically to evaluate progress.
49
Benefit monitoring and evaluation has now been superseded by ADB’s project performance monitoring system.
38 Appendix 1
METHODOLOGY DETAILS
A. Study Direction Determined at Methodology Workshop
1. Chapter I gives a brief description of the methodology used for the study including
methodological assumptions, case study selection, and study tools used to validate and
attribute impacts (Main Report, paras. 6–9). This appendix gives further details that steered the
study methodology including the criteria for selecting the projects, study areas, and
effectiveness of various study tools used. Following the literature review and the review of
documents, the study team developed a preliminary methodology approach, which included a
list of suggested instruments, definitions to be used, and possible key indicators. During the
methodology workshop, the participants clarified the direction the study should take. The
following indicate the key points determined:
(i) The study was to focus primarily on rural road investments undertaken in the
recent past, either as a transport sector investment or as a component of an
agriculture sector integrated project.
(ii) Given time and financial resource limitations, a case study approach was to be
adopted to capture a deeper understanding of a variety of scenarios rather than
selecting a proportionately representative sample from each project.
(iii) The study tried to maximize the learning experience by understanding the factors
that drive the impacts rather than measuring the magnitude of the impacts,
relying rigidly on the double difference technique.1
(iv) The study consciously focused on poor areas to ensure that impacts on the
poor’s livelihoods could be better captured.
(v) Based on the limited information available in a retrospective study, a variety of
tools were used sequentially with each intending to inform the next and all
geared to support and validate the findings.
B. Methodological Assumptions
2. Multifaceted Impacts. Instead of assuming an automatic link between constructing rural

roads and poverty reduction, the study methodology gave due consideration to the intervening
sociocultural, economic, environmental, and institutional factors that determine how people
respond to the road and that shape livelihood constraints and opportunities. This meant that
primarily relying on classical road assessment tools such as traffic movements, vehicle
operating costs (VOCs), and freight and passenger prices to assess impacts was not sufficient.
They were also not readily applicable to rural roads2 where traffic volumes and services are
often extremely low. Rural roads can also be social and economic arteries for communities in a
broader sense, with a myriad of localized impacts not readily accessible to analysis using
traditional means. This may include the ability to make periodic social visits, a broadening of
choice about where and whether to sell surplus agricultural produce, migration for seasonal
work, access to health services, and a greater readiness of professionals to work in remote
areas (as doctors, teachers, etc.). Negative impacts could include a greater openness to outside
interests to exploit natural resources. To understand these multifaceted processes, deeper
qualitative understanding and closer interaction with the villagers were needed. This was the
primary basis for using a case study approach.
1
The double difference technique is a tool that uses the differences in impacts before and after an intervention and
also between a project site and a control site (with and without project) to attribute impacts of the intervention.
2
Rural roads were defined as roads through rural areas. Different countries have varying classifications for roads
depending on technical specifications, sources of funding, and the importance of the points they connect.
Appendix 1 39
3. Multidimensional Poverty. The study assumed that poverty is a multidimensional

condition, and that lack of income was only one component. This is a central tenet of the
poverty reduction strategy of the Asian Development Bank (ADB), which states that “poverty is a
deprivation of essential assets and opportunities to which every human is entitled.”3 The study
adopted an assets-based approach to defining and understanding poverty in line with recent
comprehensive work done on poverty and transport.4 The study builds on this work to apply and
adapt some of the principles and approaches in the context of road investments in Asia. Key
elements of conceptualizing poverty holistically include (i) lack of income and consumption;
(ii) lack of private and social assets (human, social, natural, physical, financial);
(iii) voicelessness, lack of dignity, and inability to participate; (iv) insecurity and vulnerability to
risk and to shocks; and (v) lack of ability to influence institutions and processes at the macro
and micro levels.5
4. Direct and Indirect Impacts. The study considered the direct impacts to be wholly
attributable to the road improvement and its use. They could be either positive (reduction in
VOCs) or negative (increased road accidents). Indirect impacts are those that occur after a
passage of time as a result of the direct impacts. They too can be positive (reduction in
agricultural input prices) or negative (spread of diseases).
5. Road Infrastructure and Transport Services. The study assumed that rural roads are
a necessary but not sufficient condition for provision of rural transport services. Rather, this
relationship was a critical area of inquiry. In remote locations where transport demand is low,
there may be few incentives for private transport operators to develop services to rural
communities, even where infrastructure is adequate.
6. Access and Mobility. The study adopts a broad focus on “access.” Both physical and
nonphysical barriers to access are assumed. Improving transport for the poor requires both the
provision of infrastructure and services and the consideration of the key socioeconomic
parameters, which may condition people’s ability to make use of wider transport opportunities.
Mobility refers to the ability to travel regularly outside the community in search of work or other
opportunities. Mobility is increasingly important, particularly as the importance of agriculture
fluctuates in the face of changes to global terms of trade for primary products, and opportunities
for migration and commuting increase (Main Report, para. 80).
7. Gender and Transportation. In understanding impacts, a gender perspective is critical

because women and men have different productive and household responsibilities and,
therefore, different transport needs. Women are more likely to lack access to means of transport
or income for travel, but may have a heavy transport burden in meeting productive tasks,
household needs, and community obligations. They are, therefore, more likely to be time and
energy impoverished.6 Women also face significant nonphysical barriers to access due to
cultural restrictions. Extreme poverty can also increase women’s vulnerability to exploitation
through trafficking and commercial sex trade.
3
ADB. 1999. Fighting Poverty in Asia and the Pacific: The Poverty Reduction Strategy. Manila, p. 5.
4
Hanmer, Lucia, Elizabeth Lovell, Robert Chapman, and Tom Slaymaker. 2000. Poverty and Transport: A Toolkit.
Overseas Development Institute.
5
Drawn from various sources including Carney, Diana, ed. 1998. Sustainable Rural Livelihoods. What Contribution
Can We Make? Department for International Development (DFID). United Kingdom; and Ashley, Caroline, and
Diana Carney. 1999. Sustainable Livelihoods: Lessons from Early Experience. DFID.
6
Coedited by Priyanthi Fernando and Gina Porter. 2002. Balancing the Load. Women, Gender and Transport.
International Forum for Rural Transport and Development. Zed Books.
40 Appendix 1
C. Study Tools Used
8. All research activities were intended to be mutually reinforcing components to the

methodology to ensure its robustness in attributing the impacts of rural roads on poverty
reduction.7 The following tools were used sequentially, with each intended to inform the use of
the next.
9. Key Impact Indicators and Baseline Data. Local consultants recruited in each country
worked according to an agreed methodology. An important early task in each country was the
collection of key data indicators for socioeconomic development and agricultural production in
the study areas across the lifetime of the project (usually the past 10 years). They were also
tasked with collecting all available data from the project locally and from existing ADB
documentation. Both these sources were expected to provide the study with baseline data on
trends against which to evaluate emerging impacts. Both of these sources proved difficult to
collect in all study locations.
10. Although many of the projects studied were supposed to have a monitoring and
evaluation component, for beneficiaries particularly, this component was seldom implemented in
practice. Any monitoring and evaluation done was often restricted to project physical progress,
in terms of completion of works, rather than the actual impact that the project had had. This lack
of baseline information restricted the ability of the study team to assess project impacts over
time.8 Where some form of data had been collected, it was not kept or used by implementing
agencies, particularly once the ADB loan had been disbursed, and so was not available to the
study team.
11. A list of key data indicators was also developed for collection in each location. These
data were intended to establish broad background trends within which project impacts could be
located. However, the team perhaps overestimated the degree to which these data would be
available, as in practice, they proved difficult and time consuming to collect. Many of the data
indicators were simply not collected regularly and monitored by provincial or district level
authorities. Consequently, this component of the methodology was only partly completed.
12. Key Informant Interviews. Semistructured interviews with key informants at national,
provincial, and local levels provided background information and perceptions about the impact
of roads. These interviews proved to be very useful, given the time constraints of the study. Key
informant interviews were an important complementary activity to the village and community
level participatory assessment activities in informing researchers of wider dimensions not
necessarily apparent to the poor themselves. The research team, prior to the start of the
interviews, developed preliminary guide questions to semistructure the interview. Often, these
interviews enabled the study team to disaggregate groups in the study area, particularly in
identifying subgroups of the poor. They also helped develop an outline for stakeholder analysis,
which could be further adapted by subsequent participatory research activities.
13. Depending on the knowledge level of the interviewee, it was also possible to gather
explanatory information not available from secondary data. Mid-level provincial administrative
informants and technical informants helped in the site selection and provided historical
information on the selected sites, i.e., in terms of the conditions prior to road improvement. The
7
The study teams were resident in or near each of the case study areas for a period of 2 weeks during February–
April 2002 and had the opportunity to collect detailed information from all types of project stakeholders.
8
This seriously affected the quantification of benefits in an economic sense, as recall techniques could not be
reliably used to quantify the status of the village activities prior to the construction of the road.
Appendix 1 41
technical officers who worked on a particular project and at the local level were naturally able to
provide more detailed and accurate information on the project site. The local level interviews
were particularly helpful in selecting the control site and identifying poor groups within them.
Types of persons interviewed were:9
(i) project management offices at national level;

(ii) project management offices at provincial level;
(iii) other provincial level key informants in agriculture, transport, health, social
services, and infrastructure;
(iv) project management units at district level relevant to case study areas;
(v) extension services staff;
(vi) village level key informants
(a) village leaders
(b) community-based organization and self-help group representatives
(c) schoolteachers
(d) shopkeepers
(e) transport operators
(f) traders within and outside the village
(vii) those who had graduated from poverty since project intervention (interviewed all
possible households available); and
(viii) those who remained poor despite project intervention (interviewed a sample of
households).
14. Participatory Assessment. Structured participatory rural assessment (PRA) techniques

conducted by trained researchers were important in actively engaging the local communities
(and particularly the poor) in the study process. PRA techniques were used primarily to increase
understanding both of the issues facing the poorer segments of the study communities and of
the impact of rural roads on their livelihood patterns and quality of life. A range of PRA methods
were used in the different locations based on the suitability for the particular location, group of
individuals, and priority issues under discussion. The methods used included:
(i) village transect (physical inspection of resources, houses, roads, topography);

(ii) social maps (mapping poor, nonpoor houses, roads, schools, health centers, and
other public facilities);
(iii) resource maps (agricultural field processing centers, water sources, forests);
(iv) income resources analysis (sources of income, roles of men and women,
marketing, and other related issues);
(v) social network (relationships and networks, community-based organizations);
(vi) analysis problem ranking (identification of common problems of men and women
separately and ranking them as a group);
(vii) seasonal calendars (how seasons affect way of life, especially in using the road);
(viii) classification of village groups and definition of each; and
(ix) poverty analysis, trends and attribution analysis, own perceptions, and
community poverty.
15. A wealth of information was generated through these exercises, which was critical in
gaining a greater understanding of the importance of roads and connectivity to different groups
within the village. This information was transferred to the framework for comparison and
9
Most of the household case studies are a result of key informant interviews.
42 Appendix 1
analysis.10 The familiarity gained through these exercises also helped the enumerators to
subsequently penetrate the village for the household surveys. While the study team gained a
greater understanding of the issues, the villagers got the chance to learn about the study and its
objectives. The study team provided as much information as possible, especially regarding the
objectives of the study in an attempt to overcome the recurring perception of the villagers that a
feasibility study for road construction was being carried out. In most cases, the villagers
accepted the study design and orientation. Being in close contact for a length of time facilitated
this understanding among the villagers to a great extent.
16. Traffic Analysis. Traffic analysis was carried out by a trained team who assessed the
volume of traffic, its mix, usage of different modes of transport by the population, ownership of
vehicles, and role of vehicle operators. Vehicle traffic counts were generally undertaken on a
“normal” and “market” day in study locations. These data were intended for comparison with
existing project records of traffic before, during, and after implementation, but this information
was usually not available. Nevertheless, the information generated was useful in informing the
study team of the volume and mix of traffic in each location. The study team carried out the
following surveys:
(i) cordon survey (traffic counts on market and nonmarket days for motorized and
nonmotorized movements) to assess the types and volumes of people and goods
using the roads;
(ii) passenger and freight survey (on moving vehicles) to gauge purpose of travel,
time expended, value of goods transported, and transport prices; and
(iii) interviews with vehicle operators to understand impacts on VOCs, frequencies,
and travel time.
Table A1.1: Average Daily Traffic by Project Sitea

Mode Sorsogonb Negros Kurunegala Matara Bengkulu Yogyakarta
Bicycle 53 254 123 10 13
Cart 7
Motorcycle 28 257 155 153 252 1,301
Tricycle 55 276
3-Wheeler 1 45 1
Tractor 8 16
Truck 4 162 20 45 8 154
Car/Van 7 99 10 43 62 202
Jeepney 22 53
Minibus 40 5 16 34 217
Van/Truck 28
Bus 32 4 24
Total 116 972 457 469 374 1,911
a
These counts reflect the average of a market and nonmarket day in the project site between 6 am and 6 pm of the
traffic in both directions.
b
Due to security concerns, the traffic surveys were carried out only from 8 am to 4 pm in Sorsogon, and the 2-week
fieldwork period did not include a market day. Since this was the first field site, only motorized traffic was counted
and bicycles were not counted. However, there was some bicycle usage in the area, especially among poor
children going to high school as this did not involve a daily transport cost.
10
See a sample framework in Appendix 2.
Appendix 1 43
17. Household Surveys. The household survey was used in common (as far as this was
possible) across all three countries to enable cross-country comparison and identification of key
trends and impacts. It was also collected in both project and control communities. In terms of
time and effort, this occupied a major part of the fieldwork. Data were collected from about
80 households in each study area (40 in the project site and 40 in the control site).11 The
questionnaires were pilot tested in Sorsogon, Philippines and subsequently slightly changed to
adapt to differences in the context of each study area. The enumerators all spoke the local
dialect and generally interviewed the respondents in their homes, enabling sufficient time to
capture information on structured questions, as well as on open-ended ones. The total number
of households surveyed and the gender distribution of the household head is given in Table
A1.2. The survey data were coded and digitized in the field to check for their validity and any
errors, and were analyzed subsequently on return to each country base.
Table A1.2: Gender Distribution of Household Head in the Survey Locations

Location Sex of Household Head Total
Male Female
PHI-Sorsogon Project Site 29 2 31

PHI-Sorsogon Control Site 38 2 40
PHI-Negros Project Site 38 1 39
PHI-Negros Control Site 38 1 39
SRI-Kurunegala Project Site 35 5 40
SRI-Kurunegala Control Site 28 1 29
SRI-Matara Project Site 35 4 39
SRI-Matara Control Site 35 6 41
INO-Bengkulu Project Site 38 2 40
INO-Bengkulu Control Site 38 2 40
INO-Yogyakarta Project Site 34 6 40
INO-Yogyakarta Control Site 38 1 39
Total 424 33 457
Percentage 93 7 100
INO = Indonesia, PHI = Philippines, SRI = Sri Lanka.
18. Workshops. Country feedback and discussions were held in one location in each
country following the completion of the draft country report. The preliminary findings were
discussed with implementing authorities, local officials, and community representatives, allowing
the team to validate findings. The workshop held in ADB also enabled the dissemination of
preliminary findings and discussion of emerging issues and recommendations.
19. Framework and Triangulation of Information. As discussed above, a variety of

different methodological tools were used to ensure effective cross-checking and validation
through the “triangulation” of findings. In the field, the means of documenting information from
PRA and key informants was the reporting framework. It developed as a common means of
analysis in each country.12 Findings from the qualitative research activities were fed into this
framework by the study team and used in writing up country case study reports. The framework
adopted an assets-based conceptual approach to poverty, and was intended to capture
information at a number of different levels including the “process” and dynamism associated
with changes arising from the road project. In conceptualizing poverty through individual and
11
In some areas, the low density of population in the control site did not allow for 40 household surveys.
12
Sample framework in Appendix 2.
44 Appendix 1
community assets, the framework also analyzes the institutions and processes that determine
opportunities and constraints for the poor; the operation of government at different levels (local,
regional, national); the private sector; nongovernment organizations and community-based
organizations; traditional institutions and authorities; and the role of prevailing social norms,
laws, and regulations in shaping the institutional and governance context. The framework
proved to be particularly helpful in documenting the participatory and qualitative research
activities. The qualitative information from the framework was combined with household and
traffic surveys on the one hand and key impact indicators and secondary data on the other to
cross-check and validate, thus improving the robustness of the data. This triangulation of
information is depicted in the figure below.
Triangulation of Information
Household survey and

traffic passenger surveys
Robust
information
Key impact indicators
and secondary data
PRA activities and key
informant interviews
PRA = participatory rural assessment.
Source: Study team.
20. Attribution of Impacts. Given the absence of useful baseline information, it was difficult
to use double difference technique quantitatively. Instead, recall techniques were used to
compare before and after conditions in a qualitative sense. Even the recall techniques proved to
be quite difficult to use in these areas where the road appeared not to be an important
determinant in the villagers routine struggles. Therefore, even a modified double difference
technique was ineffective, making rigorous attribution of impacts impossible (Main Report, para.
8). Instead, the control areas were used to understand the difficulties the poor face without
accessibility to the outside and the extent to which easing these constraints fits in with their
poverty status.
D. Project Selection Criteria
21. It is important that in the case studies selected, sufficient time has elapsed since the
physical completion of the rural road. Impacts that may occur through changes in economic
activities, increased accessibility, enhanced communication, etc., take time to develop
subsequent to opening of a rural community. If too much time has elapsed, it would be more
difficult to discern the impact of other major interventions from those of the rural road. Therefore,
a careful balance is needed in terms of the time frame for selecting case study projects. As
such, case study projects selected had to have been physically completed a year or more
before the fieldwork. Therefore, projects approved mostly in the 1990s, which satisfy the other
criteria, were selected for a first review. To be cost effective in terms of time and resources
spent, three countries with two projects in each were to be identified. The three selected were
Philippines, Sri Lanka, and Indonesia. This was influenced by the decision not to duplicate other
Appendix 1 45
ongoing efforts but to complement them.13 In each country, a rural road component under a
transport sector project and an integrated rural development project (under agriculture sector
projects) with a rural road component were selected with the intention of comparing the impacts
of road investments on their own, or as part of a wider program.
E. Study Area Selection Criteria
22. Given the limited time available for the study and the extensive geographic scope of
ADB operations in Asia, it was clearly important the selected fieldwork sites were both
representative of the prevailing conditions in rural Asia and broadly representative of the access
conditions and restraints faced by its rural people. One of the projects selected in each country
was a sector-based road investment, and the other an integrated rural development project with
a rural roads component. In selecting these different kinds of projects, it was intended that the
impact of road investments on their own, or as part of a wider program, could also be compared.
23. Cumulatively, the six field sites selected across the three countries covered a broad
range of both the physical and nonphysical factors likely to condition the context for rural road
interventions across ADB operations in Asia. But road projects by their linear nature can cover
long distances, and the likelihood of changes in physical, environmental, and social conditions
is, therefore, increased even within one project, making a uniformly representative site selection
difficult. Besides, each country, region, and specific area has characteristics particular only to
that location, and the case studies were to capture these contextual scenarios which flavor the
findings.
24. Prior to the start of fieldwork, the local consultants visited possible project locations to
establish preliminary contacts and collect baseline data as well as key impact indicators. They
also visited a few road segments selected for inspection based on the road selection criteria.
The road area selected had to be (i) representative of the prevailing poverty conditions faced by
rural people in the project area; (ii) broadly representative of the access conditions and
restraints faced by rural people; (iii) a reasonably long project road segment (about
10 kilometers) to have an impact on the level of economic activity; and (iv) have another
community in the vicinity that could act as a control area.
25. The selection of a focus study road link was done in consultation with local staff, village
leaders, and officials responsible for implementation of the projects. They had knowledge and
experience of the area, its poorer areas, and were best placed to identify a road section suitable
for the study purpose. Some local authorities later became key informants. It is important to
emphasize that road segments in poor districts were purposely selected during this process,
and that conclusions from the study locations are, therefore, not necessarily applicable across
all sites covered by the projects. In keeping with the poverty focus of the study, locations were
selected where the density of poverty was assumed to be high. A focus community (project site)
was selected from along the road link where research activities were concentrated. The same
13
The People’s Republic of China, India, and Thailand were selected under the ongoing TA 5947-REG: Assessing
the Impact of Transport and Energy Infrastructure on Poverty Reduction, for $800,000, approved on 25 October
2000, which was a broader study but nevertheless covered rural road projects. A similar impact evaluation of
Bangladesh rural roads was being done simultaneously under the Road Network Improvement and Maintenance
Project Feasibility Study (2002). Fieldwork in Pakistan was not possible at the time due to security concerns, and
work in the Pacific islands was deemed too expensive. The projects in the Mekong region were ongoing and too
recent for discerning the impacts. Nepal, due to its difficult geographic terrain, was considered to be not very
representative of the rest of the Asian landscape.
46 Appendix 1
criteria for selection of both roads and study communities were used in all countries to ensure
cross-country comparisons.
F. Practical Difficulties Encountered During the Study
26. Data. A common difficulty across all case study projects was the lack of baseline data.
Monitoring and evaluation components built into project design were often limited to assessing
physical completion rather than capturing the impact on beneficiaries. This restricted the
assessment of impacts and their quantification in an economic sense. Recall techniques were
not sufficiently reliable in quantifying the economic status of village activities prior to the
construction of the road. The contribution of key data indicators was overestimated, and many
indicators were simply not collected regularly and monitored by provincial or district level
authorities. Consequently, this component of the methodology was only partly completed.
27. Control Sites. Finding good control areas against which to compare project and
nonproject cases proved difficult to different degrees. In practice, there were very few locations
that could be described as lacking completely any form of road access. Hence, the study team
was forced to identify control sites as being at least 45 minutes travel time away from the
nearest available form of public transport, rather than lacking completely any form of road
access. In Yogyakarta, the study team adopted other access constraints to define the suitability
of a control site. In Kurunegala, the sites did not have a sufficient population, so the study team
adopted a full census of the control area or used more than one control site. A lot of effort was
also spent on finding areas that were similar (especially in terms of agricultural potential) in most
ways to the project site study area, except for the lack of a road. As such, the selected areas
proved to be suitable as control sites with minimum selection bias and, hence, did not
compromise the quality of the findings.
28. Locational Difficulties. In most locations, the participation of local officials was
excellent and this was a key factor in the successful implementation of the study. However, in
some instances, there was variation in the interest of these officials driven by local elections,
staff transfers, and the length of time that had elapsed since project implementation. In the
Philippines, the law and order situation in the study sites imposed a considerable constraint on
the study team’s activities in the sites. This prevented the enumerators from freely observing the
household conditions of the respondents in one location. In other sites, there were various
logistical problems (power failures, lack of accommodation, etc.). Despite this, the quality of the
data collected was very good due to the dedication of the study team who spent long hours in
trying conditions in the field during the day and many hours discussing, validating, and recording
findings in the evenings.
Appendix 2 47
SAMPLE FRAMEWORK FOR DOCUMENTING FIELD RESEARCH1
Condition Field Observation Impact Indicator + or - Link to Transport

A. Structural Conditions
Geography Type of geographic conditions Physical difficulties Sets the conditions for the type
and terrain observed associated with movement of transport which prevails or is
of people and goods possible
Levels of air and noise Influences the cost of improving
pollution transport infrastructure
Degree of soil erosion and Road construction and road itself
slope instability may bring environmental
degradation through air and
noise pollution, slope instability,
and soil erosion
Location Proximity to marketing centers Connectivity of location to Distance of households from
and centers of administration/ outside locations marketing center, community
information Household travel patterns facilities, capital, coast, roads,
Proximity to coast and to and other infrastructure
natural resources (mineral
deposits, forests, etc.)
Seasonality Degree to which vulnerability Food security through the Lack of all-season access to
increases at different times year community and all-season roads
during the year Household income through Seasonal demand for transport
the year services
Impact of changing climatic
conditions through the year
Population Reduced availability of land— Per capita land distribution Level of effective demand for
Density high incidence of landlessness Rate of population increase transport services and potential
Subsistence agricultural for demand growth in the future
production with no generation Population pressure as a
of surplus—high incidence of restraint to agricultural
food insecurity accumulation through lack of
land
Shocks Population impoverished or Frequency of natural Susceptibility to natural and
infrastructure recently disasters human shocks and disasters
damaged by natural disasters Recent incidences of (floods, famine, drought, and
disasters tsunami)
B. Capital Assets
Natural Size of landholding of Concentration of land Changes in land use through
cultivable land ownership and incidence of greater exploitation of local
Security of land tenure landlessness resources (minerals, forestry),
Distance to nearest available Common property resources and increased competition for
water source Availability of water and land and resources (alienating
Distance to energy source and fuelwood the poor)
nontimber forest products Distance and time to household’s
Quality of available land and productive assets may change
natural resources
Social Levels of access to the means Frequency of trips to Increased contact with other
and opportunity for different marketing and social centers social groups, strengthening
kinds of travel (household Frequency of travel within social capital through better
chores, productive, and social) and outside the immediate opportunities for travel
between and within area More opportunity for social travel
households Number of trips made for at lower cost
Social and cultural restraints different tasks Access to information
for women to travel Travel patterns for different technology, new services, and
Division of household labor household members opportunities
and the availability of time Attendance at public Exposure to social ills, like drugs
1
Hanmer, Lucia, Elizabeth Lovell, Robert Chapman, and Tom Slaymaker. 2000. Poverty and Transport: A Toolkit. Overseas
Development Institute.
Appendix 2 48
Community-based groups and meetings and levels of and prostitution

organizations that are active voting Improved access strengthening
Participation in decision making Levels of different kinds of the hand of exclusionary groups
Scale of seasonal and migration (local mafias)
permanent outmigration
Frequency of travel to visit
friends and relatives
Impact of improved
communications and
opportunities upon local social
structure
Physical Location of markets and Levels of bicycle ownership Improved ability to travel for
efficiency and accessibility of Levels of motorized and water and energy needs
marketing system nonmotorized vehicles on Formal and informal employment
Road maintenance regime, the road generation in construction,
whether or not local Frequency and affordability maintenance, and related service
communities are involved of transport services delivery
Provision of other Road condition Better communications service
infrastructure services to the Degree to which road is delivery (post, telephone, etc.)
location (electricity, water, and regularly maintained Improved access to seeds,
communications) Infrastructure provision fertilizers, and agricultural inputs
(number of telephones, Quality of road versus level of
electricity connections, etc.) investment and perceived
benefit. Is it good value?
Availability and affordability of
transport services on the road
Financial Availability of employment Existence of local industries Opportunities for outside
(seasonal or regular) and other labor opportunities employment
Prices for agricultural inputs Local craft production for Reduced transport costs
State of banking and formal sale (operation and maintenance) for
credit facilities and credit/ Level of remittances operators, and passed on (or
lending environment for the sent/received not) to users
poor Commercial environment for Reduced travel times and
Household dependency upon transport operators increased levels of service
remittances Improved transport, increased
Affordability and availability of labor productivity
local transport services Decreased farm input costs and
produce prices
Improved income, improved
access to markets, credit and
savings facilities
Location of credit facilities
Human Availability of teachers and Days teacher/doctor in Improved access to health care
educational materials residence and to education and extension
Availability of qualified health School and health center services
professional staff, medicines attendance Poor road and services limits
Frequency of health and Visits by extension staff ability to reach education/health
education staff’s attendance in Effort expended on travel services
the location (gender disaggregated) Lack of education/health
Incidences of disease and Comparison of women’s and infrastructure facilities and the
state of health of population men’s labor tasks, and means of transport to reach them
Level of education and skills transport needs Susceptibility to spread of
training Number and frequency of HIV/AIDS
Availability of time road accidents Increased risk of road accidents
Increased time available for the
poor
C. Institutions and Processes
National Competitiveness of local National laws, regulations, Transport sector tariffs and
Government agricultural produce in the and guiding documents for regulatory framework for trade
(including l market place rural transport sector Environment for transport
Appendix 2 49
laws and Supply of specialized Import tariffs on bicycles operators

regulations) agricultural produce from the Regulatory framework for Linkage of area to national
location transport operators markets
Impact of government support National programs for Contractual arrangements for
programs for the poor microcredit and support to infrastructure tendering and
the poor maintenance regimes
Road safety laws Spending priorities on pro-poor
transport
Regional Participation of local people in Frequency of Effectiveness of local level
and Local decision making representative’s visits to the infrastructure delivery and
Government Participation of local people in area, knowledge of local maintenance
(including infrastructure and services people on who their Provision of local “social”
laws and planning, provision, and representatives are transport services (i.e., farm to
regulations) maintenance Frequency of extension market)
Availability of extension service delivery Linkage of local people into the
workers and effectiveness of Participation in meetings, market
dissemination activities voting Linkage of communities into
Effectiveness of local Effectiveness of interaction decision-making bodies and
government development between local government processes
activities agencies and NGOs/CBOs Provision of government bodies
Tax and regulation Local government pro-poor support in agricultural and other
Existence of local government programs services
pro-poor programs
Private Affordability and regularity of Number of enterprises Provision of transport services
Sector transport services supplying transport-related Local contractors in partnership
Local participation in services with government, responsible for
enterprises supplying transport- Sales of transport goods and infrastructure delivery and
related services and inputs—bicycles, tires, etc. maintenance
infrastructure maintenance Provision of affordable and
regular transport services
(taxi, bus)
NGOs and Ability of NGOs to visit and NGOs and CBOs active in Interventions to improve
CBOs operate in the area the community and their transport at local level through
Mobilization of CBOs and their activities appropriate technologies
connectivity to other groups in Mobilization of community
the area groups for lobbying and
community-based initiatives for
locally-defined needs
Traditional Influence of traditional leaders Decision-making processes Determine the distribution of
Authorities and village structures on within the community and opportunities and benefits, set
behavior and access to the dissemination of norms for cultural practice
resources information Relationship of locality to the
Relationship of local leaders state
with the state and with
surrounding communities
Cultural Perception of the community Shared practices and beliefs Excluding certain ethnic or
Norms and and different groups within the in the community particular religious groups from transport
Practices community of the outside world to the homogenous group or businesses
area Determine whether women can
travel for trade
Gender Women’s engagement in Women’s access to Women excluded from taking
Relations income generating and social household resources public transport, using bicycles,
activities within, and outside, (money and modes of travel) or trading outside the community
the household Women’s household Women lack time to diversify
Women’s willingness and responsibilities and time to their livelihoods—are time and
motivation to diversify their engage in trading and social energy deficient as have a heavy
livelihoods activities burden of household work
CBO = community-based organization, NGO = nongovernment organization.
50 Appendix 3
POVERTY DETAILS
1. Chapter III describes the essential characteristics to understand the poverty situation in
the case study areas and the factors that influence the extent to which rural roads can
contribute to poverty reduction. This appendix shows how the study dealt with the issue of
poverty in capturing its multidimensional aspects, as well as its relativeness across communities
and countries. The community definitions of poverty in each location, as well as key poverty
themes common to all locations, are also presented in this appendix.
2. It was important for the study to try to capture processes of impoverishment. Poverty
appears to be not simply the lack of one asset or entitlement, but a juxtaposition of a number of
different pressures and vulnerabilities, and is a dynamic process. It is often the combination of a
number of different factors that locks households into downward poverty spirals from which it is
increasingly difficult to escape: “Poverty never results from the lack of one thing, but from many
interlocking factors that cluster in poor people’s experiences and definitions of poverty.”1 Just as
poverty is complex and multidimensional, the study has been concerned with capturing what
responses are necessary to address these downward spirals, and what role roads play, either in
isolation or in combination with other factors, in reducing susceptibility to poverty. In particular,
the study team has sought to understand what role roads play in a successful strategy for
escaping from complex poverty situations. Figure A3.1: Relationship of Livelihood Assets to
Structural Conditions, Institutions, and
Processes at Work
A. Who Are the Poor?
STRUCTURAL CONDITIONS
3. Definition. Poverty is here defined as CAPITAL ASSETS

a deprivation in assets and entitlements Natural
essential to life, and a susceptibility to periodic Human

Social
physical and economic shocks, and seasonal
crises. It also describes a state of PROCESSES PROCESSES
Physical Financial
voicelessness and an inability to influence the
structures, institutions, and processes, which Livelihood Livelihood
shape rural livelihoods. Fieldwork revolved Outcomes Strategies
around the analysis of livelihoods through

using a capital assets-based approach.2 It was INSTITUTIONS
assumed that people had five principal types

of capital assets, and the relative strength and Source: Carney, Diana, ed. 1998. Sustainable Rural Livelihoods:
weakness of these assets determines What Contribution Can We Make? Department for
International Development, United Kingdom.
vulnerability, susceptibility to shocks, and
ability to participate in wider institutions and
processes. The relationship of the assets (physical, financial, natural, human, and social) to the
institutions, structures, and processes at work is represented in Figure A3.1.
4. Poverty in the Country Context. As mentioned, the study does not rely on income
poverty line definitions but uses poverty as a multidimensional concept. Therefore, the concept
is not directly comparable with the income and headcount definition of poverty. This
multidimensional definition for the poverty study was operationally defined by the communities
themselves. Within a community, the socioeconomic groups were classified as very poor
(structural poor), poor (transient poor), and better off (nonpoor) in a variety of ways and showed
1
Narayan, D. quoted in Lucia Hanmer, Elizabeth Lovell, Robert Chapman, and Tom Slaymaker. 2000. Poverty and
Transport: A Toolkit. Overseas Development Institute.
2
Carney, Diana, ed. 1998. Sustainable Rural Livelihoods: What Contribution Can We Make? Department for
International Development, United Kingdom. See also footnote 1.
Appendix 3 51
three robust groups that were used in the analysis (Main Report, para. 16). However, the areas
selected for the study were those with a high incidence of poverty because of the poverty focus
of the study. As such, these areas are not representative of the country in general. Although in
this context it may be desirable to understand how these groups relate to the national poverty
definition (particularly to see whether the better-off groups in these locations would be
considered poor under the national definition), it is not correct to compare these different
definitions. There are no directly comparable data available for this comparison within the study.
However, some idea of this comparison can be considered by looking at the proportion of
socioeconomic groups using both definitions. In all project and control areas, the better-off
proportion of the community was very small, about a third of the community. The nonpoor
percentage above the poverty line in each of the three countries (top 65% in the Philippines,3
60% in Sri Lanka, and 77% in Indonesia) averages about two thirds. Therefore, even in these
high poverty incidence areas, those who belonged to the better-off group could likely be
classified as nonpoor in a national sense.
5. Although two types of definitions are used, Table A3.1 shows that the better off in each
location is a much smaller subgroup than the nonpoor group defined in a national sense.
Table A3.1: Percentage of Better Off in the Selected Communities and the Nonpoor
Proportion in the Country
Location Better-Off Proportion According Nonpoor Proportion According to

to Community Definitions (%) National Poverty Line (%)
PHI-Sorsogon Project Site 30 65

PHI-Sorsogon Control Site 20 65
PHI-Negros Project Site 40 65
PHI-Negros Control Site 10 65
SRI-Kurunegala Project Site 20 60
SRI-Kurunegala Control Site 20 60
SRI-Matara Project Site 15 60
SRI-Matara Control Site 10 60
INO-Bengkulu Project Site 25 77
INO-Bengkulu Control Site 15 77
INO-Yogyakarta Project Site 20 77
INO-Yogyakarta Control Site 30 77
INO = Indonesia, PHI = Philippines, SRI = Sri Lanka.

Sources: Study data and secondary data from National Statistics Office (PHI); based on data from Central Bank of
Sri Lanka/Department of Census and Statistics (SRI); and Statistical Yearbook of Indonesia (1999).
6. Relative Poverty Across Communities. Material designation of poverty can be

different in different locations of the same country. For example, in the study communities of
Matara and Kurunegala in Sri Lanka, the material designation of poverty was very different. In
Matara, for example, the community has a higher level of integration into the cash economy and
the relative level of material goods owned, even for poorer members of the community, is higher
than in Kurunegala, a more remote and isolated area lacking this integration into a cash
economy. Poorer groups in Matara, however, are susceptible to other poverty dynamics that
may result from this greater degree of integration and closer proximity to markets, which are not
present in Kurunegala, such as fluctuations in cash crop prices and delays in payment of
wages. Indicators of poverty in both locations would be different, as isolation and integration into
3
In the case of the Negros project site, the community-defined better-off group was 40% compared to the national
nonpoor group of 65%. This is because skilled workers and professionals lived in the area because of their work
assignments and they would naturally qualify as nonpoor in a national sense.
52 Appendix 3
wider networks themselves bring different poverty dimensions into play. The situation in the two
study areas of Indonesia is similar (Appendix 7) with differences in poverty in a relative sense.
7. Poverty Within the Community. Because of difficulties in establishing absolute income

and material benchmarks of poverty across very different locations, the study instead sought to
understand how communities themselves define and understand socioeconomic status in
relation to each other. This relational approach was also intended to capture better some of the
complexity and dynamism of the broader aspects of vulnerability present in each location, which
are of critical interest to the study in seeing what the necessary conditions are for the poor to
make use of rural roads. Lower income and vulnerable groups in each study location have
different coping strategies to deal with deficits and challenges they face.
8. The poverty groups can be broken down into the following three distinct groups:4
(i) The “structural” poor or very poor.5 The members of this group face structural,
long-term factors of lack of access and entitlements to land and resources, which
severely restrict their ability to produce beyond subsistence levels. They are most
susceptible to crises resulting from natural disasters, as they have a small asset
base and few networks on which they can rely for support. Often, families with a
high care factor (where members are elderly, sick, or disabled) are heavily
represented in this category.
(ii) The “transient” poor or poor.6 This is a broad categorization covering a range
of people who are extremely vulnerable to shocks and fall into “severe”7 poverty
at times, but who also have some asset base or entitlements, which gives them,
potentially at least, the opportunity to move out of poverty. Many may slip into
severe poverty through seasonal vulnerability. Others in the top end of this
bracket are progressively securing themselves against poverty by building up
their asset base.
(iii) The better off.8 These households have an income and assets beyond their
subsistence needs and can be categorized as having diverse livelihood options,
wide networks often extending beyond the immediate community, and a strong
and diverse asset base.
9. Through the participatory study activities in each location, villagers grouped themselves
according to perceived poverty status and identified what factors were important in establishing
this status. The results are shown in tabular form in Table A3.2 for each project and control
location. These participatory poverty classifications and definitions were supplemented by data
in the household survey in two ways. First, all respondents were asked the question “How would
you rate your quality of life compared to that of most of your neighbors in this village?” Study
enumerators were trained to facilitate a response to this answer by getting household
respondents to consider a broad range of factors, including material assets, income,
4
The self-perception of “quality of life” classification has been used in the report when analyzing the survey data.
This proved to be a straightforward measure and as effective index in evaluating poverty. The three “quality of life”
strata conform very closely to combinations of other variables that have been used to verify this classification of
groups.
5
Ranged 10–40% of households according to community definition.
6
7
A state where households are forced to use what little assets they have to borrow in order to meet their immediate
consumption needs.
8
Appendix 3 53
consumption, health, and feelings of well-being.9 Respondents’ classifications were better than
neighbors (15%), worse than neighbors (34%), and about the same (51%).
10. The information within the survey was then analyzed using a number of combinations of
factors to try and build a representative “poverty index” which could be applied across all three
countries. Given the different socioeconomic context and livelihood strategies in each of the
areas and the different levels of general development in each, this was necessarily a subjective
exercise. Nevertheless, by these criteria, it was found that household survey responses could,
in fact, be banded into three broad groups and that the distribution of each group corresponded
quite closely to the levels of self-perception articulated in response to the survey “quality of life”
question and the participatory research findings.10 Through three different sources, then, the
classification of poverty groups and their approximate size were validated.
9
The survey captured how many households were receiving state assistance in some form. This was not a useful
indicator of poverty because the level of households receiving assistance was very high (71%). Fieldwork clearly
indicated that qualification for this state assistance in Indonesia and Sri Lanka was often determined by political
allegiance and party membership, as much as any assessment based upon need.
10
The variables used in this exercise were percentage of household income from agricultural labor, percentage of
household income spent on food, status of land ownership and size of landholding, months of food insecurity, and
materials used in construction of house.
Table A3.2: Characteristics of Socioeconomic Groups in Each Case Study Area
Richa
54
Very Poor Poor Better Off
Sorsogon Project Site
10%: The absolutely, chronically poor, 60%: High levels of gradation within this group from the 30%: Have a significant current
Appendix 3
with no land and solely reliant upon the seasonally vulnerable to those who are now fairly accumulation of capital, and have
income from daily wage labor. secure. All have some form of livelihood enhancement the potential and opportunity to
potential such as engaging in small craftwork or selling accumulate more. This group has
forest products. Many still rely upon wage labor, but have a high degree of land ownership
regular employment with landowners and so have a and access to income and
guaranteed income source. employment opportunities outside
the immediate area.
Sorsogon Control Site

10%: Similar to the situation in the 70%: As with the project area above in a state of 20%: As above, although the
project area, with this group landless transient poverty, i.e., still extremely vulnerable and number of better-off people is
and overwhelmingly reliant upon wage prone to slipping back into acute poverty, but having smaller, partly because the
labor. some potential for development. concentration of land is more
acute, but also because the
potential for engaging in other
activities is smaller because of the
area’s remoteness.
b
Negros Project Site 60%: Those who have difficulty in meeting even their 35%: Tenant farmers, those in 5%: Large landowners who
daily food requirements throughout the year. Usually business, and fixed income own 80% of the agricultural
landless and work as a laborer in the sugarcane professionals. Skilled workers land of the village, 65% of
haciendas and as general but unskilled laborers such as drivers, mechanics, and which is planted with
elsewhere. carpenters. sugarcane.
Negros Control Site

20%: Landless and largely jobless, with 70%: Access to some land, upon which they plant 10%: Those owning lands or who
only temporary or seasonal laboring staples for subsistence, such as corn and rice. Some are professionals, such as
opportunities. Low level of education of self-employment in other livelihood activities. teachers living in the area as part
household heads and large family size. of their work assignment.
Kurunegala Project Site

30%: Those who are landless and 50%: The broadly poor who are midway between being 10%: Comfortable through having 10%: Having a good
totally dependent upon wage labor for acutely poor and comfortable. a fixed source of income or standard of living, primarily
subsistence. salaried employment. owning a lot of land in the
village.
Kurunegala Control Site
Control area is made up of two locations due to low population in each: Walahinikalla and Wembuwa.c
a
Some communities defined a fourth group (rich) although a few or none in their community belonged to this group.
b
Classification of the poor in Negros project site is slightly different because of the structure of agro-industrialization in the area; there is a large wage-labor base
working in sugar plantations. The socioeconomic groups identified by villagers are those who labor (poor); those who have some alternative source of income (better
off); and the landowning rich. The category of poor includes those who, in other study areas, have been identified as poor and very poor.
c
Walahinikalla control area is an extension of the Kurunegala project site and the same classifications, therefore, apply. In the Wembuwa control area, less time was
spent and poverty ranking by community was not completed. The self-perceptions and poverty index from household surveys show that the categories are broadly
similar, with approximately 40% as very poor, 50% poor, and a small group (10%) as being better off.
Very Poor Poor Better Off Rich
Matara Project Site

40%: Solely dependent upon wage 45%: Also dependent upon wage labor to a large degree, 15%: State sector employees,
labor and with a low bargaining but with more consolidated security in the form of some those with some form of
position, so often having to wait for full temporary or seasonal income alternative or a small enterprise or owning more than 2
payment. piece of land. acres of land.
Matara Control Site
30%: The very poor who are solely 60%: The broad swathe of villages whose condition is 10%: Classified as economically
dependent upon wage labor and, described as average, i.e., not comfortable or secure, but stable, primarily through owning
therefore, prone to income insecurity. not totally insecure. tea lands of more than 2–3 acres.
Bengkulu Project Site
15%: Have no land and no permanent 60%: Do not have permanent income sources. If they 25%: They have a permanent job, 0%:d Their income is more
job. They work casually as daily have land, the landholding is limited. Monthly income is they have enough to eat, and they than they need. They have
laborers. Have a very poor or insufficient to cover living expenses for a month. Have have no debt. a good house and property.
temporary house. They do not have inadequate housing. They have no debt, and
anything. they have savings. They
have authority in the
community and people
listen to them.
Bengkulu Control Site
25%: Household members suffering 60%: Insufficient money for the family’s daily 15%: Sufficient food and 0%: Have a motorcycle or
from a physical handicap. An acute consumption needs and lack of food. Bad house with possessions to lead a simple life, car and a good house. Have
lack of funds and a lack of clothing. bamboo floors and wall. enough money for daily family a lot of land and savings in
needs. Have a bicycle for the bank. Have gold for
transportation or money to buy a wives and daughters and
bicycle, and enough money to money to educate children
provide for the children’s daily at university.
needs.
Yogyakarta Project Site
20%: No land for farming and no cattle. 60%: Have less than 0.5 ha of farmland, and most of the 15%: Have approximately 1 ha of 5%: Have more than 2.5 ha
House has bamboo walls and an earth land they have is hilly. No cow or bicycle. Main farmland, either paddy or of farmland, more than two
floor. No material possessions like a occupation is farm laboring. plantation land. Have a well- cows, and often more than
radio, and very rarely make a trip constructed house, a cow, and one good house. Have
outside the village. maybe some goats. Also have a motorcycle, car, and a
motorcycle and television. colored television. Have
Appendix 3
jewelry and a plantation.
Yogyakarta Control Site
25%: Have a very basic house, no 45%: Have land of approximately 0.25 ha. May have 20%: Have farmland of 10%: Have more than 1 ha
land, and no cattle. No bicycle or radio, chickens and a goat, have a badly constructed house, no approximately 0.75 ha. Have a of farmland. Many animals,
and no petromax lantern. petromax lantern, and no bicycle or radio. cow, chickens and goats, a well- a good house, radio,
constructed, permanent house, television, and a
and a radio and bicycle. motorcycle.
55
ha = hectare.
d
No villagers in the study area were deemed to fall into this category, but community was clear about what distinguishes this group.
Source: From participatory rural assessment poverty ranking and definitions activities and village social mapping.
56 Appendix 3
B. Key Poverty Themes Common to All Study Locations
11. Research activities in all study locations revealed a number of key common themes,
which condition household susceptibility to poverty. These are summarized in the following
sections.
1. Resource Endowments
12. Land and Labor. The ability to command access over natural resources is critical in
determining whether or not a household is poor. This is the case across all the study
communities. Without any form of land, it is clearly difficult for households to “graduate” from
poverty, as they have no capital other than their labor. Without land, most of the productive time
is spent in wage labor to meet the subsistence needs of the household, and there are few
opportunities to accumulate savings as rates for wage labor are at subsistence levels
throughout the study locations. The situation is compounded where the educational level of
household adults is low, as the potential then for engaging in more lucrative, skilled employment
is also reduced. A lack of land means that the poor are unlikely to have much to trade or sell
outside the community and are, therefore, less likely to use a road, no matter what its condition.
13. Among all survey respondents11 in all case study areas (project and control sites), 26%
of respondents do not farm any land whatsoever. Figure A3.2 shows the percentage distribution
of farming activities among survey respondents. Although 74% of household farm, either their
own land or land belonging to others through tenancy, a large majority (78%) also work as
agricultural laborers at some time during the agricultural cycle. This is because the farmers are
working extremely small plots of land that are barely sufficient to meet their needs. The size
distribution of landholdings among all survey respondents can be seen in Figure A3.3. It shows
that 40% farm up to 0.5 hectare (ha) and 60% up to 1 ha. Among the very poor (50%) farm, less
than 0.5 ha and 75% farm less than 1 ha.
Figure A3.2: Distribution of Farming Activities Figure A3.3: Size of Landholdings
10.00+ ha
26 3.00-10.00 ha 1%
9%
up to 0.25 ha
2.00-3.00 ha 26%
47 8%
2
5
1.00-2.00 ha
22%
19 0.25-0.50 ha
14%
0.50-1.00 ha
Do not farm any land Squat on private land
20%
Farm under traditional rights Tenant farmers
Owners
Source: Household survey data. Source: Household survey data.
11
The terminology “all survey respondents” is used to reflect all project and control site groups in all three countries.
Appendix 3 57
14. Livestock. Animals, are an important source of livelihood security in all of the project
locations, and are also a principle source of income in some areas. Ownership of livestock is,
therefore, a critical socioeconomic indicator. For the poor and very poor, small animals act as
both savings and investments and are usually sold to meet a short-term need for capital to
overcome some difficulty or crisis, such as the costs for medical treatment of a sick family
member.
15. Forest Products. For the poor and very poor, forest products are key assets because
they usually lack land upon which to grow cash crops. They are gathered from vacant lands and
forests in or outside the village, and often require considerable time and effort for collection.
Forest products are more frequently collected in control locations (26% of households) than
project locations (18%), and are sold primarily in the market (34%), to neighbors or shopkeepers
(19%) and to visiting buyers (16%). Forest products collected include firewood, green-leaf food
plants, herbs and medicinal plants, bamboo, mushrooms, and raw materials for making baskets
and other craftwork. Seasonal crops like mushrooms are also sometimes collected.
2. Debt Cycles
16. The level of household indebtedness is a key poverty indicator. Many poorer farmers
and wage laborers in the study communities are heavily reliant upon credit extended by
agricultural middlemen and moneylenders in the community. Tenants and poor farmers borrow
money against the next season’s crop, at times of the year when they are most vulnerable.
Middlemen consequently dictate the price, at which this debt is relieved, and often the
households cannot pay off the debt from one crop. Hence, a debt cycle develops. For
middlemen, this system ensures their supply and enables them to dictate the price at which they
buy. Indebtedness also compels the poor to sell any surplus they have immediately upon
harvesting, at a time when the price is likely to be at its lowest point. Those dependent upon
wage labor also get credit from large landowners for whom they work regularly as a means of
overcoming periodic crisis and food shortfalls during the off-peak agricultural season. Credit
from landowners to laborers ensures a labor obligation in the future when they will most need it,
during peak harvest times, when labor might otherwise be difficult to get. Formal credit sources
have numerous requirements usually outside of the scope of the poor to meet. They also
require a degree of literacy; and the credit outlets are located at centers away from the village.
In contrast, middlemen operate in the village and offer immediate credit at any time. High levels
of indebtedness and chronic debt cycles mean that the poor have no capital with which to invest
in the new opportunities that better roads may bring. It also means that they have no opportunity
or incentive for exploring further markets, where prices for crops may be better, because they
are “locked” into relationships of debt from which they cannot easily escape.
3. Family and Community
17. Household Size and Composition. In all of the study locations, there appears to be a
high correlation between large family size and poverty. Whether having a large number of
children is a symptom or cause of poverty is a moot point, but is probably both. Children offer a
potential source of future livelihood security for poor families, but also require a high level of
care in terms of income, resources, and time. When the care burden is heavy, productive
members of the household are forced to stay at home to engage in care tasks, and the potential
labor base of the household is, therefore, reduced.
18. Social Networks and Longevity of Residence. Close relationships of family and group
are features of rural societies throughout Asia, and all of the study locations clearly show how
58 Appendix 3
important, complex, and dense these networks can be. The social capital that these
relationships embody is the critical glue for communities and shapes many aspects of rural
livelihoods. These connections offer security in times of hardship and are an important social
safety net for the poor, with those better off in the group obliged by tradition to look after the
more vulnerable.12 Family and kinship networks are also important in working productive
resources, like land, and are critical in enabling households to accumulate and to diversify
incomes. It is the lack of access to an immediate family or social support network that really
distinguishes the critically poor from other members of the community. The very poor often lack
these dense social networks because they have recently moved to the area from elsewhere, or
because they do not have immediate family and social ties in the community. Longevity of
residence enables family groups to build up relationships of trust and mutuality with neighbors.
When longevity of residence across a number of generations is combined with available family
labor, this would appear to be a successful remedy in enabling the family group to move from
poverty (Appendix 8, Box 9).
12
In many of the study locations, extended families would live in a “compound” arrangement, with a variety of
standards of housing reflecting the relative affluence or impoverishment of different members of the extended
family group.
Appendix 4 59
STUDY ROADS, PROJECT SITES, AND CONTROL SITES
A. Sorsogon Integrated Area Development Project (SIADP)—PHI
1. Sorsogon-Philippines, Bulan-Magallanes Road. This road segment is 21 kilometers

(km) long, connecting the municipal centers of Bulan and Magallanes. It was rehabilitated and
graveled under the project in 1995, and subsequently asphalted in places using district funds.
Following the improvement in the road surface (still in good condition), there has been a
moderate increase in the volume of 3-wheelers and they have replaced jeepneys, the previously
dominant transport mode. Though the cost of using a 3-wheeler is higher, they are more popular
due to short waiting times. Competition has kept transport prices below the rate of inflation in
gasoline prices in recent years.1 However, the increase in traffic volume has not been as large
as anticipated at project design stage, but villagers in Palale have benefited from more regular
transport and reduced traveling and waiting times.
Sorsogon Study Road—PHI Sorsogon Control Site Access—PHI
2. Sorsogon Study Areas. The Sorsogon project site (Barangay2 Palale) and the control
site (Barangay Bical) had similar socioeconomic profiles. The main production in both is
coconut, with little else existing as a cash crop. The traditional abaca production had failed in
late 1990s due to bunchy-top disease, a problem throughout the Sorsogon area. There is a
noticeable lack of intercropping in both areas, though there is some banana production. Both
communities are consequently heavily dependent upon the price of copra. Much of the copra is
sold to a dealer in the nearby center of San Francisco, midway between Palale and Bulan.
There is a high concentration of land ownership in both the project and control locations, with
the majority of local inhabitants being either tenant farmers of large landowners or agricultural
wage laborers. Households are closely associated through kinship. As in many other areas in
the Philippines, both project and control sites have their own elementary school within the
barangay although facilities are inadequate. In terms of accessibility, the sites are different.
While Palale is on the study road 10 km from Bulan town, Bical is 30–40 minutes walk from the
nearest road, which is a further 3–4 km from the study road. Access between Bical and the
roadhead is by track, with agricultural goods and other larger items transported by buffalos.
Access to Bical is over two rivers without crossings and during the rainy season, these rivers
1
See the Philippine country report for a fuller discussion of this issue in regard to the Bulan-Magallanes Road. ADB.
2002. Philippine Country Report: Impact Evaluation Study of Rural Roads on Poverty Reduction. Manila.
2
A barangay is the smallest administrative unit locality in the Philippines. It is headed by a barangay captain who is
elected by the community for 2–3 years.
Appendix 4 60
flood and the barangay center becomes inaccessible. Palale has a protected tap water source,
but water quality is an acute problem in Bical, particularly during the rainy season, because
open wells and rivers are the main sources of water. Unlike high schoolers in Palale who have
year-round access to the high school on the project road, Bical high schoolers have difficulty of
access during the rainy season.
B. Fifth Road Improvement Project (FRIP)—PHI
3. Negros-Philippines, Moises Padilla-Guinpana’an Road. This segment was part of

72 km of road rehabilitated in 1996 under a countrywide project. The segment is 13.5 km long,
and links the national highway with Canla-on City after rehabilitation. It is now used as a bypass
for intercity buses and large sugarcane trucks, as the surface is better although the route is
longer. Many of the larger vehicles using the road and passing through the project site do not
stop, and are of little value to its inhabitants. Fast speeds, however, impose an increased risk of
road accidents. During harvesting season, large sugarcane trucks service the haciendas, which
predominate in the area. The increased volume of private transport and competition means that
villagers have benefited from reduced transportation costs, both for local travel (motorcycles
and 3-wheelers) and for longer distance travel (buses) from nearby centers. Villagers
overwhelmingly classified the available transport services as “good” in the study passenger
survey.
Negros Study Road—PHI Negros Control Site Access—PHI
4. Negros Study Areas. The Negros project site (Barangay Magallon Cadre) and control
site (Barangay Macagahay) are similar in their sociocultural backgrounds and asset holdings.
However, accessibility constraints resulted in their agricultural production being different. Project
site production is overwhelmingly sugarcane. A large proportion of the study community works
in the haciendas, providing seasonal labor to large absentee landowners of these estates during
the peak season (October to February); at other times, employment is scarce. The dominant
form of production in the control site (Macagahay) is corn for subsistence food needs.
Opportunities are sparse due to the poor terrain and long distance to the road and other
settlements. Sugarcane is not grown due to the absence of a road network. There are schools
in both areas and a health center in the project site. In terms of accessibility, the sites were quite
different. Magallon Cadre is located 3 km from the national road junction along the project road,
and the community enjoys ready access to the nearby town center where there is a market. The
control site is separated from the road by a large river. It is accessible over a footbridge or by
boat. The center of the barangay is a further 45-minute walk through steep terrain. During the
rainy season, the control site is isolated and travel becomes dangerous. Its nearest market
Appendix 4 61
center is on the other side of the river where the control site villagers buy and sell commodities
during market days and access the health center.
C. North Western Province Water Resources Development Project (NWP-WRDP)—

SRI
5. Kurunegala-Sri Lanka, Ambale-Inginimitiya Road. This road was rehabilitated as part

of a wider irrigation project to improve irrigation tank facilities and market access for farmers in
the area through better roads. The study road segment is 16.5 km long, running from the
junction of the national highway to the Inginimitiya irrigation tank. The section through the study
community was completed in December 1998. A small bridge reconstructed under the project
has been particularly important in guaranteeing all-weather access along the link. The road
section is in good condition, perhaps due to the severe lack of rain over the past 3 years. Due to
frequent overspilling from irrigation tanks, accessibility was difficult before rehabilitation.
Bicycles are the main means of transport along the road with a few 3-wheelers, which come
only on request. The main public form of transport is the state bus service and by hand tractors
when needed. The road rehabilitation has improved the regularity with which the bus comes, as
it is no longer hampered by seasonal access difficulties. Motorcycle and bicycle traders collect
fish and milk from the villages and transport them to nearby areas.
Kurunegala Study Road—SRI Kurunegala Control Site Access—SRI
6. Kurunegala Study Areas. The project site (Nugannoruwa) and control sites
(Walahinikalla and Wembuwa)3 are similar in socioeconomic profiles. Rice production is
dominant. Other crops are grown in traditional drylands and the communities increasingly rely
upon cattle as a source of livelihood due to a water shortage that has prevailed over the last 2–3
years. In both sites, most of the paddy fields are dry, with just one area fed by a tank with
sufficient water for one crop per year. Production potential is somewhat better in this control site
than in the project site because of greater water availability in the control site, which is why
people continue to live there despite accessibility constraints. The people in the other control
site Wembuwa moved to the area 15 years ago. There are strong social ties of marriage and
kinship within each village. The school used by all the sites is in an adjacent village, and the
health facilities are at the nearby town of Nikaweratiya. Access to the project site and control
sites is quite different. The project site village is adjacent to the study road halfway between the
provincial road and the irrigation tank. Both control communities are, however, more than
3
Due to low density of population in the control sites, two sites were selected to allow a sufficient number of
households to participate in the survey (Appendix 6).
Appendix 4 62
45 minutes from a bus route on the study road. Despite this, control site villagers have to access
the same community hall, schooling, and health facilities when needed, but must first access the
road via the project site. Wembuwa, the other control site, lies at the end of a motorable track,
which then branches off into a series of footpaths. The main access to the study road is across
the fields and along a nearby tank bund.
D. Southern Provincial Roads Improvement Project (SPRIP)—SRI
7. Matara-Sri Lanka, Urubokka-Katuwana Road. The study road segment is 10.5 km

long, and runs between two major parallel transportation routes. The existing gravel road was
poorly maintained and had become virtually impassable prior to rehabilitation. The road was
rehabilitated under SPRIP in 2001, and is currently in good condition. There is evidence of a
substantial increase in traffic volume along the road, and of roadside development (green-leaf
tea collection centers, small stores). The much heavier volume of traffic here (than on
Kurunegala project road) is partly because the road now serves as a bypass between the two
parallel road networks, with larger trucks and goods vehicles taking advantage of the better
surface. The villagers prefer motorcycles to bicycles because of the steep gradient. The high
integration of this area into the cash economy means that some households have motorcycles.
A bus and a minibus offer public transportation services. A new state bus service is also
planned. Modified trucks transport tea estate workers to the estates. Since the rehabilitation, the
transportation price on 3-wheelers has fallen by half due to a better road surface and increased
competition.
Matara Study Road—SRI Matara Control Site Access—SRI
8. Matara Study Areas. Both project site and control sites are very similar in their
socioeconomic profile with tea as the main product in both project (Heegoda) and control sites
(Makiliyathanne). There is some spice production. Conditions are good for lowland green-tea
leaf production, and there is a high level of commercialization and integration into the cash
economy. Villagers are either smallholders of land or work in the commercial tea-growing
plantations. The poor in both areas work in the tea plantations. There is a fluidity in the concept
of the village here in both sites, with few villagers sure of which village they actually belong to.
There are no community organizations in either site though there are strong social and family
networks in operation. The nearest school to both areas is on the main road through Heegoda.
Accessibility though is very different in these locations. While the project site is adjacent to the
road, the control community lives on the upper slopes on either side of the study road and is
accessible by long winding stone paths, which cut across the slopes. The villagers in
Makiliyathanne are very conscious of the difficulties they face without road access; they use the
Appendix 4 63
same services and institutions as the project site villagers, but feel the benefits of the road
beneath are passing them by. Due to the steep track access, everything has to be headloaded,
because even tractors get stuck on their access path.
E. Tree Crop Smallholder Sector Project (TCSSP)—INO
9. Bengkulu-Indonesia, Talang Padang-Tanah Abang Road. The TCSSP project built

basic, narrow, graveled plantation roads to better access farmers’ rubber plantations and rice
fields, which were newly established under the project. The project road was established in
1993, and was built by a farmers’ group, and initially at least, maintained by them. However, this
group stopped maintaining the road after a few years, as the members felt that the benefits
received from these types of plantation roads were not commensurate with their efforts in
maintaining them. Their decision also coincided with the decline in intercropping as the rubber
trees matured. The plantation roads have now reverted to basic access tracks. To be
comparable with other communities studied, the community adjacent to a plantation road and
the asphalt road (16 km) linking the plantation road to the main road was considered as the
project site. This link road has a close connection to the TCSSP plantation roads. It was
asphalted by the provincial government in 1996 in response to farmers’ requests for a better
road to complement the TCSSP project activities. The area is remote and much of the travel
local. Since the asphalting, a regular bus and minibus services run to Bengkulu, the provincial
capital. The speed of travel and thus time saved have also increased markedly since
rehabilitation.
Bengkulu Study Road—INO Bengkulu Control Site Access—INO
10. Bengkulu Study Areas. Villagers in both the project site (Talaatan) and the control site
(Talang Padang) are smallholders of land and the poorer families also work as agricultural
laborers, fisherfolk, take up basket weaving, or collect forest products. Crops in both sites
include paddy, corn, cassava, sweet potatoes, groundnut, and soybean. Tree crops include
rubber, coffee, coconut, jackfruit, and durian. Accessibility is very different in the two locations.
Talaatan is next to the district road, and the villagers there use public transport to go to the
weekly market or to the provincial capital. The control site is completely inaccessible by road as
it lies on an island surrounded on all sides by a river. Access is either across the river or via a
small suspension bridge. During the wet season, the river can swell and wading across is not
possible. The village extends to the banks on the other side, and access from there to the
Appendix 4 64
nearest road is along a track and then on through farmland to the road. Although the road is
motorable by 4-wheel drive during the dry season only, the conditions are very difficult and there
are no transport operators servicing the village.
F. Third Local Roads Project (TLRP)—INO
11. Yogyakarta-Indonesia, Semanu-Giri Panggung Road. The TLRP rehabilitated

10.5 km of the 22 km stretch of this road, which runs from the municipal capital to another
arterial road in the south. The road passes through agricultural lands, and the study community
at Candirejo is at approximately 4 km along the road. There is a higher level of motorized
transport on this route compared to the other study areas, minibuses being the predominant
mode. Bicycle use is very low, perhaps due to the terrain. Since the rehabilitation of the road, a
high frequency of transport services has developed, with a consequent decrease in waiting
times. Decreased traveling times are also evident.
Yogyakarta Study Road—INO Yogyakarta Control Site Access—INO
12. Yogyakarta Study Areas. The study areas are similar in terms of physical aspects,
geography, topography, agricultural crops, and culture. Both the project site (Candirejo) and the
control site (Kelayu subvillage) are situated in hilly areas, meaning that most of the available
farmland is sloped. The condition of the soil is very poor with a major portion of the land
classified as unfertile. The majority of cultivation in the area is dryland rice and cassava, which
is intercropped with groundnuts and some banana. The rocky land topography in some places
makes agricultural production difficult, and water is a real problem throughout the region. The
main occupations among villagers are a mix of smallholder farmers and farm laborers. Most of
the villagers have “noncertificate” status of land ownership. The accessibility is much different
between the sites. While the study road runs through the project site, the control site community
is 6 km from the asphalt road. The villagers have to walk along a rocky road and pathways to
reach the central village.
Appendix 5 65
CASE STUDY DETAILS IN THE PHILIPPINES
A. Introduction
1. The two road projects selected as case studies were (i) Sorsogon Integrated Area
Development Project (SIADP), and (ii) Fifth Road Improvement Project (FRIP). SIADP is an integrated
area development project with a road component, while FRIP is a road sector project.
1. Sorsogon Integrated Area Development Project
2. Description of SIADP. SIADP is located in Sorsogon Province, Bicol Region, which is one of
the least developed regions in the Philippines with a widespread incidence of poverty covering 78.6%
of its population at project approval in 1988. The project focus was on poverty reduction by generating
employment and improving living standards of subsistence farmers and fishing communities. The
Project included five components: (i) improvement and rehabilitation of 215 kilometers (km) of
provincial and 94 km of farm-to-market roads; (ii) rehabilitation of 15 communal irrigation systems and
construction of flood-control facilities; (iii) health services through schistosomiasis control and
provision for water supply systems; (iv) support services for agriculture and fisheries, including abaca
rehabilitation, plant nurseries, and artificial reefs; and (v) project management and training.
3. SIADP aimed to improve the level of accessibility of its areas, reduce the cost of
transportation, and enhance the distribution and marketing of commodities and transport for humans.
The basic thrust of the project was to strengthen the productive capacity of the poorer population
groups through eliminating major physical constraints and making basic social and economic services
easily accessible to the population. The following criteria were used for prioritizing the road projects:
(i) access to high-density population centers, (ii) access to areas where project production program
were being implemented, (iii) transportation of products to markets, and (iv) access to agricultural and
fisheries production areas.
4. The executing agencies and the subcomponents for which they were responsible are as
follows: (i) Department of Public Works and Highways for road improvement and rehabilitation and
flood control; (ii) National Irrigation Administration for the rehabilitation of communal irrigation
systems; (iii) Department of Health through its district office in Irosin for schistosomiasis control;
(iv) Department of Agriculture for abaca rehabilitation extension support and artificial reefs;
(v) Sorsogon Provincial Governor for domestic water supply; and (vi) Office of the Provincial
Agriculturist for plant nursery establishment. SIADP was completed in December 1997.
5. SIADP Setting. Sorsogon Province is at the southernmost tip of Luzon Island. A main road
artery, the Maharlika Highway, connects the province to the other provinces and to the southeastern
tip of the province, where ferry boats operate. It occupies an area of 214,145 hectares (ha), largely of
volcanic cores interspersed with broad and level farmlands. It has one city and 14 municipalities, 13 of
which are along the coast. There is heavy rainfall from November to January and hardly any dry
season. The 2000–2001 cropped area statistics for Sorsogon shows that about 151,000 ha are
planted with coconut, rice, bananas, root crops, corn, etc. However, despite a large cropped area,
output for key crops fall short of the consumption levels in the area. About 75% of its total population
(0.6 million) are in rural areas.1 The average annual per capita income of Sorsogon in 1997 stood at
P18,032, with a huge variance between the lowest average level (P3,546) against the highest
average level (P101,755). The poverty threshold in Sorsogon for the same year was P7,760.2
1
National Statistics Office. May 2000.
2
This compares with P8,319 for the entire Bicol Region in which Sorsogon Province is located.
66 Appendix 5
The percentage of the population below the poverty threshold in Bicol Region increased from
55.0% in 1997 and 62.2% in 2000 and Sorsogon experienced similar pattern. The unemployment rate
for Sorsogon in 2000–2001 was 18.2%.
6. School participation rates in Sorsogon are high at 97.7% and 70.7%, respectively, for
elementary and secondary levels. Generally, the health situation in Sorsogon improved in terms of
infant and child (<5 years old) mortality rates between 1993 and 1998. About 1% of the population in
2001 was diagnosed as suffering from severe malnutrition, an improvement over the 3% in 1990. The
better health status of the people in the area is attributed to improved health facilities. Inadequate
transport facilities such as roads and bridges to connect production centers to markets remain one of
the major constraints in the overall development of Sorsogon Province. The current state of roads
varies considerably from very good to very bad condition, with most in bad condition. The province
has approximately 1,429 km of roads. The number of vehicles registered in the province increased
from 7,470 units in 1997 to 10,707 units in 1999. The selected Bulan-Magallanes Road connects two
urban centers and the municipalities of Bulan and Magallanes.3
2. Description of the Study Areas in Sorsogon
7. Among the 15 road projects under SIADP in Sorsogon, the Bulan-Magallanes Road was
selected for the study. Selection of these project sites was based on a variety of factors such as the
representativeness of the project area, existence of rural poor, security concerns, and availability of
local enumerators (Appendix 1). Barangays Palale and Bical of Bulan Municipality were selected as
project and control areas.
8. Bulan-Magallanes Road, Sorsogon. The Asian Development Bank (ADB) supported the
rehabilitation of 21.12 km of road out of 26.77 km of total road length from Magallanes to Bulan as the
remaining stretch between Bulan to San Francisco was already paved. Before rehabilitation, the road
was motorable for a stretch of 8.6 km from Magallanes to Siuton during the dry season. At the other
end, the road was passable from Bulan town to Barangay Cadandanan, a stretch of 9.4 km. Between
these points, before rehabilitation, the road was impassable to motor vehicles throughout the year.
Currently, the Bulan-Magallanes Road has a 10 km concrete section from Bulan to Barangay Palale
and carries largely local traffic. It has steep grades and sharp curves along its mountainous section
and flat sections at either end of the road segment with rolling hills in between. The road follows the
ridge alignment. Therefore, the numbers of cross drainage structures are only few and there are
almost no retaining walls. There are six bridges with a span of 15–25 meters (m). The side slopes of
the road are fully covered up by vegetation, therefore, no soil erosion was noticed. The quality of the
study road is still good with the road surface still holding, 7 years after reconstruction, despite heavy
seasonal rainfall. The low volume of traffic and good quality of graveling could be the reasons. It
seems that the road was rehabilitated at high cost and overdesigned for the current level of traffic.
With the shoulders of the road not filled, there is a significant risk of road accidents. After ADB-funded
graveling work, the local government funded the paving of the road, part double lane and part single
lane. The rigid pavement is appropriate for tricycles (motorcycle-drawn sidecars). As a result, the
majority of the tricycles operate on both sections that are paved with reinforced concrete.
3
Municipalities in the Philippines are not necessarily confined to urban areas.
Appendix 5 67
9. Sorsogon Project Site. Barangay Palale is approximately 10 km north of Bulan Municipality

and has one administrative sitio4 called Patag. The main market centers of this village are either
Barangay San Francisco or Bulan town. It has a total land area of 726 ha and a population of 96
households. Palale is highly dependent on copra production volume and its price. Copra is produced
in six cycles spread over the year. The steep decline in copra prices between 1997 and 2000 resulted
in widespread impoverishment and difficulties for the locals. The previously existing abaca production
ended due to disease. Currently, the Government is
encouraging a new crop (pili nut) to minimize
overdependence on coconut. There is an elementary
school in Palale, but the roof leaks severely during rains.
Although Palale has a level 2 water supply (tap in the
village), about 20–30 households are facing difficulties in
accessing water. The village has no market. There are
only two small convenient (sari-sari) stores mainly
carrying the basic food requirements of the households
in the village. Villagers have all-weather access along
the Bulan-Magallanes Road. The barangay road leading
to sitio Patag is no longer passable by motorized
vehicles. Farmers in the village either have to transport
Transporting forest products produce from Patag manually or use a buffalo (carabao)
from Patag to the roadhead to the roadhead. Most tenant farmers are obliged to sell
the copra to their landowners who give them credit in
times of need. Even others who do not have this arrangement still find it hard to sell copra in an
“open” market, as the local trade is controlled by a small number of copra dealers.5 The ability, then,
of local tenant farmers to take advantage of any potential increase in opportunity for trade is severely
restricted by the manner in which they are locked into local relationships, primarily of debt.
10. Sorsogon Control Site. Barangay Bical is one of the most inaccessible settlements of Bulan
Municipality. Village access is via tracks to the roadhead of about 40 minutes walk. From the
roadhead, tricycles can be caught to the main Bulan-Magallanes Road. As a conflict-torn area, Bical
has not historically received many government services. However, in December 2001, the village
received electricity and there was an ongoing road-opening project in the area during the field visit.
Only 14 households had electric connections as access may be beyond the means of most
households. Like Palale, the village economy depends heavily on coconut production. The village has
a total land area of 455 ha and a population of 96 households and an elementary school. The water
supply is level 1 (open source, spring, or river). There are only few middle traders at the road junction.
But most farmers continue to Bulan or San Francisco to market their goods. The dealer in San
Francisco buys a large portion of copra from the area. The village has no public market. There are
only two small sari-sari stores carrying basic food requirements of the households.
3. Fifth Road Improvement Project
11. Description of FRIP. FRIP was approved in 1990. Its objectives were to improve the project
roads to an economically maintained condition, reduce transport efficiency and accessibility
constraints, and assist the Government in improving road maintenance. FRIP comprised (i) civil works
for improvement of about 420 km of national roads and 420 km of rural roads in 11 provinces;
(ii) periodic maintenance of both rural roads in the same 11 provinces; (iii) consulting services; and
(iv) a project benefit monitoring and evaluation study for the Third Road Improvement Project. The
4
A part of a barangay.
5
In addition to the indebtedness, the lack of drying and storage facilities also affects the price that farmers can demand.
68 Appendix 5
EAs were the Department of Public Works and Highways for the national road component and the
Department of Interior and Local Government for the rural road component.
12. FRIP Setting. Among the 11 provinces covered under FRIP, Negros Occidental (West)
Province was selected for the case study. It is located in the northwest portion of Negros Island within
the Western Visayas or Region VI. Negros Occidental has a total land area of 792,607 ha. It has two
pronounced seasons, the wet (June–October) and the dry. Sugarcane is the major crop of the
province, occupying 49% of the cropland area, followed by rice at 28%. The minor crops are corn,
coconut, banana, cassava, and mango. The province had a population of 2.5 million (National
Statistics Office, May 2000) and is expected to reach 2.8 million in 2002, with 52% living in rural
areas. Population density was 284 in 1990 and is expected to reach 361 persons per square kilometer
in 2002. In Negros, the national road junction (NRJ) Moises Padilla-Guinpana-an Road Project, in
Moises Padilla Municipality, was selected as the study road. Its level 3 water system can serve only
7 out of its 15 barangays throughout the year, which are all situated in the town area. Other
barangays are mainly dependent on deep and artesian wells for their water supply. The predominant
dialects of the area are Hiligaynon, Cebuano, and Caray-a.
4. Description of the Study Areas in Negros
13. Moises Padilla-Guinpana-an Road, Negros. The Moises Padilla-Guinpana-an Road, the
study road, is approximately 80 km south of Bacolod City situated in Moises Padilla Municipality. It
begins at an NRJ and proceeds northeast passing three barangays and links to Canla-on City. This
road is paved and is about 13.5 km long. Terrain along the
road is varied, with flatlands starting at the junction and
becoming rolling hills, and then mountainous near the end.
The road project, which was formulated in consultation with
the municipality, followed economic and sociopolitical
criteria set out in the FRIP guidelines. The Department of
Interior and Local Government implemented the project
with the assistance of the Project Engineer’s Office.
Nonetheless, barangay officials of Magallon Cadre, the
study area, were not involved during the formulation of the
project, reflecting a top-down planning approach. The
quality of the project road is still good 6 years after
rehabilitation. The road is underdesigned for the current
traffic load and volume. There is a good supply of transport
Tricycle on study road in vehicles for local needs of villagers in Barangay Magallon
Negros Cadre, especially tricycles, as it is between two markets.
There is high connectivity of barangay roads to the main
roads, but these are in poor condition. The rehabilitation equipment included bulldozers, road rollers,
dump trucks, and graders. All bituminous works were also carried out by capital-intensive technology.
14. Negros Project Site. Among the three barangays along the NRJ Moises Padilla-Guinpana-an
Road, Barangay Magallon Cadre was selected as the study area. It is located 3 km away from Moises
Padilla and 86 km away from Bacolod City. The village is composed of seven sitios. The Intiguiwan
River cuts across the village and is being tapped for irrigation purposes. The local economy is highly
dependent on the sugar industry: sugarcane plantations (haciendas) dominates the landscape of the
village. Magallon Cadre has a total land area of 2,490 ha and a population of 616 households. Out of
its total land area, about 52% is planted with sugarcane followed by rice (27%); social structures such
as school, playground, and houses (11%); and cash crops and trees (10%). The concentration of land
ownership is particularly acute. Most of the inhabitants in the project site are employed in the
haciendas of large landowners, and most of the villagers themselves have no land. Even the lands for
Appendix 5 69
settlements along the road belong to the hacienda. Residents are, therefore, required to work in the
hacienda system for wage labor at peak periods in the agricultural season. Since they have no land
and few other productive assets with which to diversify livelihoods, they become locked into a
continuing relationship with the hacienda. Since haciendas are the primary source of employment for
the villagers, the majority of its workforce are sugarcane workers. A daycare center, two
kindergartens, and an elementary school are available to the villagers. There is a barangay health
center, which mainly provides maternal and child care services. The village has no public market, but
there are about 20 small sari-sari stores to cater to the basic requirements of the households in the
area. Despite ready access, only one third of all the households have electricity connections.
15. Negros Control Site. Barangay Macagahay was selected as a control area. Access to the
village is through Barangay Montilla, another barangay of Moises Padilla Municipality. A big 30 m
wide river, Binalbagan River, separates the village from Barangay Montilla, forcing the villagers to
take a boat or walk along a floating bamboo bridge to cross the river. After taking a boat, it takes
about a 30–45-minute walk for the villagers to reach the village proper, while the other route over the
floating bridge is another 20–30 minutes of extra walk to reach the village center. Barangay
Macagahay has seven sitios. It has a total land area of 2,742 ha and a population of about
335 households. Settlements in Macagahay are sparsely distributed. About 83% of its total area is
considered mountainous. During rains (typhoons), Barangay Macagahay is completely isolated as it is
extremely dangerous for villagers to travel due to difficult terrain. Due to the absence of a road
network, sugarcane is not grown in the village and a
considerable number of bamboo trees also mature
without being harvested every year. The main production
is staple crops: irrigated rice (15 ha), rainfed rice (40 ha),
and corn (300 ha).
16. There is no public market in the village and the

market center in Montilla is the nearest. Since the health
center in the village has inadequate medicine supplies,
villagers go to the adjoining village. Also, the two schools
in the village have to merge classes due to lack of
classrooms and teachers. Extension workers are rarely
visible in the village due to considerable difficulty in
reaching the area. Transporting goods across the
river in the Negros control site
B. Transport Situation
17. Sorsogon Project Site. The Bulan-Magallanes Road is passable throughout the year with the
majority of villagers claiming that the condition of the road is better now than 5 years ago. Before the
road rehabilitation, people also used to travel from Bulan to Magallanes or vice versa by banca
(nonmotorized boat) as an alternative to using the road. People originating from impassable sections
used to hike. Their freight was transported by carabao sledge and manual backloading. With the
improvement of the quality of road, it was anticipated that the use of motorized traffic would increase
in volume. Similarly, the banca traffic would then be diverted to the road. The present volume of
average daily traffic (Appendix 1, Table A1.1) shows a low volume. The existing projection on traffic
data is not directly comparable with the 2002 data.6 Other routes that have opened up since then and
6
This data are not directly comparable to the cordon survey data generated in February 2002, as the locations of the
previous traffic counts are unknown as well as the period over which they were taken. There is also evidence of significant
seasonal fluctuations in the levels of traffic using the roads. During the fieldwork, the traffic counts were done from 8 am to
4 pm instead of 6 am to 6 pm as in the other study locations for security reasons. There were no market days during the 2-
week study period.
70 Appendix 5
the failure of abaca production and the poor copra market may have
affected the expected traffic growth.
18. Only jeepneys and tricycles operate along the Bulan-Magallanes

Road with jeepneys servicing only part of the study road. There is no
fixed time for jeepneys to operate because they wait for about
30 passengers to get in. Tricycles wait for at least seven passengers for
them to operate. The tricycles can carry up to 20 passengers (children)
or the equivalent load of commodities. The frequency of tricycle trips is
higher during morning and late afternoon, and very low during the rest of
the day. Both types of vehicles do not operate if they do not get a
minimum number of passengers or freight. However, there is unsatisfied
demand during peak hours. The area generates a higher volume of trips
during August to December harvesting months. To meet demand,
transport operators add more vehicles during the peak season. The
tricycle fare from Palale to Bulan was P5 before the rehabilitation of the
road but increased to P10 in 2002, about 6 years after the rehabilitation.
Similarly, the jeepney fare of P4 increased to P8.7 The tariff rate did not
A jeepney carrying increase in proportion to the price of gasoline and prices of other tricycle
passengers and spare parts. This could be partly attributed to the level of competition
freight on Bulan- among tricycle and jeepney operators. Before rehabilitation, only two to
Magallanes Road three tricycles were servicing road segments, but now there are 15.
However, the full savings in the vehicle operating costs are not
transmitted to the passengers immediately.
19. Apparently, availability of better road and transport services contributes to the
commercialization of farm production in the village. Almost all of the households that farm in the
village sell their crops. Nevertheless, the majority maintain that they are selling a lower percentage of
their produce now than 5 years ago. This implies that the presence of better road and transport
services are not sufficient conditions for villagers to achieve higher farm productivity. They are selling
their farm products at the market or to local merchants. Most claim that there is no change in the
number of agricultural buyers visiting their village now compared to 5 years ago. The declining pattern
may be because some producers are taking their produce directly to the market and partly because of
a declining market for copra. The villagers’ preferred mode of transporting products to market is
tricycle.
20. All travel movements near and within the village are by foot except for visiting a local healer or
a health facility when adult women use a shared taxi. Adult men predominantly travel locally to visit
fields for crop production, collect water and fuelwood, and process agricultural products. The adult
women’s local travel is for buying provisions for their family’s daily needs, securing medical
assistance, and maintaining good relationships with relatives and friends. Almost all of the travel
movements of villagers outside the village are done by shared tricycle. Adult men travel mainly for
processing agricultural products, selling their harvests and other products, and securing documents
from the state office. Adult women travel primarily for buying provisions for their family’s needs and
securing medical assistance from sources outside the village. Some adult women travel outside their
village for business or employment purposes. Male and female children travel outside the village by
shared taxi mainly for attending school on weekdays.
7
If inflation for tricycles and withdrawal of the gas subsidy are factored in the present tricycle tariff rate, the fare should have
reached more than P14 for tricycles; the jeepney fare should have been more than P11 in 2002.
Appendix 5 71
21. There appears to be no difference in purpose of travel within the village between the different
socioeconomic groups but the mode of travel is different. The poor and very poor travel on foot
whereas the better off use a tricycle sometimes, saving time on routine tasks. Outside the village, the
very poor and poor travel primarily to buy provisions, while the better-off households travel for
processing and marketing agricultural products. Overall, the adult men undertake productive tasks
related to income-generating activities, while the adult women do household tasks except for few who
are engaged in business/employment. The area has a mobile shopping facility that vendors visit once
a week and sell daily food necessities to the poor. There is also a local government mobile clinic
along the Bulan-Magallanes Road to provide emergency medicines and treatment for common
illnesses. The villagers benefit from the easy access from this, and the lower cost of medicine due to
savings in transport cost. Better roads and more available transport services have increased the
mobility of teachers, health workers, and extension workers, allowing them to provide services to the
villagers. It also enables provincial health, education, and agriculture officials to monitor the quality of
service delivery in a more time-efficient way. Likewise, better road and transport services have greatly
contributed to the sociopolitical participation of villagers. In the 2001 local elections, a high voter
turnout of 95% of the voting population is an indication that villagers are aware of their roles in
community management.
22. There is no organized transport association for operators along this route. There are three
institutions directly involved in regulating and licensing the operation of transport services. Issuance of
vehicle registration and licenses is the responsibility of the Land Transportation Office. The Land
Transportation Franchising and Regulatory Board administers the franchising of routes. For jeepneys,
the process involves a court hearing and payment of an application fee of about P3,000. Jeepneys
can operate up to 65 km. The license to operate tricycles is administered by the concerned
municipality. At present, the Provincial Engineer’s Office maintains the road because it is categorized
as a provincial road. It sends a maintenance crew with the required construction equipment.
Therefore, there is no employment generation opportunity. A majority of villagers are not prepared to
contribute to the future maintenance of the road, but those who are willing would prefer to contribute
their labor. National and local authorities feel that the rural road projects should be integrated with
other economic activities. Better extension services and a program of skills development could have
been initiated to create an entrepreneurial environment.
23. Sorsogon Control Site. The only village access

is via tracks to the roadhead of about 40-minutes’ walk
across two rivers. Bical villagers currently carry medical
patients in an abaca stretcher to the nearest road to be
transported to the clinic or hospital. Consequently, home
births are common in Bical. Despite the absence of
better road and transport services, agricultural production
in the village is commercialized. All the farming
households claim that they are selling almost the same
percentage of their produce now as 5 years ago, but
most say more agricultural buyers visit their village now
compared to 5 years ago. As most farming households
have difficulty in transporting their produce, they sell it at Tricycle taxis waiting for
lower prices to agricultural buyers visiting their village. hire at Magallanes
Those selling directly to the market transport their
products on foot or with an animal, and then transport further by tricycle or with an animal.
24. Within the village household, daily travel activities are all done on foot similar to Palale.
Responsibilities for transport are also similar. The shared tricycle is the most common mode of
transport for activities done outside the village. In addition, villagers also transport their agricultural
72 Appendix 5
products to the market to seek medical assistance on carts drawn by buffalos. Because of the lack of
road access, villagers spend too much time moving from one place to another to carry out their
activities. On average, villagers spend about 2 hours to accomplish a task involving travel outside the
village covering a distance of about 20 km. As in Palale, there seems to be no difference in travel
within the village between the very poor, poor, and better-off households. Purpose of travel patterns
across socioeconomic groups are also similar to Palale as are the gender differences in transport
responsibilities. The very poor and poor households travel on foot due to their lack of access to
animals. They spend more time reaching the roadhead compared to the better off. This reduces
substantially the time that the very poor and poor households may spend on productive activities,
adversely affecting their income-earning capacity. In the 2001 local elections, voter turnout was 91%.
A majority of villagers are willing to contribute to the maintenance of a future road through provision of
labor.
25. Negros Project Site. Magallon Cadre is strategically located along the NRJ Moises Padilla-
Guinpana-an Road, the study road. Travel in and out the village is fast, easy, with all-weather access
due to the better road and better transport facilities available to the villagers. But the village roads
leading to its different sitios are in a bad condition. Transport availability is not significantly affected by
changes in season, as several alternative means are available. It is better now than 5 years ago. The
average daily traffic passing through the study road substantially increased from about 300 in 1990 to
nearly 1,000 in 2002. All types of vehicles are seen on this road such as buses, minibuses, jeepneys,
automobiles, tricycles, trucks, tractors, and motorcycles (Appendix 1, Table A1.1). After the project,
the frequency of transport services increased considerably. Motorcycles are the most used mode of
transport followed by tricycles. Trucks are used mainly for transporting sugarcane. The volume of
traffic is high in comparison to the Bulan-Magallanes Road in Sorsogon. Most of the motorized
vehicles originate from Bacolod City and the road appears to be playing the role of a national highway
acting as a better route connecting Bacolod City to San Carlos City and La Castellana to Canla-on
City. The direct route to these cities is in poor condition. Therefore, the vehicle operators prefer to
take long detours to avoid poor roads. Only three 60-seater buses used to operate between Bacolod
City and Moises Padilla passing through the study road before its rehabilitation. They left in the
morning and returned in the afternoon. At present, “Ceres Liner” with reasonably large and
comfortable buses plies the Bacolod to Canla-on City route every half an hour. Other minibuses
operate every half hour from Cabacungan to Bacolod and vice versa. This means that long-distance
transport services are available every 15 minutes along the study road. On the other hand, tricycles
are available as and when necessary. Out of 300 tricycles in Moises Padilla Municipality, one third are
owned by the drivers themselves, and two thirds by nondriver owners.
26. There was a reduction in the vehicle operating costs after rehabilitation of the study road.
However, the nominal tariff rate for passengers did not decrease after rehabilitation of the road but the
real rate did. The bus fare for the Bacolod-Cabacungan section increased from P35 (before
rehabilitation) to P40 in 2002 (6 years after rehabilitation).8 According to government norms, a
transport vendor can raise fares up to P0.90 per km for a passenger. It means the bus operators can
collect up to P85 per person for a road length of 94 km from Bacolod to Cabacungan. Nonetheless,
they are charging only P40. Therefore, the savings in the transport costs are passed on to the
passengers. It appears that availability of better road and transport services contributes to the
commercialization of farm production in Magallon Cadre. All farming households sell some of their
agricultural production. The majority of farming households are selling a higher percentage of their
produce than before, and they have more buyers visiting the village now than 5 years ago. About 82%
of respondents using transport services say they are good and a further 4% report them as fairly
good.
8
If the nominal tariff rate had increased in parity with the gasoline price, it should have been P99.
Appendix 5 73
27. Most villagers buy their household provisions and agricultural inputs from the public market
center in the town of Moises Padilla and sell agricultural products in this market. They normally use
the medical facility available in the town. A minor health treatment facility is available in the village
itself. There are employment opportunities, though limited, in the town. Since Magallon Cadre has
only an elementary and intermediate level educational facility, high school or secondary level students
must travel to the town. Beyond secondary level, the students have to travel to Bacolod City or other
larger cities. Nevertheless, most of the transport demand of the local people is within the municipality,
especially for the poor, and is met by an abundant supply of tricycles. Bacolod City serves as the
regional market center for business people, and the town of Moises Padilla serves as the local market
center.
28. The pattern of local household travel within Magallon Cadre shows that most households
travel on foot as they perform their daily activities. The patterns of household travel outside the village
are also similar in responsibility and task to Sorsogon. On average, outside travel of the villagers
covers a distance of 5 km requiring only 12 minutes. Moreover, the transport fares incurred in
accomplishing the regular activities are relatively stable at P10 per return trip. Travel patterns by
socioeconomic group are also similar to Sorsogon. Within the village, all groups have similar travel
needs. However, most poor households who travel outside the village restrict themselves primarily to
buying provisions to meet subsistence needs, while the better-off households travel for processing
and marketing agricultural products to generate more income. As most of the poor households are
situated far from the road, they have to spend considerable time to access the transport services
along the road. As in Palale, better roads and more available transport services increase the mobility
of teachers, health workers, and extension workers, allowing them to provide efficient and effective
services. There was a 94% voter turnout in the 2001 local elections.
29. The Land Transportation Franchising and Regulatory Board has the authority to define the
transport routes and the fare rates. The Franchise Office of Negros Occidental has determined the
fare rate at P0.90 per km for buses. The Land Transportation Office undertakes the licensing of the
vehicle driver and vehicle registration. The local municipal government administers the licensing of the
tricycles. The registration fee for a tricycle is P300. The issue of traffic accidents falls under the
jurisdiction of the traffic police. In addition to the formal regulations, the large transport operators like
Ceres Liner have developed their own policies on the number and frequency of services according to
the transport demand. The United Negros Drivers and Operators Center also makes decisions on the
queuing system. Since this is classified as a provincial road, the Project Engineer’s Office undertakes
routine and periodic maintenance work. The maintenance activities generate limited employment for
the local people. Considering the volume of traffic it handles, this road functions as a national road.
Maintenance will be a problem for the continuing operation
of the road in the near future, as it was not designed for
the heavy traffic and vehicles carrying loads weighing
20 metric tons or more. Surface deformations are already
visible in a number of locations along the study road.
Understandably, villagers are not willing to participate in
the maintenance of a road that is used mainly by outsiders.
30. Negros Control Site. Although Barangay

Macagahay has no road access, it appears that this is not
a major constraint in the commercialization of agricultural
production. They are selling a higher percentage of their
produce now than 5 years ago. They are selling primarily
their farm products to visiting buyers in the village because Waiting to transport goods that were
most of them have difficulty in transporting their agricultural manually brought from Negros
products to the markets. Most of the villagers who are control site
74 Appendix 5
selling directly at the market are transporting their products manually and upon reaching the
roadhead, they reload their products in tricycles or in buses/jeepneys or use an animal for the rest of
the journey to the market. The village local government recently acquired a truck of 3.5 tons capacity
for villagers to market their agricultural products upon reaching Barangay Montilla and to reach
Moises Padilla at subsidized rates. This provides an opportunity to the poor to be hired as porters of
agricultural products from the village to Barangay Montilla. They are usually paid P10/sack of
25 kilograms of charcoal, banana, camote, and cassava and P10 per pole of bamboo.
31. There is minimal travel movement of households outside the village. No one travels outside
the village to undertake processing agricultural products, to transact business or for employment, or to
attend classes. Minibus is the preferred mode of transport of most households in the village upon
reaching Barangay Montilla after walking for 30 minutes to 1 hour. Adult men travel outside their
village to buy provisions for their family and to sell harvests or other products. Adult men and adult
women travel together to procure and secure medical needs of their family and documents at the
municipal office. On average, one has to travel about 25 km and spend 2.3 hours to perform a task in
the town. Since traveling outside the village is very costly involving an average expense of P32 for
each return trip, villagers schedule trips outside their area once every 2 months. However, a lack of
transport services has not constrained the villagers from participating in the 2001 local elections as
shown by the 93% voter turnout. Most villagers are not prepared to contribute to maintenance.
C. Poverty Situation
1. Common Characteristics
32. The poverty situation is severe in all sites. The community classification of socioeconomic
groups is shown in Appendix 3 in tabular form. The very poor households are generally in a chronic
state of deprivation, they are mainly landless and are highly dependent on wage labor with no regular
employment. The poor have some livelihood potential, depend on wage labor, but have regularity of
employment due to long service to landowners in their village. Within this group of households, there
is a high variation from those who are perilously close to chronic poverty, to those with a degree of
security, and potential for accumulation and development. Better-off households are those that have
significantly accumulated capital, a high degree of land ownership, and access to income and
employment opportunities outside the locale. The majority of households have been resident in the
village for over 5 years. Others are usually among the poor in the village because they are more
vulnerable to shocks, not having established a strong social network within the community.
33. The average village household size is about six members with poor households usually having
more children typical to most rural villages. In addition, households with a high number of dependents
such as school-age children, the elderly, and handicapped members are usually the poor and very
poor as well. About 5% of households have female household heads and these families belong to the
very poor or poor groups. They are either single parents or widows with six or more dependents. The
average education of household heads and spouses is 5 to 6 years (except 3 years in the Negros
control site). The very poor and poor households are usually those households with heads who have
a low level of education; they have little understanding of new farming technologies; or find it difficult
to get nonfarming employment. Laborers are subject to the seasonality of production. With no regular
income, they have no opportunity to accumulate savings to invest in more profitable activities that
better road and transport services may offer.
34. Most farming households claim that there are no changes in size of farm area, or in the
amounts and types of crops that they have been growing over the last 5 years. In general among all
socioeconomic groups, about half of the household income is spent on food and about 6% on
transport and 6% on education. However, the very poor and poor households spend at least 65% of
Appendix 5 75
their income on food and this can go up to 80% in the control areas where scarcity of opportunities is
higher. This prevents them from building up their financial capital to invest in productive endeavors.
Most villagers claim that there has been no change in their income sources over the past 5 years.
They would like to increase their income through raising animals or starting a small business.
However, this opportunity is beyond the reach of the very poor and most of the poor households in the
village, as the initial capital required is high for animal raising. Without access to credit, they cannot
find the start-up capital. Most of the very poor households mainly raise chickens instead for an
emergency such as school fees or medical treatment. Considering all members of the households, a
great number of villagers (over 80%) never worked away from the village despite the improved road
and transport services. Those with an opportunity to work outside the village are those with adequate
education and skills and are mainly from better-off households. Opportunities that exist are mainly for
temporary/seasonal work (over 70%), either in urban areas or in a rural zone in the same district.
Opportunities for temporary work outside the village are biased against women.
35. Most floors of the poor houses in the village are made of packed clay or dirt, with bamboo
walls and thatch roofs. In the Negros control site, even raised floors are made of bamboo as there is
an abundance of bamboo in the village, which is not marketed due to problems of transportation. As
the poor do not have enough money for food, they have few savings or little money for the
improvement of their houses. Most households have invested in hand tools as they are dependent on
agriculture, but the poor have usually fewer such tools. Only the better off have electric connections,
while the very poor and poor have mostly kerosene lanterns for lighting their houses. A majority of the
households (about 70%) in the project sites get their drinking water from protected sources in the
village, while most of the very poor usually get it from unprotected sources since they are located
away from the village center. But in the control sites, there are only a few protected sources. Most of
the very poor households have no toilet facilities and use open fields. Many of the very poor and poor
households seek medical treatment from the village health post clinic and traditional healers, while the
better off go to hospitals. A high percentage (80%) of the households experience food scarcity ranging
from 1 to 2 months, but very poor face up to 6 months of food scarcity. During months of scarcity, they
get credit from store owners and seek extra work. During these periods, they depend primarily on root
crops for subsistence. Over the last 5 years, most of the villagers claim that their food scarcity has not
reduced. Due to food shortages, the very poor and poor households usually resort to slash-and-burn
farming.
2. Sorsogon Project Site
36. In Barangay Palale, the villagers classify the socioeconomic groups as very poor (10%), poor
(60%), and better off (30%). The primary occupation is agricultural wage laborer (50%) for most
households and others (26%) are coconut farmers. As harvesters and processors of copra, laborers
work for a daily wage rate (P100/day). The 2-month cycle of copra creates an income gap in between
months for the very poor and poor households. A high percentage of average household income of
villagers is spent mainly on food (54%); the very poor and poor households spend at least 65%. About
half the households (57%) have private toilet latrines, but most of the very poor households have no
toilet facilities and use open fields.
37. Most of the farming households are tenants farming over 1 ha of land and growing coconut.
Farming households have sufficient family labor and do not need to hire, resulting in even fewer
opportunities for the very poor. Banana and some root crops are also grown. There have been no
changes in size of farm area and the amount of crops that they have produced over the last 5 years.
This development has negative implications on the earnings capacity of the very poor and poor as
well. As little capital is flowing into the local economy, the poor are forced to engage in highly labor-
intensive businesses to earn a little, e.g., selling bananas or sand from the river. In addition, the very
poor collect and sell firewood, a tedious task. Despite the presence of the “Greater Market Access
76 Appendix 5
Scheme” van along the road in the village (a government-supported program to sell staples at
subsidized rates including rice, eggs, and sugar), the poor and very poor are unable to maximize the
impact of these services as they have limited money to buy adequate amounts of food for their needs
as they need to pay in cash. By self-assessment, most of the households consider themselves having
the same quality of life as their neighbors (49%), 29% consider their quality of life as worse than their
neighbors’, while 20% feel they are better off.
3. Sorsogon Control Site
38. The poverty situation in Barangay Bical is more severe than in Barangay Palale. They
consider 10%, 70%, and 20% as being very poor, poor, and better off, respectively. In Bical, similar to
Palale, agricultural wage laboring (66%) is the primary occupation but in contrast, the primary
occupation of the spouses of household heads is not farming but also working as a wage laborer. This
is an indication that poverty is more intense in Bical with more households not having a regular
income stream. As their mobility is highly confined in the village, the very poor and poor have difficulty
in scouting for job opportunities that may improve their income capacity. The farmers generally
cultivate over 1 ha of land under a sharecropping arrangement with owners. In contrast to Palale, the
majority of the households in Bical are collecting forest products for sale. The primary forest products
are traditional plants. The abundance of these forest products helps the very poor and poor
households in bridging the gaps in their income due to seasonality of copra production. But high
transport costs are constraining them from selling more forest products. There has been no change in
access to common land for grazing animals or collecting forest products now compared to 5 years
ago. With 5–6 days of continuous rains, Bical gets flooded. This has a serious impact upon the poor
who lack the safety nets and security assets to recover quickly. As the very poor and poor are paid on
a daily work basis, they do not earn during natural calamities. The severity of the situation depends on
the duration of the natural calamities. To attend classes, high school children have to walk across the
flooded river, change, and proceed to their school. Much less capital is flowing in Bical than in Palale,
resulting in an acute lack of capital for villagers to engage in business. Most (63%) of the poor
households consider themselves as having the same quality of life as that of their neighbors, while the
rest claim they are worse off (very poor, 23%) and better off (15%).
4. Negros Project Site
39. In Barangay Magallon Cadre, the poverty incidence is high. Villagers consider a high
percentage of households as poor and very poor (60%), better off (35%), and rich (5%). The poor in
the village are those households that have difficulty in meeting even their daily household food
requirements. They are usually landless and work as unskilled labor in the plantation. The better off
are skilled workers (drivers, mechanics, carpenters), tenant farmers, those in business, and fixed-
income professionals. The rich are big landowners who own 80% of the agricultural land of the village,
65% of which is planted with sugarcane.9 As the local economy is dependent on sugarcane
plantations, livelihood and employment opportunities in the village are very limited and seasonal. It is
only in the peak season (October to February) that all workers in haciendas are employed. During the
off-season, the poor have practically no other source of income than the haciendas.
40. Villagers’ primary occupation is agricultural wage laboring (48%) and farming (22%). There is
not much alternative employment opportunities except for a few variety stores. The absence of
microenterprises in the area is mainly due to villagers’ lack of capital and credit history, and collateral.
9
The socioeconomic classification in Magallon Cadre is slightly different because of the agro-industrialization in the area.
There is a large wage labor-base providing services to absentee landowners in sugar production. The main socioeconomic
difference identified by villagers is, therefore, between those who labor (and thus fall into the category of the poor and very
poor); those who have some alternative source of income (the better off); and the landowning rich. Within the category of
the poor are included those who, in other study areas, have been identified as the poor and very poor.
Appendix 5 77
Several informal credit schemes are available (even prior to the road improvement) for villagers to
start a microenterprise, but these are not easily available to the poor households in times of
emergencies. There has been more participation in the credit schemes after the road rehabilitation.
Most of the target clients are the poor rice farmers who have high repayment rates as they have the
flexibility to pay in cash or in kind. Variety stores provide informal credit to the poor during hard times.
Wage rates for hacienda workers in the village are very low (males from P70/day to P100/day and
females P60/day). The poor do not seek jobs outside the hacienda because their exposure is only
limited to work in haciendas. Unless they develop other skills, they cannot take advantage of the
transport services to seek employment elsewhere. They usually have no savings at all to draw money
to look for other opportunities. Because of a labor surplus in the village, poor households make
themselves always available for work in the hacienda, otherwise they will be replaced.
41. Most poor households experience food scarcity, especially in August. This coincides with the
milling season when there is no demand for labor in the haciendas, and with the heavy rains in the
area. Some cut down on food expenses by just having coffee during breakfast, bread and coffee
during lunchtime, and rice during dinner. Farming households mostly farm not more than 1 ha of land
and most of them are tenants. Although sugarcane plantation dominates the landscape of Magallon
Cadre, the majority of households rank paddy as their most important crop. These farming
households seldom hire the services of the poor households in the village as they have enough family
labor for their farm requirements. The majority of the households (68%) consider themselves as
having the same quality of life as their neighbors; worse off 22%; and better off 10%.
5. Negros Control Site
42. The impoverishment of Barangay Macagahay is mostly due to inaccessibility to the area.
Although the economy is diverse, its development is highly
constrained by low agricultural productivity of staple crops
such as corn and rice due to difficulty in transporting farm
inputs such as seeds and fertilizers, and considerable
difficulty in marketing agricultural products such as
banana, bamboo poles, and farm animals. Villagers
classify households in their area as very poor (20%), poor
(70%), and better off (10%). The very poor households are
landless, have no regular job or are even totally jobless,
household heads with no formal education. The poor
households have access to land where they plant mostly
staple crops and are mainly self-employed. The better off
are professionals who settled in the village due to their
Children carry water in
work assignment and those owning agricultural land.
Negros control site
43. Due to absence of access roads, villagers have considerable difficulty in traveling during the
rainy season as it is slippery terrain with flash floods. Farm products perish when these are not
transported during the rainy season. Farming (46%), followed by agricultural wage laboring (39%), is
the primary occupation of most members of the households. Despite inaccessibility, agricultural
buyers come to the area as they can buy livestock and other agricultural products at about 10–20%
cheaper than in Moises Padilla town. Most households in the village are dependent on farming, rather
than on the agricultural wage labor market. They farm less than 1 ha of land and are either owners or
tenants. The majority of the households rank corn as the most important crop for them as it is their
staple food. Banana is their next important crop followed by root crops such as camote and cassava.
In particular, most of the very poor households survive day-to-day by just having root crops for
breakfast and lunch. Most only have a full meal for dinner; steam-milled corn with dry fish or fish
paste, or even salt.
78 Appendix 5
44. Unlike Magallon Cadre, a relatively larger number of households (38%) collect forest
products—mainly charcoal and bamboo—and sell these at the markets or to neighbors. Usually, the
very poor and poor households collect forest products as it is time consuming. At the village, they sell
charcoal at P25–P30 per sack and bamboo poles at P20–P25 per piece. To this price, they just add
the porter charges of P10 per sack and per piece for buyers in Montilla. Most households in the village
do not wish to sell more animals and forest products, particularly bamboo, as they have considerable
difficulty in transporting the products to the market. The majority of the households (70%) consider
themselves as having the same quality of life as that of their neighbors, while the rest feel worse off
(22%) and better off (8%).
D. Transport Impacts and Linkages to Poverty
45. Beyond ensuring basic access, transport is a low priority for the very poor. Much of their travel
is localized, for productive or subsistence purposes. In particular, they have little capital to set up
small businesses or any income-generating activity along the road to take advantage of transport
services. Even for long-term residents who may have the potential to slowly graduate out of poverty,
significant restraints exist. For small tenant farmers of copra, for example, conditions of tenancy
include selling copra to the landowner at a predetermined price. The potential for making use of better
transport links to explore wider markets is, therefore, irrelevant. The very poor and poor households
use jeepneys rather than tricycles since they are cheaper. As a consequence, they spend more time
in traveling and this prevents them from engaging in other activities where they could earn income.
46. After the rehabilitation of the study road, there is some outmigration evident from Barangay
Palale to urban centers. There is also outmigration in Bical but at a lower rate. However, there is no
necessary link between graduation from poverty and migration for employment, unless this temporary
migration is sufficiently secure and regular to replace the existing income and subsistence function
from agriculture. The opportunity to get well-paid and regular work outside the community is often
closed to the very poor, who lack the access to information and social networks to be able to take
advantage of these opportunities. The road project has substantially influenced the setting up of
roadside businesses especially in Negros project site. Small vending stalls (particularly for fish) and
variety stores along the road offer an opportunity as main and supplementary sources of income for
the poor and a possibility for them to move from a poverty status to being better off.
47. Based on the overall perception of the villagers in the study areas, the positive impact of
improvements to roads as ranked in order of importance are (i) availability of more transport,
(ii) cheaper transport and better road surface, and (iii) increased opportunities to sell things. The
negative impact to making improvements to the roads are ranked as (i) increased dust and noise
pollution, and (ii) outside interests buying up land and resources. There was no significant movement
away from poverty, although marginally, the project sites have less poverty than control sites in both
the study areas even with the existence of good quality roads and transport services. Household
income and consumption levels remain low and in some cases are not sufficient for their basic needs.
This is evident in the continuing high incidence of household poverty in the study areas, 70% for
Barangay Palale and 60% for Barangay Magallon Cadre. With no direct access to a road, the
incidence of household poverty in the control areas is 85% for Bical and 90% for Macagahay. As
construction/rehabilitation of the road project relied upon machinery and outside contractors, there
was little involvement of the poor in the road projects. There were few opportunities for the villagers,
particularly the poor in the study area, to supplement their income either by being involved in the road
construction phase or its subsequent maintenance work.
48. Lack of access to land remains a critical determinant of poverty in both study areas.
Households that are unable to access regular income from agricultural production usually belong to
Appendix 5 79
the very poor and poor categories as defined by the villagers themselves. Agriculture did not progress
as expected, with the same traditional crops grown and cultural practices adopted despite the removal
of major constraints to the flow of technology through the availability of good access roads and better
quality transport services in the project areas. There are no innovations in primary production both in
the coconut and sugar industries, as they remain vulnerable to market price fluctuations that put the
locals at the mercy of external forces. Farming households in project sites have more direct access to
markets and contact with agricultural buyers due to availability of better roads and transport services.
They are no longer solely dependent on buyers coming into their area. On the other hand, there is no
influx of other cash crops into the study areas nor are villagers practicing intercropping or multiple
cropping for diversifying their activities and increasing their food options. It seems that the transfer of
agricultural technology is lacking in both study areas. Livestock and poultry sectors are still backyard
growers. As a consequence, villagers in both study areas suffer food deficits as the quantity that they
produce is less than the quantity required by the households for food consumption.
49. The graduation from poor to nonpoor through migration may be possible only if outside
employment is sufficiently secure and regular to replace the existing income and subsistence function
from current livelihood activities. Based on the situation in the study areas, the opportunity to get well-
paid and regular work outside the community is often closed to the very poor. Better roads and
available transport services enhance the delivery of various government services to the poor in the
project areas. It also gives government officials the opportunity for timely and efficient monitoring of
the progress of their programs for the poor. Thus, roads are critical as social arteries for the delivery
of government services, penetration of ideas and cultures, and technology dissemination to poor
people in the project areas. The potential for using the roads to facilitate economic activity is evident
in the project areas, but the poor are in need of financial assistance to start their own businesses like
sari-sari stores along the road, which might enable them to get out from poverty.
80 Appendix 6
CASE STUDY DETAILS IN SRI LANKA
A. Introduction
1. This appendix presents the information on the Sri Lanka case studies in summary form.
More details are available in the country report.1 The two road projects selected as case studies
were North Western Province Water Resources Development Project (NWP-WRDP) and the
Southern Provincial Roads Improvement Project (SPRIP).
1. North Western Province Water Resources Development Project
2. Description of NWP-WRDP. The main objective of the Project is to improve the economic,
social, and nutritional well-being of the people living in the project area by increasing the income of
local farmers, diversifying to higher value crops and increasing the employment opportunities for
the unemployed and women. In order to meet these objectives, the Project is made up of the
following components:
(i) rehabilitation, improvement, and restoration of rural infrastructure including major,

medium, and minor irrigation systems and rural roads;
(ii) credit lines to promote development of farmer-owned irrigation systems (wells and
lift pumps) and to rural women for establishing income-generating activities; and
(iii) institutional support in the form of facilities, equipment, training, consulting services,
and monitoring of project implementation, benefits, and environmental impact.
The development of rural roads was only one component of a large and diverse project. Four
hundred fifteen kilometers (km) of D and E class roads and 573 km of agricultural roads were
completed under the Project. The selection of roads was based on the intention to provide access
to individual villages and irrigation schemes and link them to provincial highways. The work
primarily entailed broadening carriageways, graveling the surface, and improving culverts and
bridges.
3. NWP-WRDP Setting. The North Western Province (NWP) of Sri Lanka comprising districts
of Kurunegala and Puttlam has a diverse ecology spanning three major zones: wet, intermediate,
and dry. The primary crops vary from being predominantly coconut in the coastal areas to paddy in
the interior dry zone. The road density in NWP is higher than the national average, particularly for
rural roads. The topography of the region enables the low cost construction of basic tracks, which
has been a popular activity for self-help components of a number of rural development programs.
Tracks over time become upgraded to gravel roads. The area focused upon by this study was the
Kotewehera Divisional Secretary’s (DS) Division of the Kurunegala District. The road connecting
the provincial capital of Kurunegala to Puttlam runs by the side of the Kotewehera DS Division. The
study area is in the interior of the division.
2. Description of the Study Areas in Kurunegala
4. Ambala-Inginimitiya Road, Kurunegala. The road segment selected is a D class road,

from Ambala Junction in Kotewehera to the Inginimitiya tank dam, a length of 16.5 km. It falls
under the purview of the Wariyapola Divisional Engineering Office of North Western provincial
council. The road was developed in three phases: the first 4 km was metalled and tarred in 1995,
1
Asian Development Bank. 2002. Impact Evaluation Study of Rural Roads on Poverty Reduction: Sri Lanka Country
Report. Manila.
Appendix 6 81
4–8 km graveled in 1996, and 8–16.5 km going through the case study village was graveled in
1998.
5. The first 4 km are in a good condition. The

presence of potholes was moderate along the
gravel road over the 4–8 km stretch. But the
condition deteriorated with many potholes
thereafter, but overall, the road was in a usable
condition. The culverts along the road and the
bridge at 16 km are also in a good condition. The
drains on either side are not continuous and are
partly filed with loose gravel. This may cause the
road to become inundated during heavy rains. The
road is maintained by the North Western Provincial
Engineering Department, which gets Rs12,000 per
km per year for the maintenance of a gravel road.2
The current method of maintenance is by direct Schoolchildren taking trip to town on a
labor. Maintenance of gravel roads are mainly in hand tractor on the Kurunegala study road
terms of motorgrading especially after heavy rains.
This was done at the end of 2000.
6. Kurunegala Project Site. The case study village of Nugannoruwa is a traditional village.
The houses are clustered together, bordered by the Nugannoruwa tank and (wet) paddy fields on
one side and the dry farmlands on the other. The project road runs across the side of the village. A
gravel road circles the inner side of the village with a network of tracks and footpaths leading off to
individual households. The 118 households, of which four are female headed, are linked with very
strong kinship ties perpetuated by the high level of intermarriage. Lower income households rely to
a great degree on the support of relatively more economically stable households.
7. This area borders Sri Lanka’s dry zone and had suffered from a severe drought for the past
3 years. The irrigation tanks surrounding the village are dry, and farmers who are unable to depend
on paddy cultivation as a livelihood are relying now upon traditional forms of dryland agriculture,
akin to shifting cultivation. Problem ranking with different groups of villagers showed that this crisis
afflicts all members of the community, and not just the landowners, as poorer villagers too are tied
into the paddy economy through daily waged labor. While richer landowning villagers have assets
and savings upon which they can rely on in a crisis, the poor seldom have this support, and so fall
quickly into a cycle of impoverishment from which it becomes increasingly difficult to escape.
8. Due to the prevailing water crisis, more and more food items have to be purchased and the
debts to village shops increase. The shopkeepers generally give credit for small items till the end of
the week, but full payment is made only rarely. The shopkeepers buy stocks on credit from vendors
in the town with whom they have good relations. The high degree of social cohesion has also
resulted in very active community organizations. The funeral aid society3 of Nugannoruwa has
expanded to include activities such as managing the village preschool in the community hall built
by it. It is also the controlling body for fishing in the Nugannoruwa tank.4
2
A section of the road of length 5.6 km that goes through Inginimitiya dam is maintained by the Irrigation Department.
3
Funeral aid societies (maranadhara samithi) are the most dynamic and successful community organizations in both
rural and urban Sri Lanka. These societies are self-mobilized and have the primary function of providing bereaved
families with support after a death. Most societies have now spread to include other functions.
4
Fisherfolk are charged a 50% levy on their catch and the funds are used for tank maintenance.
82 Appendix 6
9. Kurunegala Control Site.5 The control area studied was made up of two villages,
Walahinikalla and Wembuwa, which are at least 45 minutes’ walk from the bus route. Both are
populated by families that have moved into these less connected regions mainly due to the
availability of cultivable land. Walahinikalla is an extension of Nugannoruwa. All households in
Walahinikalla participate in social activities and the community organizations of Nugannoruwa.
Villagers in Nugannoruwa own paddy lands in the Walahinikalla fields. These are the only fields
that have been cultivated during the drought. It is the availability of water that has lured families to
move to this more remote area. Of 11 households, two are occupied during the daytime by elderly
widows who go back to Nugannoruwa to sleep, as access to health care in an emergency is
difficult from Walahinikalla. Similar to the project site, the primary livelihood source is paddy
farming supplemented by livestock rearing, which has increased substantially over the last few
years (Appendix 6, para. 40). The land in the high ground enables pasturing, home gardening, and
the cultivation of coconuts. The main preoccupation of the community is the lack of a road, an
issue that came up frequently in communication with the study team.
10. The 28 families that make up Wembuwa are mostly young families that have migrated from
villages in the region over the last 15 years in search of land to encroach on. The structure of the
community is much looser than in Walahinikalla. The village is approached by a track that dwindles
out to a network of ill-defined footpaths. The boundaries of the village are not very well defined.
There are no community organizations. However, this village is not as inward looking as would be
expected given the level of isolation. Almost every household has a migrant member, either in the
Middle East or in other towns in the province or the capital city. Though the villagers of Wembuwa
are as isolated as those of Walahinikalla, they seem less preoccupied with the problem, which may
be because they can access the road by creating footpaths across the flat land. Walahinikalla, on
the other hand, can be completely cut off from the road network, especially during the wet season
when the tank spills over.
3. Southern Provincial Roads Improvement Project
11. Description of SPRIP. The rationale of the Project is to improve rural accessibility and
reduce road transport costs, as these are prerequisites to achieve economic growth and poverty
reduction. The primary benefits of the Project are perceived in terms of reduced road transport
costs as a result of improved accessibility and reduced maintenance costs. The main beneficiaries
are expected to be the rural population and other low-income groups in Southern Province. It is
currently ongoing, with implementation commencing in April 1998 and scheduled for completion by
December 2003. The implementation is through the Southern Provincial Road Development
Authority, which reports through the Southern Provincial Council to the Ministry of Provincial
Councils and Local Government of the central Government.
12. The project objectives are to (i) rehabilitate about 550 km of the 1,800 km secondary road
network in the Southern Province; (ii) improve the capacity of the Southern Province Road
Development Authority (SPRDA) to manage the network efficiently and effectively; and (iii) develop
the institutional, financial, and private sector contracting resources needed for sustainable road
maintenance. To meet the above objectives, the Project includes three components: civil works for
road and bridge rehabilitation, SPRDA capacity building, and consulting services for project
implementation.
5
The identification of a sufficiently populated control site was a problem. Due to high road density, the accessibility
criteria used was a minimum of 45 minutes walking distance to a bus route. But it was not possible to find villages that
were sufficiently populated to enable a random sampling of 40 households. Therefore, two villages that were at least
45 minutes’ walking distance from the Inginimitiya Road bus route were selected to carry out a census rather than a
sample survey. Even then, the number of usable household surveys completed in the control area was 29.
Appendix 6 83
13. The SPRIP developed a system of prioritizing roads for selection.6 Each road under
consideration is ranked based on key indicators of present road condition, economic development
potential, and social needs. The economic development potential focuses on the number of
villages and the land use patterns, commercial activities such as shops, industrial activities such as
factories, and the availability of public transport and traffic levels. The social needs concentrated
on identifying disadvantaged communities. The percentages of the population receiving
government benefits for the economically disadvantaged and with access to public utilities were
the key indicators used in the ranking. The weighting given to the economic and social aspects
reflects the orientation of the Project toward economic growth and poverty reduction.
14. SPRIP Setting. The Southern Province is made up of the three districts of Galle, Matara,
and Hambantota. It displays a great diversity in ecology and correspondingly in the economy.
Approximately 45% of the population is employed in the agricultural sector made up of coconut
cultivation in the wet and intermediate coastal areas and tea, rubber, and spices in the hilly wet
zone in the north. Paddy is cultivated extensively in the dry flat zone as well as in the river valleys
throughout the province. The levels of connectivity in the Southern Province does not compare well
with national averages, particularly in terms of rural roads. The case study area was located in the
Pasgoda DS Division in the Matara District. Here, of the existing roads, only about 50% can be
used by all types of vehicles while over 15% cannot be used by any vehicles.7 This area’s
economy is based on smallholder tea, coconut, and cinnamon.
4. Description of the Study Areas in Matara
15. Urubokka-Katuwana Road, Matara. The selected road section considered for study in the
Southern Province is a C class road from Urubokka to Katuwana, 10.5 km in length. The SPRDA of
the Southern Provincial Council manages it. Makandura divisional engineer (under Matara district
engineer) is allocated Rs10 million annually to maintain 171 km of paved roads and 40 km of
gravel roads. Normally, roads are patched three times per year while those that are extremely
poorly motorable gets patched as much as 12 times per year. The current method of maintenance
is by direct labor, and this is expected to change to tendered contracts in the future. The study road
is presently in good condition since construction was completed in 2001. The road surface is single
bitumen surface treatment with 14 millimeter metal chips and is expected to be strengthened in the
future with a chip sealing. The ground surface throughout was strengthened with an asphalt-based
concrete layer prior to construction. A high gradient section of around 300 meters is concreted to
avoid erosion. There are two span bridges at the 2 km and 5 km points.8 In 1988, the road was
widened, graveled, and it was tarred in 1994. It deteriorated to a poor condition due to its
narrowness and the lack of maintenance of the surface and was almost unusable for vehicles prior
to the SPRIP rehabilitation.
6
Information based on communication with General Manager, SPRDA, and Southern Provincial Roads Improvement
Project; Phase 1 Final Report.
7
Wijekoon Banda, T.M. 2001. The Study of Rural Road Network in Southern Province. Unpublished draft report.
Intermediate Technology Development Group.
8
The bridge at 5 km is about 3 m and at an intersection with a sharp bend. It was observed that the 2-door bus heading
from Katuwana had to back up four times to cross the bridge.
84 Appendix 6
16. Matara Project Site. The sections of the villages in Heegoda Grama Niladari (GN) Division,
which were in close proximity to the project road, were used as the project site. The Heegoda GN
Division is made up of seven loosely structured
villages populated by 490 households, 6% of which
are female headed.9 Villagers were not always
clear as to which village they belonged to within
the division. This fluidity in the concept of the
village was in stark contrast to that of the project
site in Kurunegala. The terrain over which the
villages were spread may make it less possible for
the boundaries to be identified. The fact that some
households living by the roadside had agricultural
land on the higher slopes may also have
contributed to the fluidity. The project sites did not
have any functioning community organizations,
Rocky path to Makiliyatenna is either self-generated or externally facilitated. The
difficult and dangerous only society that operates at some level and has a
village-wide membership is the clientele of the
temple. Community level organizations that are sponsored by the State do exist, but come into
operation only when the state bureaucracy necessitates it. However, friendship and kinship ties are
quite strong and are frequently relied upon in livelihood forms. Currently, the community is
benefiting from a strong price for low altitude tea. Prices for farmers’ produce are good, and
laboring opportunities in the plantations are available.
17. Matara Control Site. The control area households were on the upper slopes of
Makiliyatenna GN Division. While 310 households make up the eight villages of Makiliyatenna GN
division, about half of them live on the upper slopes of the hills on either side of the Katuwana
Road.10 The villagers of Karawilakanda, Diddenikanda, and Darandala made up the majority of the
control site sample. As in the project site, a very fluid sense of village exists. In common with the
project area, the control area also had no public institutions or common space in the villages. The
existence of only two very small retail shops and the network of water sources meant that even the
traditional informal meeting points did not function. Overall, the community in the control area is
very similar to that of the project area. The differences stemmed almost exclusively from the
location in terms of being on the higher slopes and, therefore, having poor access to a motorable
road. The hardships and indignities faced by the villagers when accessing the towns and the need
for a road was articulated frequently at focus group meetings.11 There is a strong feeling that the
benefits of the improved Katuwana Road are passing them by. The track that leads up to the upper
slopes is in very bad condition. Even if a group may hire a tractor to bring up some fertilizer, it gets
stuck quarter mile up the track requiring the operators to push it. In such an instance, headloading
is the other alternative. The villagers of the control site have to access the same institutions and
services as those in the project site, and this has increased their sense of being marginalized.
18. Kurunegala Project Site. The primary transport asset of the villagers of Nugannoruwa is
the bicycle. All households have access to a bicycle: either their own or a neighbor’s. This mode is
9
Data from Heegoda GN office.
10
Data from the Makiliyatenna GN office.
11
During the rainy season, they carry fresh clothes to change into before getting into the bus because they have to slide
on foot to get to the road.
Appendix 6 85
used extensively by both men and women within the village and for short distances out of the
village. Mobile traders on bicycles trade fish, milk,
fruits, etc. Transporting of cargo (water, firewood,
grain, etc.) on bicycles is primarily a male task
though women transport children to school. The
main mode of transport from village to the nearest
town (Nikaweratiya) is by bus. The goods are
charged at the rate of approximately Rs20 per
(gunny) bag in the bus. Salaried employees and
older schoolchildren use the bus daily to go to
Nikaweratiya town. The majority of the younger
children attend the local school and either walk or
cycle to school. The morning bus on market day is
overloaded. There is an increase of passengers
using the bus and decrease in bicycles use Bicycles are used for collecting firewood
indicating reduced local travel and increased and transporting grain in Nugannoruwa
external travel on market days.
19. The only regular public transportation available are the Regional Transport Company (RTC)
buses. The bus fare from Nugannoruwa-
Nikaweratiya is Rs10. The bus service is very
regular and thus reliable. There are six Land
Master hand tractors of 12.5 horsepower that are
owned by the villages. Most of the tractors are
used for ploughing their own lands and for their
own daily uses like transporting water. Two
tractors are used to plough for a fee. Tractors are
also used for small freight transport jobs and
passenger transport on an ad hoc basis. The
charges levied range from Rs150 to Rs600. All of
the tractor owners have contributed to the
improvement of the road at some stage through
Government bus along study shramadana programs by lending their vehicle for
road in Kurunegala transporting earth, etc.12
20. Another villager owns the only truck in the

village. The truck is used for transporting cow dung
to Kalpitiya from the village. There is excess
demand for more trucks for transporting cow dung.
The same villager also owns the only van in the
village, which is an air-conditioned 15-seater
vehicle. The demand for this vehicle is less other
than on auspicious or religious days. The driver is
paid 20% commission from the hires. There is no 3-
wheeler in the village. However, villagers access
the 3-wheelers in the neighboring villages and
those at the 3-wheeler park near the main road at
Kotewehera. Most of the garment sector workers Going to market at
Urubokka, Matara
12
Shramadana is a community activity for which each household contributes whatever possible such as labor, food, or
equipment.
86 Appendix 6
returning to the village for holidays use the 3-wheeler. It is also used to access health care and
other emergency needs as well as to transport cargo.
21. Kurunegala Control Site. The transport infrastructure and services available to the control
site community are the same as those available to the project site community. The crucial
difference is, however, that they have to first access the road. At certain times of the year, it is
possible to use a bicycle over the footpaths created over the dry paddy fields and tank bund. In
keeping with the high ownership of bicycles in the project village, over half of the houses in the
control sites too own bicycles. As in the project area, both men and women use bicycles. Very
often, bicycles are used to transport cargo even if it has to be wheeled a significant part of the way.
Adults prefer to wheel young children to school as it is much faster and less tiring than walking with
them. The greatest problems are faced in the case of emergencies. Transferring patients to
hospital is the major constraint. Unless the patient can sit on a bicycle, the only option is to use a
modified stretcher or a “sick chair.” Both of these entail significant hard labor to carry the chair over
ill-defined tracks, especially at night.
22. Matara Project Site. The traffic on this road is much heavier than on the NWP-WRDP
study road (Appendix 1, Table A1.1) because it provides a short route between two parallel road
networks running from the coast to the interior of the Matara and Hambantota districts. The most
popular personal vehicle in this area is the motorbike. The popularity of the motorbike over the
peddle bicycle is a reflection of both the hilly topography of the area as well as the relatively higher
income levels from cash cropping in Matara compared to Kurunegala sites. However, a majority
are bikes which have very low engine power. There are only two buses (both privately owned)
plying the section of the road past the study sites. The one-way fare is Rs20. Usually, the buses
have a specified schedule. The operators informed that the passenger demand is high on Saturday
(market day) and that buses do not operate on a time schedule but commence the journey as soon
as the bus is loaded to capacity. On Saturday, an additional private van runs between
Udagomadiya and Katuwana. According to the sources from RTC, a new bus plying between
Katuwana and Urubokka was expected soon after the fieldwork period. The transport service is
very clearly in a dynamic stage of expansion, with even long distance RTC buses going through. If
more buses start using this road as a link route, the service available to the villagers will increase
dramatically, provided that the buses stop along Katuwana Road.
23. Tea estate trucks are cargo vehicles which have been modified to carry village labor at the
tea companies’ cost to large and medium-scale tea
estates. There are two modified trucks and one bus,
which operates to Kiriwanaganga Tea Estate, which is
the largest in the area. A vehicle is paid Rs1,450 per
day for the two trips in the morning and in the evening.
The timing is adjusted to enable the transport of both
day and night shift workers in the two directions. In
addition, there are about 10 more modified trucks
plying to estates in Morawaka, Deniyaya, and Waralla
areas (all over 30 km away). These vehicles are also
paid in the range of Rs1,500 per day. Two 3-wheelers
are parked near the Nagaha Junction (close to the
project site) and near the Udagomadiya Bridge (close
Trucks to pick up women for work in to the control site). There are other 3-wheelers in
large tea estates in Matara households along the road that do not park at the hiring
points but are available for hire. The 3-wheel hire
charges for distances less than 3 km is between Rs30 to Rs50 per km. Each additional km is
Appendix 6 87
charged from Rs10 to Rs20. For those from outside the village, they charge more compared to the
villagers. The 3-wheel operators were of the opinion that their income was reduced after the
completion of the road project. Three-wheelers are used essentially for nonroutine travel and for
visiting the hospital and most frequently when children are taken to the hospital. Hand tractors and
trucks are used to transport green leaf (unprocessed tea leaves) from the collection points to the
factories. Cultivators headload the tea to the collection points on the road.
24. Matara Control Site. As in the case of the Kurunegala control site, the households in the
control area rely on the same transport infrastructure and services as those in the project site. The
critical difference arises from the much harder terrain faced by the control area prior to accessing
the motorable road. A few households have invested in motorbikes, primarily as an income-
generating asset. Most use it to access employment outside the area or, as in the case of a small
retail shop owner, to assist in transporting cargo up the slope. The situation in terms of taking the
sick to hospital is far worse than in the control area in Kurunegala. While the same methods of
stretchers and “sick chairs” are used here too, the hilly and slippery terrain make it a far more
difficult and dangerous task. Very often, there is no option but for a sick person to walk unaided at
least up to the track that leads up from the road.
25. The poverty situation in Sri Lanka presents an unconventional duality as the level of human
development is far higher than can be expected on the basis of its gross domestic product (GDP).
The GDP per capita of $820 ranks Sri Lanka as a lower income nation and 137th in the world.
However, in terms of human development, Sri Lanka ranks 81st in the world with a human
development index of 0.735.13 However, consumption poverty still remains a major issue as
approximately one third of the population fail to achieve an acceptable quality of life. The incidence
of poverty at the lower poverty line of Rs791 per person per month at 1995/96 prices is 25.2%. At
the higher poverty line of Rs950, the incidence moves up to 39%.14
1. Kurunegala Project Site
26. Discussions on the levels of economic stability and poverty in the village were strongly
colored by the present water crisis. The level of
dependence on paddy farming and agricultural
wage labor as a factor defining poverty came out
strongly. The causes of poverty were identified as
the lack of a livelihood source, large number of
young children, lack of land and livestock, and,
specific to women—being abandoned by the
breadwinner. In the focus group discussions, 20% of
the village was deemed to have a good standard of
living (better/off and rich) because of fixed/salaried
income or other sources. This group was better able
to cope with the crisis situation. The worst affected
were identified as 30% (very poor) of the
Participatory rural assessment
households that were totally dependent on
activities in Kurunegala project site
agricultural wage labor. The rest (50%) of the
13
United Nations Development Programme. Human Development Report 2001.
14
An official poverty line does not exist for Sri Lanka. The accepted working levels are those calculated by researchers
based on the data of the Central Bank of Sri Lanka or the Department of Census and Statistics.
88 Appendix 6
households (poor) were in between. In the survey, the self-perception of comparative quality of life
was quite similar with 28% perceiving themselves as “better than others,” while 20% thought
themselves “worse than others,” leaving 42% “about the same as others.”
27. Landholdings. This problem of lack of water—rainfed or irrigated—has been gradually

aggravating over the last few years and has now begun to take on crisis proportions. Its impact is
felt by the entire village irrespective of their income status. The only fields that are cultivated for at
least one season are those that are fed by the Walahinikalla tank, which is adjacent to the control
site. The project site tank is conserved for consumption. Access to land is fairly equitable with only
about 10% of the households being landless and most family groups (but not all household units)
have some land in the Walahinikalla fields as well as the Nugannoruwa fields. The size of
landholdings varies dramatically: about three family groups hold a majority of land. The situation
could have been rather different if farmers with larger holdings used up the existing water at the
expense of the smaller farmers. The strength of the farmers’ society, kinship ties, and community
cohesion seem to be factors that have prevented this from arising.
28. Changing Livelihood Patterns. Despite the current attempts to distribute water equitably,
being deprived of the primary source of income and livelihood has had different impacts on
households. Those households with a large landholding, which previously had a strong income
flow, have been able to build on their assets to cope with the changing situation. The case study of
Mr. Lalith (Appendix 8, Box 16) shows that the movement into transport and trade-related
enterprises is an example of taking advantage of family savings and wealth to diversify
successfully when cultivation deteriorated. These households also have the advantage of having a
concentration of better paid and stable outside employment. In addition to better sources of
information and external links, the long-term higher income levels, which have enabled family
members to stay in education longer, is now paying off as almost all these households have
employed graduates or teachers. One of the more positive aspects of the conditions at
Nugannoruwa is education. The level of education in the village is very high. The majority of those
under 40 years have received 8–10 years of education. Over the last 10 years, seven students
have graduated from Sri Jayawardena Pura University Colombo, either as internal or external
students.15 The interest in education is increasing as the need to look for external employment
increases.
29. Households with a smaller landholding are also turning to other sources of income.
Increased dependence on livestock is a trend that is clearly visible. Cattle and buffalos are reared
primarily for milk and dung. Due to the existence of a system of leasing animals, households with
few assets can move into dairy farming. The milk is sold either to individual traders who collect
door to door or at the milk cooperative situated in the adjoining village (Appendix 8, Box 10).
Trading in fish, and free-growing seasonal fruits such as tamarind, lime, and mango is yet another
source of income for lower income households. The most vulnerable are those households that
have always relied on wage labor. The asset base of these households is very low and is
frequently limited to a bicycle and basic farming tools. The reduced opportunities for wage labor
due to reduced agricultural activity was a critical factor in the low-income problem ranking. The
shortage of water, exacerbated by crop destruction by wild elephants, has created a situation
where over 60% of the families in the household survey acknowledged the scarcity of food. This is
felt more keenly by the lower income groups with little supplementary cash income. They had to
survive on a day-to-day basis purchasing some vegetables from the few surviving home gardens
or the village shops. In the case of the higher income group, it was nutrition rather than food that
was identified as a problem.
15
Source: Probation and Childcare Officer.
Appendix 6 89
2. Kurunegala Control Site
30. Since most families in the two control sites of Walahinikalla and Wembuwa had moved
there to seek better livelihood opportunities, they are generally poor. The self-perceptions of the
households in the two control sites are worse than others (43%), better than others (7%), and
about the same (50%). While the possibility of accessing external sources of income is much lower
here than in the project site, the potential for subsistence home gardening is greater. There is little
difference in the conditions faced in paddy cultivation as the fields around Walahinikalla are farmed
by households in the project site as well. Most families in Walahinikalla own very small plots of
land. As such, food insecurity is a problem faced equally in the project as well as control sites. The
crucial dimensions of poverty identified by the focus group in Walahinikalla related more to issues
of access to health, education, employment opportunities, and isolation, which constrained the
improvement in housing. Unlike in the project area, the issue of access overrode the issue of
quality of health care. The lack of access is not only a matter of life and death but it is also one of
human dignity.16 Despite the difficulty of access, most families encourage children’s education, and
the average levels differ little from those in the project site. However, there are no undergraduates
or graduates in the control site.
31. The situation in the second control site is quite different. Reflecting the history and structure
of the community, the villagers of Wembuwa have varied sources of livelihood. Few own and
cultivate paddy land. Most of those who do cultivate paddy do so as tenants. As in other villages,
the lack of water and encroaching wild elephants have made paddy farming an insecure livelihood,
and livestock rearing is extensively practiced. Traders from towns in the region lease out cattle to
be reared and then sold back to them. A comparatively large number of families have members
who are employed outside the village. Middle East employment, the garment sector, and other low-
level private sector employment is quite common. The standard education is considerably lower
than that of the project area or Walahinikalla. The children of this village go to the same school as
those in the project and other control area. However, in addition to a 45-minute walk on average,
they have to take a bus to reach school. It is particularly difficult to send younger children to school
as they have to be accompanied by adults. The migrant nature of these families is also a factor
that has disrupted schooling, especially in the case of young adults. Housing conditions in both
Walahinikalla and Wembuwa fare badly. There are only a total of 11 households that have a brick
structure—of this seven are unfinished. All the rest are clay and mud houses. The cost and
inconvenience of transporting building material have to a large degree resulted in this situation.
Correspondingly, the existence of safe sanitation is very low. Over 40% of the households use
open fields. This contrasts sharply with the fact that only 12% of the households in the project area
do not have access to a private latrine.
3. Matara Project Site
32. The size of landholdings is a crucial factor in sustaining poverty. In the Heegoda GN
Division, to which the project area belongs, 36% of the households own 16% of the land of less
than 1 acre each. Twelve percent of the land, plots of between 2 and 5 acres each, is held by
10 households. The majority of the households (54%) hold between 1 and 2 acres of land and
make up 40% of the total land area.17 The group with less than 1 acre of land cannot generate
sufficient income from cash cropping and depends heavily on wage labor. The intermediate group
can meet their basic needs but also engage in supplementary activities of wage labor, skilled labor,
16
A patient who died from a snake bite, after a week in hospital, could not be brought home for burial due to floods.
17
Data from Heegoda GN Office.
90 Appendix 6
or microenterprise to augment their household income. The advantage of tea smallholdings is that
household labor can easily be divided between own land and wage labor. A parcel of land of up to
2 acres would need only 1 day of labor per week. Although the level of land ownership is not highly
concentrated, because landholdings are small and the poor are required to work in the tea
plantation throughout the year to keep their jobs, this limits their opportunity for exploring more
lucrative livelihood alternatives. The participatory rural assessments identified dependence on
wage labor as the critical factor differentiating the poor and nonpoor. Over 80% of the households
were dependent to a high degree on wage labor and thus could be considered poor or very poor.
Of these, families that had to depend totally on wage labor were seen as the chronically
disadvantaged. A factor identified by the focus group as aggravating the impoverishing effects of
wage labor was its low bargaining power. This meant that full payment was rarely made at the end
of the day’s work. The 15-day payment cycle caused problems of liquidity to those working on tea
estates as they had no savings. Focus groups were reluctant to identify any group as rich. Instead,
they categorized those who work only on their own land and/or have state sector employment as
economically stable and better off. To this group were added those who have other sources of
income based on enterprise. It is also this group that benefits from the opportunities offered in the
project area due to the proximity to the road. The majority of enterprises—such as retail shops, tea
collection points, hiring of 3-wheelers, vans, and trucks—have been established by this group.
33. The levels suggested by the focus groups and the self-perceptions of the households
correspond fairly closely. According to the self-perception, 15% were better than others,
42.5% were about the same, and 42.5% were worse than others. The groups that consider
themselves “about the same” and “worse than others” roughly equal the 80% suggested by the
focus group as being poor. In terms of education, just 1% had a tertiary education.18 Given the
project area’s proximity and potential accessibility to a high quality school, these are very
unfavorable achievements. In the poorer families where both parents were wage laborers, children
are frequently left unattended for the entire day. There are also opportunities for older children to
find income-generating activities informally in tea smallholdings deterring them from attending
school. On the other hand, the educated young were very well informed of external events and
used all possible means to derive information to expand their opportunities for income generation.
4. Matara Control Site
34. Similar to the project area, the control area focus group identified the ownership of sufficient
tea lands as a critical factor in the households being nonpoor. A smallholding of 2–3 acres was
perceived as being sufficient to prevent a household being dependent on wage labor and thereby
prone to income insecurity. The major difference between the control and project sites lies in the
fact that in the control area, wage labor is often away from home from very early hours in the
morning till late evening due to the time taken to walk down to the pick-up points of the estate
vans. Focus group discussions segmented the village along the lines of economic ability as
economically stable (10%), average (60%), and very poor (30%). The self-perception in the
household survey reflected 10% as better than others, 55% about the same, and 35% worse than
others. Education in the control area is even more unfavorable than that of the project area. They
have to deal with issues of child supervision as well as access. Their children attend Makiliyatenna
Primary School (a lower standard than Urubokka school) as Urubokka is too far to walk. Only
secondary school children access the Urubokka Central College, but it is difficult to qualify to enter
the school. Families who want to ensure a complete education for their children take the option of
boarding their children at a relative’s or friend’s house in Urubokka or other areas. The villagers in
the control area use the same health facilities as those in the control area. For a patient to walk or
18
Data provided by the Heegoda GN office based on those collected for the 2001 National Census.
Appendix 6 91
be carried down this route is extremely hazardous. From the road, they can hire a 3-wheeler or use
the bus. Since the rehabilitation, the number of 3-wheelers by the road has increased and the cost
has decreased.
D. Transport Impacts and Linkages to Poverty
35. This section looks at how the road and transport system interact with poverty in different
micro socioeconomies and different groups within the communities. While it is understood that
isolating one contributory factor from the host of issues that combine to create an impact is a
difficult task, what has been attempted is to draw plausible links based on the information
gathered.
1. The Road as an Enabling Factor in Income Generation
36. The impact of the road on the primary occupations of paddy (Kurunegala) and tea (Matara)
cultivation is very different. This is both due to the difference between the subsistence and cash
cropping orientation of the two livelihood forms, as well as the crisis situation faced by paddy
farming. However, the road is linked strongly to other sources of income, albeit in very different
ways.
37. Despite the water crisis, paddy remains the primary occupation of the majority of the
households—at least in perception, if not in reality. The prolonged water shortage, aggravated by
wild elephant encroachments, has made the harvest very poor for the last 3 years. The sale of
paddy has fallen to negligible levels because the households maintain stocks of paddy to ensure
food security as well as a form of saving. Hence, few traders come to the village to purchase rice.
Transport is important in bringing the harvest home from the fields and taking it to the mill to be
ground. While the modes of transport differ for each activity, the project road plays little part in the
moving of produce, as most travel involved is internal to the village. The dry, hardened paddy fields
where the bunds have been opened enables hand tractors to drive right up to the threshing site.
Once the paddy is stored near the homes of the farmers, small quantities are taken to the village
mill on bicycle as the need arises. In stark contrast to tea cultivation, headloading is seen very
rarely.
38. For all tea smallholders, irrespective of their socioeconomic group, the benefits of the road
are both in monetary and/or nonmonetary forms. Given the demand for green-leaf tea, the
improved road has brought better prices, greater reliability in regular collection, and delivery of
fertilizer to the starting point of the track. The collectors can no longer justify a high and fluctuating
transport cost deduction from the factory price, and increasing competition among collectors has
brought a better price. All collection points are on the project road. While the plots are on the
higher slopes, the crop has to be headloaded in both the project and control areas. In most
households, family members headload the crop to the road side. For the economically less secure
group, the benefits have been primarily nonmonetary. The price paid for daily labor has not
increased significantly. However, access to larger estates have become less difficult as the estates
now provide free transport to all registered labor. Twelve converted trucks pick up labor from
appointed points along the road in the morning and drop them back in the late evening. There is
more free time now for other livelihood or leisure activities. Even control site households now have
greater potential to register with the larger estates as permanent labor and thus reap the monetary
benefits of year-end and new year bonus and provident fund contributions.
39. The occupational forms are very closely tied with the road. All salaried employees outside
the village rely on the public transport system. Migrant wage labor predominantly uses the bicycle
92 Appendix 6
or the bus. Other sources of income generation are being developed by the villagers as a direct
reaction to the problems faced in farming.
2. The Road as a Safety Net
40. In Kurunegala, the water crisis has hit hardest the households with no diversified income
base and low levels of assets. While those with diversified income sources are rapidly expanding
them, others are entering into new sources of income based on existing asset, knowledge, or
network base. In most cases, the households have had to look
beyond the confines of the village or fields and thus have come into
direct contact with issues relating to transport and the road. Mr.
Lalith’s household (Appendix 8, Box 16) is a good example of
diversified income sources with a strong asset base, which has
expanded directly into transport-related economic activities. Even
households with a much lower asset base are increasingly turning
to mobile trading as a coping mechanism as farming becomes
unsustainable. As the majority of the households own a bicycle, this
is frequently the asset base they work with. Networks open to the
household are frequently a critical decision variable in what exactly
is traded, e.g., fish or fruits such as mango, tamarind, and lime.
Moving up the asset base, those with access to credit have
purchased motorcycles to carry out trading activities. The
rehabilitated road enables all-weather timely access for a fish-laden
bicycle, ensuring the success of the activity. As can be expected,
the condition of the road has a direct bearing on the ability and
motivation of households to rely on these coping mechanisms. The
ease of travel and safety (in terms of wastage of cargo due to
Cow dung collected in bicycles or motorcycles sliding off the road) have been facilitated
Nugannoruwa since the road improvement. The road also offers livelihood options
to households with no private transport assets. Another livelihood
that has developed is the production of milk for a cooperative that was established in the next
village. The cooperative relies on the truck arriving every day to collect the milk to transport for
onward distribution. Before the road was rehabilitated, the road surface was poor, the truck was
slow, and the road was often impassable during the rains and, therefore, not reliable. Now,
villagers know that the truck will come every day, and this certainty has given them the security to
invest their time and resources in developing their cattle holdings. In addition to milk, cow dung has
become an important source of supplementary income. Trucks collect cow dung directly from the
stalls. For poor families where paddy farming barely provides sufficient rice even for subsistence,
quite apart from being an income-generating source, these new sources of income have become
critical for survival.
41. The complete immersion of the SPRIP study site in the cash economy has meant that a
greater number of households are seeing the increased connectivity resulting from the improved
road as providing income-generating opportunities. The number of small retail shops along the
Katuwana Road has increased rapidly. More tea collection points have opened since the road
improvement in 2001. The sustainability of these economic activities needs to be considered as
time passes. In Matara, the opportunities presented by the road development have been taken
advantage primarily by those who live by the road (retail shops) and/or have an asset base or
network to build upon. The perception that most benefits from the improved road accrue to traders
is very prevalent within the community. However, households which seem asset poor are also
taking advantage of the situation. The region’s natural ability to host a variety of wet-zone tree food
Appendix 6 93
crops, spices, and fruits has enabled even the poorest households to earn a small additional
income by selling these free-growing items to traders who now frequent the Katuwana Road.
3. Road and Transport as a Contributory Factor to Quality of Life
42. The collection of water and firewood is a primary function for survival. The two study areas
provided very different examples of how problems of access can be approached. In Kurunegala
project site, a majority of households use the Nugannoruwa tank on the village boundary for all
subsistence purposes except drinking. The road and path network of the village enables the use of
simple, manually-powered vehicles to meet subsistence needs of households in terms of water and
firewood and thereby substantially reduce the burden of carrying water and firewood over long
distances. However, little has been done in terms of increasing the carrying capacity of the bicycles
by modifying its luggage space or attaching trailers. As a result, the villagers spend considerable
time and effort on these basic tasks. In contrast, in the Matara study site, access to subsistence
requirements of water and firewood has not posed many problems in the project and control sites.
The vegetation of the region allows firewood to be collected by most households from the
immediate vicinity. A network of feeder streams and springs has been tapped for piped water for
drinking and cooking, thereby saving time.
43. Economically secure households are able to buy provisions in the weekly market at a
competitive price, while the lower income groups rely on purchasing small quantities. The increase
in the shops in the village has meant that there is better access to retail goods to meet immediate
requirements. However, the prices have not come down. The condition of the house is one of the
primary indicators of someone’s economic status. The inability to improve the housing conditions
without considerable expenses and effort was seen as a major drawback created by the lack of
good link roads along the hilly areas of the Matara control site. The movement of building material
means a costly exercise in hiring headloaders or tractors. The alternative is for the family to
headload building material whenever time permits, which is expensive and inefficient. This problem
was seen in the control site in Kurunegala as well. The much higher quality of housing in the
project sites in comparison to the control sites was more due to the problems of transporting
building material than of different income levels.
44. The control site households articulated the effect of the lack of roads on human dignity as
an issue separate from income-related deprivations. The sense of isolation and marginalization
was very high in these communities. This was especially the case with households that identified
strongly with better connected households of similar socioeconomic status in all other respects.
The high road density seems to have increased the aspirations of all communities in terms of being
well connected.
4. A Core Issue: Access to Health Care
45. Better access to health care, especially in emergency situations, was an important factor
that was mentioned as a primary contribution of the road by all groups of villagers and
substantiated by the findings of the household survey. In Matara site, increased use of 3-wheelers
to take patients to hospital and the reduction in hiring charges were cited as major elements of
change since the road was improved. Participatory rural assessment activities in Kurunegala
revealed that emergency access and prenatal access had improved since a van had been
purchased by a villager. The van in the village is accessible at all times of the day and night, which
enables quicker and safer transfer to hospital than the alternatives of using a hand tractor or 3-
wheeler. The same applies to the increasing availability of 3-wheeler taxis in Matara site. The
insecurity felt by families with young children and aging adults due to difficulties of accessing
94 Appendix 6
health care was discussed in greatest detail and with strongest emotion in the focus group
discussions of the control areas. Reliance on relationships of trust and kinship in cases of
emergency is very high.
5. Penetration of State Mobile Services into the Village
46. While certain services require the villagers to access them, others are effective only if
accessed on site. Civil security provided by the mobile police units and services of health and
agriculture extension officers services such as electricity, telephones, etc., need to penetrate into
the villages if provision is to be effective. The villagers in the control areas constantly articulated
the fact that the providers of such services rarely make the effort to service areas that are difficult
to access. This was especially the case in the Makiliyatenna area where the steep footpaths made
access particularly troublesome. The opening of a region by laying a road or upgrading one can
have a substantial impact on visits by service providers. Though this does not automatically mean
a higher quality of service, it does mean the entry into the mainstream as in the case of civil
security. While all groups within the village benefit from greater penetration by service providers, it
is the low-income groups that benefit the most. As seen constantly in the study locations, the poor
seldom travel outside the immediate environment other than for income generation. Such
households would avoid accessing services provided elsewhere unless in extreme situations of
need.
6. Social Capital, Empowerment, and Felt Need for a Road
47. Of the two study areas, the level of social cohesion, kinship networks of support, and self-
mobilization of the community were much higher in Kurunegala sites than Matara sites. Despite
this community involvement and initiative in ensuring the rehabilitation of the road, the villagers of
Matara sites showed more initiative and capacity to mobilize public opinion than their counterparts
in Kurunegala who come from a more educated and cohesive group. Community involvement was
not part of the project design in either site. However, in Matara, the village bus owner approached
the recognized traditional leadership of the village (Buddhist monk) and lobbied for rehabilitation
(Appendix 8, Box 15). The critical factor for a much less cohesive village to become a substantial
force in the road rehabilitation process could be the perceived benefits of the road. The marginal
increase in benefits due to the enhanced quality of the road was very high. In addition, the
condition of the road was critical to the economic survival of the population. In contrast, in
Nugannoruwa, external connectivity was not a critical factor for survival. Despite some changing
livelihood pattern as the traditional paddy economy deteriorates, it is still a fairly weak link. For the
majority of the population, the condition of the road was most important in terms of accessing
health care, especially in an emergency.
48. The level of labor participation in the road construction was very similar in both sites. Very
few members of the communities were part of the labor force that rehabilitated the road (three
households in Nugannoruwa and one in Heegoda). Only a total of six households (out of 80) in
both project sites expressed an interest in employment in the road construction. This was due to a
lack of knowledge regarding potential wage-labor opportunities as well as a contracting system that
encouraged the use of external labor.
7. Negative Impacts of Road Improvements
49. The identification of negative impacts of the road by the villagers in the study areas was on
average very low. Over 75% in the NWP-WRDP and 90% in the SPRIP stated that there were no
negative effects. The negative effects that were identified by a few were increase in dust and noise
Appendix 6 95
pollution, increase in traffic accidents, loss of land, and disrupted drainage. In the case of the
NWP-WRDP, increased dust and its impact on health were discussed in the focus groups and
subsequently reflected in the responses of the household survey. In Nugannoruwa, 80% of those
who said that there were negative impacts referred mainly to the dust from the gravel road.
Problems arising from increased traffic moving at faster speed was echoed in Matara, especially as
a concern for children’s safety when walking to school.
96 Appendix 7
CASE STUDY DETAILS IN INDONESIA
A. Introduction
1. The study was conducted in two provinces in Indonesia: Bengkulu Province in the island of
Sumatra, and Yogyakarta Special Region in Java island. The first case study focused on the Tree
Crop Smallholder Sector Project (TCSSP)1 at Bengkulu Province, while the other case study
focused on the Third Local Roads Project (TLRP).2
1. Tree Crop Smallholder Sector Project
2. Description of TCSSP. The TCSSP project funded by the Asian Development Bank (ADB)
($135 million)3 was implemented March 1992 to March 2001 in the provinces of Aceh, Bengkulu,
South Kalimantan, Central Kalimantan, and East Kalimantan. The project objective was to reduce
poverty by improving the income and employment prospects of rubber and tea smallholders and
landless rubber tappers of whom over 50% were living below the poverty line. This was to be
achieved by increasing the production of rubber and other tea crops through variation in crop and
cultivation practices. By focusing on labor-intensive crops, most of which were exported, the
Project aimed to contribute to equitable distribution of income and employment opportunities and
enhance nonoil export earnings. It also anticipated a positive environmental impact by limiting
shifting cultivation and reducing soil erosion caused by undesirable farming practices in high
elevation areas. The improvement of road infrastructure was expected to develop regional
economic activities and the improvement of living conditions in the rural areas.
3. The project scope was to establish and maintain the following: (i) about 75,000 hectares
(ha) of rubber plantations,4 and (ii) about 20,000 ha of tea plantations. It also included pilot projects
to (i) establish 10 small rubber tree nurseries; (ii) establish
about 2,000 ha of drought-resistant tree crops; and (iii)
test a low-cost approach to tree crop development by
providing farmers with incentives including provision of
improved planting materials, agro-inputs, and extension
services for the first year of tree crop development. The
project components comprised (i) plantation
establishment and maintenance; (ii) construction of roads
and buildings; and (iii) support services including project
administration and management, vehicles and equipment,
training, and consulting services. The Project was to
organize farmers’ groups to introduce them to improved
crop varieties and cultivation practices (see photo).5 The
Farm roads leading to
farmers’ groups received technical advice, training,
rubber plantations
organization, and management support from the
intercropped with paddy
executing agency. For this purpose, project management
units (PMUs) with individual offices were established for each project geographic area. The PMU
offices report to a provincial office which, in turn, reports to the project director in Jakarta.
1
Loan 1118-INO: Tree Crop Smallholder Sector Project, for $135 million, approved on 14 November 1991.
2
Loan 1232-INO: Third Local Roads Project, for $200 million, approved on 25 May 1993.
3
Actual loan usage was $113.7 million.
4
Actually, 81,000 ha were developed.
5
This is a photograph taken at the early stages of the project.
Appendix 7 97
4. TCSSP Setting. South Bengkulu District has a total area of about 6,000 square kilometers
(km2) with a population of about 370,000. The TCSSP in the Bengkulu Province has two project
units in North and South Bengkulu with six project subunits under each. The provincial government
of Indonesia advised the study team that South Bengkulu District, Talo Subdistrict, would be a
suitable study location. Talataan Subvillage of Talang Kabu Village was selected for the project site
after inspection of several similar sites. The isolated village in Talang Padang served as the control
site.
2. Description of the Study Areas in Bengkulu
5. Bengkulu Study Road. The Project did not build rural roads in this area. Instead, the
Project constructed farm roads in 1993 in this area, which connected the settlement area to the
farms. Talang Padang-Tanah Abang Road, 16 kilometers in length that leads from the farm roads
to the main road, was asphalted by the District Public Works Services in 1996. The villagers
claimed that they proposed the construction of this road in anticipation of the Rubber Plantation
Production Program of the TCSSP to allow easy market access. This road connects the project
area to Manna (the district’s capital) where there is a state road that connects to Bengkulu (capital
of the province). The people of Talang Kabu Village use public transportation on the road to go to
the weekly market at adjoining Penago Village or to go to Bengkulu, the capital of Bengkulu
Province to sell their produce.
6. Bengkulu Project Site. Talang Kabu Village has an average rainfall of 308 millimeters
(mm) per month with maximum rainfall in November. The land is generally flat. The village is
divided into three subvillages and has 10,000 ha and a population of 470 families or 2,024 people.
About 85% of families own land in the village, but 71% consider their occupation to be agricultural
laborers because of limited land ownership (<1 ha). Other sources of income include laboring as
construction, fishing, and basket weaving. The families who do not own land (15%) work as
laborers and also work on basket weaving, fishing, home industry, firewood gathering, etc. Most
landowners have noncertificate status of land ownership. Through TCSSP, 56 farmers from the
Sumber Bahagia farmer group were awarded 1 ha each. Due to unfavorable soil conditions, the
production of crops in the village is very low. Food crops include paddy, corn, cassava, sweet
potatoes, groundnut, and soybean. Tree crops include rubber, coffee, coconut, jackfruit, and
durian. Most farmers have rubber and/or coffee in their smallholdings. The prices of these are
highly affected by the fluctuations in world market prices. They are reluctant and unable to invest in
inputs as a result. The farmers attribute their vulnerability and uncertainty to the fluctuating prices.
It makes them less likely to risk increasing production if they expect that prices will fall. Then, they
are more likely to concentrate on subsistence food production.
7. Bengkulu Control Site.6 The Talang Padang Village control site consists of three
subvillages. The village is 30–85 meters (m) above sea level. The rainfall pattern is similar to that
of Talang Kabu Village. The topography is generally flat, while some areas are mountainous and
sloped. Farmers in the area are unfamiliar with the terrace farming system or any sloping
agricultural land technology. The population is 705 people consisting of 183 families. Based on
housing conditions, the houses in the area were categorized into three types: poor (22%),
satisfactory (70%), and good (8%). Most of the households (95%) have about 0.3 ha, thus their
economic condition is not so different from those families who do not own land (5%). Their main
source of income is farm labor. Their secondary source of income includes fishing in the river,
basket weaving, and gathering firewood. Those who cultivate less than 1 ha also worked as farm
laborers for additional income, therefore, it is difficult to differentiate the farmer from the laborer in
6
See Appendix 4 for access conditions.
98 Appendix 7
the village. An exception, however, are the better-off villagers who own more than 1 ha. Sixty
percent of farms produce paddy, 30% produce tree crops (rubber, coffee, durian, jackfruit, etc.),
and 10% produce cassava, groundnut, corn, and many kinds of vegetables. The village
experiences food scarcity from August to January, while extreme food scarcity exists during
November and December before harvest.
3. Third Local Roads Project
8. Description of TLRP. Indonesia’s public roads are classified as national, provincial, and
district roads. TLRP aimed at helping the Government to improve the condition of district roads.
This was based on feasibility studies financed by ADB in 1992 under Loan 863-INO and technical
assistance (TA) 921-INO. The implementation of the Project was funded by ADB under Loan 1232-
INO (footnote 2). Loan effectiveness was October 1993, and it was closed in May 2000. The
Project covered the four populous provinces of Bali, Central Java, East Java, and Yogyakarta. The
primary objectives of the Project were to improve the condition of the district road network in the
four project provinces by providing financing for rehabilitation and improvement of roads in bad
condition, support periodic maintenance, and strengthen the capabilities of the Public Works
Office. Its principal development objective was economic growth. Project components to fulfill these
objectives were (i) the rehabilitation or improvement of about 5,000 km of paved and unpaved
roads and replacement of about 880 m of bridges, and (ii) periodic maintenance of about 8,600 km
of roads. In addition, the Project was to support the procurement of road maintenance equipment;
construct workshops and laboratories; and engage consultants for the supervision of civil works,
including assistance with project implementation. In Yogyakarta, the Project was implemented in
two districts in Yogyakarta, namely Sleman and Gunung Kidul.
9. TLRP Setting. The study area of TLRP in Yogyakarta Province is located in the Gunung
Kidul District, Candirejo Village, and Panggul Subvillage. The control community is the Tepus
Subdistrict, Sumber Wungu Village, and the Kelayu Subvillage. The Gunung Kidul District has a
population of about 700,000 and has an area of about 1,500 km2. It is located in mountainous
terrain going up to 600 m above sea level. The land condition is dry, sloped, and rocky. Maximum
rainfall is experienced in February with an average rainfall of 551 mm per day. Paddy can be
cultivated in about 5% of the land; 45% of the land is used for dryland crops. After collaborating
with the Central and District Public Works Services and upon inspection of sites, the study team
decided on Semanu-Giri Panggung, a study road that transects four villages and is separated by
three subdistricts. The lack of irrigation facilities is a main problem for agriculture, and land
cultivation is dependent only on rain. The dryland paddy is planted in the rainy season.
4. Description of the Study Areas in Yogyakarta
10. Semanu-Giri Panggung Study Road. The road segment crosses the villages of Semanu,
Candirejo, and Giri Panggung. The actual length of the road is 22 km. However, funding for the
road rehabilitation project only covered 10.5 km. Among the four villages transected by the road,
Candirejo is between Semanu and Giri Panggung villages, both of which have weekly markets.
Villagers use the road for transporting their goods to the market at least twice a week. Villagers
also use the road to go to various towns and capitals for other business activities. The district road
is well maintained by the District Public Road Unit. The community also helps in its routine
maintenance by cutting grass and cleaning the side ditch and road shoulder periodically with
volunteers.
11. Yogyakarta Project Site. The Candirejo Village is 300–600 m above sea level with an
average rainfall of 2,500 mm per year. The area is hilly and mountainous, only 12% being flat.
Appendix 7 99
About 83% of the land is used for dryland cultivation, and there are no wetland rice fields at all.
Only 30% have certificate status of land ownership. About 63% of the villagers have less than
0.5 ha of land, while only 15% have more than 1 ha. The village has a labor force of 5,735 people,
of whom 26% are unemployed. Eighty-seven percent of families get their main income from
agriculture, while the rest are intermediaries and government officials involved in home industry,
labor, etc. Generally, farmers have a secondary source of income like laboring in farms, industries,
or construction work. Most farmers who have more than 2 ha of land usually have other
occupations such as businesses, intermediaries, or are involved in transportation services. Others
are government officials, health service personnel, teachers, etc. Due to unfavorable land
conditions (dry, unfertile soil), crop production in the area is very limited. The project road passes
through an area where water retention is difficult, the soil is rocky, and it is time consuming to
prepare fields for cultivation. Consequently, much of the cultivation is limited to cassava,
groundnuts, and some banana. Some rice is cultivated for consumption, but diversification due to
road improvement into higher-value cash crops would be difficult in these production conditions.
12. Yogyakarta Control Site. There are about 1,500 families in the Sumber Wungu Village
(1,833 ha). From the whole population, 110 people are illiterate and 223 people finished
elementary school. Due to lack of electricity in the area, 88% of the families use traditional lamps,
while 81% use firewood for cooking. The poor accessibility in the area limits the supply of
petroleum. It is similar to the project site area in terms of culture, homogeneity of the community,
history and background, land condition, and the types of crops. The Kelayu Subvillage is 6 km from
the asphalt road. The villagers have to walk through rocky roads to reach the central village. The
topography of the area is similar to Candirejo Village. Most of the area is hilly. The soil condition for
agricultural production in the area is classified into fertile (30%), sufficient (8%), and unfertile
(62%). Unfavorable soil conditions limit crop production in the area. The production is mainly fruits,
corn, and cassava.
13. Institutional Arrangements. The roads are governed by the Public Works Department,
while transport services are the responsibility of the Communication Department. This includes the
local government (province, district, and city). The transport network is divided into the main,
branch, and local roads. The local primary roads are under the local government. The financing of
roads is dependent on the status and function of the roads. Therefore, local primary roads receive
different development actions compared with local secondary roads. The process of determining
the feasibility of a road involves considering many factors such as traffic volume, road links, and
road treatment.
14. The commonly used vehicle in Bengkulu

project site is the motorcycle for short distances
(Appendix 1, Table A1.1). Other vehicles include
vans and minibuses for long distances and
nonmotorized vehicles such as bicycles. Many
people, however, still prefer walking short distances
and ride the bus for long distances. The dominant
mode of transport is the motorcycle in Yogyakarta
project site as well. In addition, vans, minibuses, and
buses are used. Few bicycles are used. In contrast
in Sumber Wungu control area, motorized vehicles
Community in Yogyakarta control site are very rare in the village. The villagers do their
maintaining intra-village road daily activities mostly by walking. The local road
100 Appendix 7
within the village is well maintained by the community. It was made of gravel using the Telford
construction method, by the local people themselves. The village people constructed it using the
self-help method without government support. Unfortunately, the community does not have the
ability to upgrade the gravel road to asphalt and have requested funding support from the
authorities.
1. Bengkulu Study Areas
15. The villagers’ focus group discussion in Talang Kabu concluded that around 75% of the
villagers are poor and that the poor lacked the following: source of income, food, education,
information, and housing (Appendix 3, Table A3.2).7 Through focus group discussions, the
following causes of poverty in the area were identified.
(i) Job opportunities for farm laborers are not available all year but only during the
planting and harvest season. Lack of skill and experience makes farm laborers less
competitive inside and outside the village. Unskilled village youth often cannot
compete with other more skilled and educated workers, even if they migrate.
(ii) The access of rural farmers to current or advanced information on farm technology
is very limited. The activities of two existing (TCSSP) farmer groups are mainly
focused on selling their product and not on accessing information technology. Crop
pests and diseases are also a major problem in these farms.
(iii) Of the many reasons for low crop production, lack of land to cultivate is critical. The
lack of funds for farm inputs also contributes to low crop production.
(iv) The lack of market information access is also a problem. Farmers’ groups need to
link up with other farming groups outside the village to establish a network and have
first-hand information on market prices to have a better bargaining position.
A men’s focus group A women’s discussion group

discussion in Talang Kabu in Talang Kabu
16. The sources of income of villagers, poor families, and nonpoor families are summarized in
the table below:
7
See Indonesia country report for more details. ADB. 2002. Impact Evaluation Study of Rural Roads on Poverty
Reduction: (ADB TA 5979-REG). Manila.
Appendix 7 101
Source of Income of Poor and Nonpoor Families at Talang Kabu Village

Sector Poor Families Nonpoor Families
Labor Farm and nonfarm labor Not a source of income
Agriculture Any kinds of beans and resin for weaving Paddy, coffee, rubber, and coconut
Livestock Chickens, ducks, goats (few) Cows, goats, chicken, and ducks
Fishing Fish Not a source of income
Processing sugar Palm sugar Palm sugar
Shop, kiosk Shopkeeper Owner
Skills used Make fishing equipment, basket making, Skills for outside salaried employment
broom making from coconut leaves, and or business skills
making fishing nets
Source: Participatory rural assessment activities in Bengkulu, Indonesia.
A very poor family has an average monthly income of Rp55,000–Rp67,000. The head of the family
usually works as a farm laborer for 8–11 days a month. The rest of the income is acquired from
other odd jobs or borrowing money from a neighbor or intermediary just to put rice on the table.
The average family spends around Rp85,000 per month for very basic needs. Both the very poor
and poor families face food insecurity. Their income is often insufficient so that the family’s basic
food requirement is not met. Many times, these families only eat cassava, banana, and/or corn for
breakfast and dinner. At times, the family does not get a chance to eat rice for 2–5 days. The poor
families, however, claim that they are used to these conditions.
17. In Bengkulu, intermediaries (tokeh) and traders

effectively control market exchange in paddy and
rubber with debt relationships. It is increasingly difficult
for the poor to escape from these debt relationships as
indebtedness increases year on year. Numerous
projects to help the poor have been implemented in the
village before. However, the villagers were not included
during the planning of these projects. As a result, the
village investment plan of the community did not
identify the common problems and concerns of the
poor. During the implementation of road projects
(district road rehabilitation and farm road construction),
poor families served as laborers during the Poor laborer headloading paddy
construction phase, which earned them Rp2,500 per from farm to village settlement
2
km or Rp75,000 per month. Since 1999, the
Government, through the social safety net program (SSNP), has been supplying and selling
cheaper rice to the poor and poorest families.
18. Most of the houses in the village were dilapidated and of substandard living conditions.
Walls were made from old bamboo, there was no flooring, and the roofs were made of dried grass.
Only a few houses had tile roofs. The poorest houses in the area were extremely unfit and unsafe
to live in and were characterized by the absence of windows, leaky roofs without ceilings, and a
single bedroom. Most of these houses also did not have electricity because it was too expensive.
Only a handful of children from poor families are able to attend secondary school, and more than
75% stop studying after elementary school. Financial constraints usually hinder these families from
continuing the children’s education. Moreover, most of these families would rather have their
102 Appendix 7
children work as laborers instead of continuing their education because this would add to the
family’s income.
19. Talang Padang is a remote village. Based on the village’s focus group discussions, around
85% of the villagers are considered poor. The following criteria for characterizing the poor were
identified during the discussions: (i) physical health, (ii) living conditions, (iii) financial capability,
(iv) food, (v) means of transportation, and (vi) level of education (Appendix 3, Table A3.2). The
focus group discussions identified similar causes of poverty as the project site, but these are more
acute due to lack of accessibility.
2. Yogyakarta Study Areas
20. The study community at Candirejo Village found that the community can be classified into
four classes: very poor, poor, sufficient, and rich. The focus group discussion reported that 20% of
families living in the village were very poor, while 60% were classified as poor. The villagers also
identified five main class indicators: (i) farmland ownership, (ii) cattle, (iii) house condition,
(iv) appliances, and (v) transportation (Appendix 3, Table A3.2). Villagers who own less than 0.5 ha
of farmland were categorized as poor. The small piece of land is not enough to produce a high
income for the family. However, according to the discussion on problem analysis, the size of land is
not the main cause of poverty. The discussion on problem analysis identified five main causes of
poverty:
(i) Low quality of produce/commodity due to farmers’ lack of skill in cultivation, pest
control, and postharvest handling.
(ii) Lack of access to market information and competitive power with other markets.
(iii) Lack of roads and mode of transportation prevents farmers from having access to
markets, and thus are taken advantage of by intermediaries.
(iv) Farmers cannot plant high-value crops.
(v) Sources of income are limited. The absence of irrigation systems further limit
farmers from planting other viable food crops.
21. Despite their collective efforts, the poor families’ generated income is still insufficient to
sustain their needs. The very poor families (often with women acting as head) only have a monthly
income of Rp40,000–Rp60,000. For example, a mother works for 4–6 times a week (half day), and
her daughter contributes Rp30,000 every 2 months working as a domestic help. They borrow about
Rp20,000–Rp40,000 from neighbors and pay back this debt by working as labor at the lender’s
farm. The family spends around Rp85,000 a month to buy rice, vegetables, water, soap, and
gasoline for a lamp. Poor farmers who own less than 0.5 ha of land have an income of Rp45,000–
Rp120,000 per month. This income comes from cassava, corn, and groundnut production from
their own farms and working as a farm laborer for other farmers. Generally, they borrow around
Rp10,000–Rp50,000 from intermediaries and other farmers around the village. To pay their debt,
they have to sell their crops to the intermediaries at a much lower price than usual or they repay
other farmers by paying cash or working as laborers. The family’s expense for basic needs is about
Rp150,000 per month to buy rice, vegetables, side dish (viand), water, soap, and gasoline for a
lamp.
22. The poor and very poor families face food insecurity. A very poor family uses 50–70% of
their monthly income to sustain their basic needs. They only eat rice during lunch and eat cassava,
bananas, and corn for breakfast and dinner. Sometimes, these families even go for 4 days a month
without eating rice. The condition of the poor is not so different from that of the poor family. The
poor family uses at least 30–80% of its monthly income on basic needs. Generally, the poor family
Appendix 7 103
chooses to borrow money from intermediaries and pays them in two ways: at an interest rate of
25% per month or by selling their crops to the intermediary at a much lower price than usual.
Borrowing money seems to be the only way in which these poor families survive. They tend to
borrow money to sustain their daily needs and are not able to invest in other income-generating
activities.
23. Most of the poor and poorest families only attended elementary school, some do not even
complete that. Since they lack skills and education,
the poor and very poor families are generally
incapable of working outside the farming sector. A
few of them tried to establish small enterprises by
making banana and cassava chips, however, their
lack of capital and marketing skills meant that these
businesses were short lived. Banks, village
cooperatives (such as KUD), and other sources of
capital are located in the capital subdistrict, which is
about 7 km from the village. Further, these institutions
require formal requirements from borrowers (i.e.,
collateral, a formal permit letter from the village head,
A woman from a very poor a copy of an identification card, and proof that the
family collecting firewood business is running well). These requirements are
unfamiliar to the poor and very poor, and they have to
rely on the moneylender instead for their financial needs.
24. The poor and very poor also do not have information about capital resources, technology,
marketing, and other information to support their income-generating activities. They are like
second-class citizens in the village community. Basic and extension services from the government
do not seem to be reaching them. During village meetings, the institutional representatives tend to
marginalize the poor. They do not disseminate the information or decisions made in village
meetings to the poor. Preparations for village development planning is normally done during
community meetings at the village office. Often, the poor and very poor are not involved during the
village planning meetings, and plans do not include means to reduce poverty. Ironically, this does
not seem to bother the poor because they feel that village planning does not have any effect on
them nor do they perceive it as any of their business. They are usually only involved in village
activities as hired labor. For example, they worked as laborers during the district public works road
rehabilitation project. The poor are also not involved in agricultural development, extension, and
health programs. Since 1999, the Government, through SSNP, has been supplying the poor with
low-priced rice (50% off the regular market price).
25. The household conditions of the poor are very similar to those of the project site. The poorly
constructed houses result in these family members becoming ill often. Generally, instead of going
to the village health center, these families rely on indigenous medical practitioners or faith healers
to receive medical treatment because they cannot afford the medical services provided by the
village health center. Only a handful of children from poor and very poor families are able to attend
secondary school. More than 75% stopped studying after elementary school, similar to the
Bengkulu project site. In 1998, a traditional farm labor group was established (not by the project) in
the Yogyakarta project site, comprising members mainly from poor and very poor families. It
involved working in each other’s farm in rotation and worked well during peak periods. Very poor
laborers were paid cash for their work. In emergencies, the very poor ask for the payment even
before they start a job.
104 Appendix 7
26. The control community in the Kelayu subvillage of Sumber Wungu Village classified the
community into four classes: very poor, poor, sufficient, and rich. Their focus group discussion
revealed that 25% of the families in the village were very poor, while 45% were poor. The following
indicators determined the family classes: (i) the type of lantern used; (ii) house condition; (iii) land
ownership; (iv) cattle ownership; and (v) ownership of transportation, radio, television, and other
facilities (Appendix 3, Table A3.2).
27. The subvillage is a remote area. In this large subvillage, there are only three small shops
with limited capital (Rp200,000–Rp450,000). Most farmers cultivate their crops with minimal inputs,
while a few rich farmers have cattle and inputs. The agricultural extension services rarely come to
the village because of unavailable transportation. Only one truck comes to this subvillage once a
week during market day because the operators choose to go to other villages with better roads.
The transportation problem directly influences the marketing of agricultural production. More
importantly, this also greatly affects the price of crops and the bargaining position of farmers who
sell their crops. Only a handful of intermediaries from outside the village come to collect agricultural
products, thus there is not much competition. All of the intermediaries are outsiders and the price
of commodities is dictated by the intermediaries.
28. Three priority problems were found common between men’s and women’s discussion
groups: (i) limited water supply, (ii) bad road conditions, and (iii) no electricity. There are no
sources of water in the area. Each family has a water tank under the roof of their houses to collect
the rain, which they save as drinking water to prepare for the dry season. This drinking water is
sufficient for 1 month after the end of the rainy season, after which they need to buy water from
outside the village. Three mobile tanks from the local district government supply the water every
2 days. Villagers usually buy the water in groups.
29. There is no asphalt road or public transportation in the Kelayu subvillage. The villagers
have to walk 7 km to go to the central village or 7 km to reach the asphalt road where public
transportation is available any time. Due to this condition, the supplies from outside the village are
very limited. The district agriculture department only conducts extension services twice a year.
Three small shops are in the subvillage. These, however, do not have goods available all the time.
Schoolteachers also have to walk 7 km from the main asphalt road to the village school every day.
During the rainy season, the teachers often do not go to school. The villagers use traditional lamps.
They have no access to daily information from outside the village because there is no television.
Only a few rich families own a radio. Small businesses that would rely on the use of electricity are
also not an option for villagers.
D. Impact of Rural Roads and Transport
1. Distribution of Benefits from the Road
30. Economic impacts from the TLRP road rehabilitation in Yogyakarta seem to have accrued
primarily to the better off in the study community and not to the poorest. There are a number of
reasons for this. First, it was those with some form of capital to invest initially, even if at fairly
modest levels, that were able to take advantage of the opportunities that improved access to
outside markets and networks provided. The case studies of Mr. Supardi and Mrs. Peti (Appendix
8, Boxes 14 and 17) show this clearly. The improved road essentially offered an outlet through
which capital could be invested through the development of small businesses or the sale of
agricultural surplus. For marginal farmers or for those dependent upon wage labor for their survival,
the road offered few opportunities as they were unable themselves to invest in any way (Appendix
8, Box 3).
Appendix 7 105
31. Second, the travel patterns of the poor are primarily village centered. The poor have little to
sell and little money to buy so they seldom travel outside the community, except to make trips to
the weekly market once a week or twice a month, or if emergency necessitates an urgent trip
outside. For them, improved access is important only in emergencies, otherwise their travel
patterns continue much as they did before the road rehabilitation. This is the case in the TCSSP
study site for travel outside the village or immediate locale. In contrast, it is the traders, merchants,
and intermediaries in both locations who make most use of the road and depend heavily upon
rapid and regular access in and out of the study community for their daily (or regular) economic
activities.
32. Third, the better off in the community have the ability to access transport services and are
more likely to own their own transport, particularly bicycles. During the study activities with villagers
in both Bengkulu and Yogyakarta, owning a bicycle was identified as a key well-being indicator.
Personal mobility and frequent travel outside the community were explicit aspects identified by
villagers as distinguishing the poor from the better off.
33. The very poor in both Bengkulu and Yogyakarta are unlikely to regularly sell surplus
agricultural produce unless illness or some household crisis requires it. In these cases, food
security for the household is seriously compromised as a consequence. The very poor are often
engaged in both study areas in selling firewood, grasses, and other forest products, which they can
gather their incomes. The better access provided by the rehabilitated road (Yogyakarta) and
construction of the road (Bengkulu) has opened up better opportunities for selling these products
on a more regular basis. More buyers and intermediaries are now visiting the communities, and
this enables villagers to sell at a higher price.
34. The basic farm access roads provided under TCSSP in Bengkulu are quite different from
the asphalt road in Yogyakarta provided under the TLRP. These basic TCSSP roads seem to
serve an important function for the poor, which demonstrates how reliant they are upon intra-village
travel over external travel. The roads are used primarily to access farmland, and they are
inaccessible to all but the most determined motorized vehicles. For poor villagers who have
traditionally been employed in portering crops from the fields to the roadhead, the better farm
roads have significantly improved their income-earning opportunities. Previously, they would have
to carry paddy and other crops on their backs, which was very difficult and tiring work. Now, they
can use handcarts to transport the crops, carry more with less effort, and make more trips. They
now also have more time and energy to do other things (Appendix 8, Box 5). However, the
improved access and better opportunities for income earning through this activity has led to an
increase in competition among laborers for this kind of work. They can now use handcarts to
transport goods along these access roads to the main roadhead. This means that there are
improved opportunities too for intermediaries and traders to recruit labor to bring in produce from
the field even during harvest times, which was previously impossible to do (Appendix 8, Box 18).
2. Provision of Transport Services
35. The availability and frequency of transport services in both the Bengkulu and Yogyakarta
study areas have increased. In Yogyakarta, there has also been a significant reduction in transport
prices for passengers, though this is less clear on the asphalt road in Bengkulu. One reason for
this may be the proximity of the Yogyakarta project site to the main town of Yogyakarta proper; at
the Bengkulu study site distances to major trading centers are much greater, and the potential for
competition among operators is, therefore, reduced. The Yogyakarta study road is also more
closely connected to the main network in the area. Nevertheless, improved services have resulted
106 Appendix 7
in both areas as a result of the road improvements. For vehicle operators in Yogyakarta, they have
lost their monopoly position through the wider availability of services, but have been more than
compensated. They can no longer run a completely full vehicle, but make more trips per day as a
result of the better road, and their vehicle operating costs are significantly reduced because of the
better road surface.
36. Evidence from the control area in Yogyakarta appears to confirm that transport providers
gravitate toward areas with better roads, irrespective of the high agricultural potential of an area
and the consequent high potential demand for their services. The income from other routes is
better as the road is better; the risk to their vehicles is much higher on the bad roads. As a result,
more inaccessible areas are left with an infrequent service usually run on a monopoly basis.
Operators can charge what they like as they face no competition. In the case of intermediaries
buying agricultural produce, this also means that they can largely dictate the prices at which they
buy, as farmers have no option of selling in a wider market outside the community.
3. Access to Services for the Poor
37. Improved access to services is a critical nonmonetary impact for the poor in both study
locations and is probably the major benefit to them from road improvements. Better access to
services means that the poor can travel outside to summon medical services in an emergency or
transport the sick or injured to medical facilities more easily. It also means that providers of state
services such as health, education, and extension particularly, have improved mobility and can visit
project areas more quickly and more regularly. Case study evidence form both locations shows
clearly how improved access and mobility have impacted on services in both communities.
38. The improved asphalt road to the study area in Bengkulu and the rehabilitated road in
Yogyakarta ensure all-season access for teachers to the village. Previously, their visits were
restricted by the rains and were dictated by the irregular timetables of the vans offering the only
means of transport into the villages. Now, better transport services and better roads have
shortened their travel times and improved their attendance. District officials also report that it is
easier to monitor the provision of services in the field with better roads. The case study of Mr.
Sumianto (Appendix 8, Box 22) the extension worker in Yogyakarta, shows how he is now able to
meet 90% of his work targets, against only the 40% that he was able to complete before the road
improvements. Mrs. Suryatini, the local nurse, also reports a qualitative shift in the level of service
that she is able to provide to the community because of the better road (Appendix 8, Box 21).
Officials report higher motivation to do their jobs and some have been able to take up secondary
occupations to supplement their incomes because of the time saving through better transport.
4. Long-Term Sustainability and Community Involvement in Roads
39. The TCSSP roads in Bengkulu were originally constructed through the active involvement
of community members, and the farmers’ association continued to maintain the road for 2 years
following construction. However, maintenance was neglected thereafter as it was felt that no
perceivable benefit would come to farmers because of the road as rubber trees grew and rice
could no longer be intercropped between them. The road has consequently reverted to a
footpath/track. At the Yogyakarta study site, the community had little involvement in the road
project; all construction involved machinery and outside labor. The prospects for some community
engagement in the maintenance of the road are, therefore, now slim.
Appendix 7 107
40. In both control areas for the study, though, it is clear that there is a great deal of potential
for community engagement with the roads. Villagers at Bengkulu recognize that their lack of road
access is a critical restraint to the socioeconomic development of the village; and in the Yogyakarta
control area, villagers have been building and maintaining their own village roads for the past
30 years. Community mobilization and involvement, therefore, do offer the prospect of devolved
local management of the roads, enhanced community ownership, and better prospects for long-
term sustainability of road investments through active community maintenance. For the poor,
regular employment in road maintenance would offer an important supplementary source of
income through which they could begin to diversify their livelihoods and attain greater economic
security.
5. Negative Impacts
41. The improvement of roads has encouraged overspeeding by motorists, and thus the
frequency of accidents has increased. Most are motorcycle accidents involving youth. Because
young students have better access to and communication with others outside the village, they have
also been exposed to drugs and other vices. Robbery is also gradually becoming a problem. Like
villagers and other motorists, criminal elements have access through these roads, which they can
use for easy getaway.
Appendix 8 108
HOUSEHOLD CASE STUDIES FROM ALONG THE PROJECT ROADS
1. This appendix shows how some people have been able to use the road
improvements to make their lives better and have the potential to come out of poverty. Yet,
other case studies show that the situation is still quite hopeless for other families despite the
road improvement.1
A. Very Poor Households Unable to Make Use of the Road to Move Out of Poverty
2. The following case studies show households in extreme poverty from different project
sites. They have still no hope of being able to get out of their poverty status. They have little
idea of how to use the improvement in the road condition to better their lives.
Box 1: A Very Poor Family—Sorsogon, Philippines

Mr. Nestor Grita and his wife Ms. Luz Luviria live on a
small piece of land beside the project road in Barangay
Palale (Sorsogon project site). They do not own the land
and have been squatting there for the past 12 years.
Their house is a basic bamboo-walled, mud-floored
structure, with a grass thatch roof that leaks during the
rains. They have eight children (5 months–12 years).
Because of income constraints, they had home births
instead of hospital births. The eldest attends the high
school in the neighboring barangay and the other
school-aged children attend the barangay elementary
school. Education is a major expense for the household
and costs P250 a year for the eldest and P50 each for
the other four. They recognize the importance of
educating the children, and their major aspiration is that all of their children finish at least high school.
Nestor works as a laborer in the community, processing coconut to copra. He works for two tenant farmers but
the work is not regular, perhaps 4 days a month on average. He is paid on a contract basis, between P100–P150
per day. They grow some gabi (root crop with edible leaves) and yams on a small piece of unused land
(0.25 hectares) some distance from the house and must buy everything else they need. They sell gabi if there is
a surplus, and Nestor does other laboring jobs when there is an opportunity. Luz does not work outside the
household. They keep some small animals, the major source of income, to cope with crisis. They currently have
1 hen, 5 chicks, and 1 pig. Nestor recently sold his rooster for P80 to have his child treated at the clinic in San
Francisco. The health of the children is their major concern. They hope that they do not become seriously ill, as
they cannot pay for treatment. They also fear the occurrence of another typhoon, which damaged their house
2 years ago. They have open credit with a storekeeper in the village and pay no interest so long as the bill is paid
after some time. Prices for goods in the village are, however, more expensive than in the Bulan market center.
Nestor travels to Bulan only when he has some money for groceries. He had money over the past week so he
bought rice, coffee, sugar, and milk for the baby. Luz travels about once a month to Bulan, but usually only if
there is some need such as provisions or something for the children.
They recognize that more transport is available along the road now since it was rehabilitated, but it has not
become any cheaper to use. They are, in any case, only able to use these services intermittently when they have
money for purchasing essential items.
1
Permission was specifically sought from those who have been featured in the case studies to identify them by
name and include photographs.
Appendix 8 109
Box 2: A Very Poor Family—Negros, Philippines

Mr. Romulo Olarte’s family is typical of families highly dependent on the sugar plantation for livelihood. He and
his wife Merlinda have been living in a small hut on a piece of land owned by the hacienda beside the project
road in Barangay Magallon Cadre for 26 years. The walls of their house are made of bamboo, floor of packed
clay, and grass thatch roof that leaks during the rains. They have 10 children (2–17 years). Two of the children
are now living and working permanently in another province to help ease the family’s financial burden. Because
of lack of income, Merlinda gave birth to all her babies at home with the aid of a “hilot,” a village midwife, instead
of at the hospital. Out of the five children of school age, only two are attending elementary school in the village:
one in grade 3 and one in grade 1. Three are not attending school: one daughter (10 years old, grade 4) and two
sons (15 years old, grade 6 and 17 years old, grade 4). The two sons of working age stopped their schooling to
help augment the family income. It is clear that poverty disrupts the educational opportunities for children of poor
families. Once the children reach working age, they give up schooling for the income they will bring to the family.
Husband and wife recognize the importance of educating the children, and despite their financial incapacity, they
at least aspire that all their children who are still interested in going to school can complete at least high school
education.
They do not own land and so Romulo works as a laborer in the haciendas in the village and surrounding areas,
while his wife mainly does household chores. Work is always available for him from September to March during
huge planting and harvesting of sugarcane in the haciendas. During these months, he and his two sons together
earn P200/day to P250/day. Due to the relatively large household size, they can barely save money. In the off-
season (April–August), Romulo and his family experience food scarcity. They have considerable difficulty in
getting work at this time and have to buy everything they need. August is the most severe month for this family to
experience food scarcity due to continuous and heavy rain. The family is still fortunate that they have two family
members sending them money especially during the off-milling season to bridge their needs until the milling
season. It is clear that family networks, especially in times of difficulty, are critical in helping the poor overcome
their poverty and offer a social safety net in times of hardship. Most of the poor families in the village are
fortunate that a number of small variety stores are ready to provide them with credit during off-season months
and charge no interest so long as they are able to pay their bill whenever they have money to do so. Prices for
goods in the village are, however, more expensive than in the market center of Moises Padilla. The health of the
children is their major concern, and they hope that they do not become seriously ill as they have no money to pay
for treatment. They keep some small animals, currently six chickens, as security for when money is needed to
cope with a crisis.
Romulo’s travels are highly confined to the village, mainly on foot and sometimes by tricycle as most of the
haciendas are within a 10-kilometer radius from their house. They recognize that more transport is available
along the road now since rehabilitation, but it has not become any cheaper to use. They are only able to use
these services intermittently when they have money for purchasing essential items. To save on transport costs,
only his wife travels about once a week to Moises Padilla usually only if there is some need, mainly for
purchasing provisions or something for the children.
Box 3: A Very Poor Family—Yogyakarta, Indonesia

Mrs. Tumilah is a 50-year-old widow, living with her two daughters of 27 and 22 years old and one son of
13 years old. This is really a very poor family with very bad and unhealthy house conditions, 4 meters x 6 meters,
tile roof, wood wall, and ground floor. She works as a farm laborer; for additional jobs, she is a firewood collector.
The oldest daughter works as a household helper in Yogyakarta, the capital of the province.
The family has 0.2 hectares of land, which is planted cassava and groundnut. The land is not very productive and
is on a slope. Because she is 50 years old, Mrs. Tumilah cannot do hard jobs. She only does planting, weeding,
and harvesting. Job opportunities are not always available through the year. Generally, she works 2–3 days per
week with 8 hours work per day or 8–12 days per month. She gets Rp2,500 for 4 hours work from 7 am to 11 am
or from 1pm to 5 pm. To avoid the cost of lunch, the farmers are not interested to recruit her for a full day. From
this job, she gets an income averaging Rp40,000–Rp60,000 per month.
Her normal expenses are Rp3,500 per day for her family’s basic needs such as rice, vegetable, etc. so that she
needs Rp105,000 per month. She gets additional income from selling firewood, which she collects from her farm
or nearest uncultivated land. Other additional income cannot be estimated. Sometimes, she gets Rp50,000 per
month, and other times she only gets Rp10,000 per month. Generally, their income is insufficient to satisfy the
family’s needs per month. To meet the family food shortage, she borrows money from her neighbor farmer and
pays back by working at the farm of the lender, or borrows from the intermediaries at an interest rate of 20% per
month. She also gets additional income from her oldest daughter. She still remembers that she was not involved
Appendix 8 110
Box 3—Continued
when the road was being rehabilitated because the project did not want to use women. Actually, she was ready
to work on the road or any other construction.
Road and Transportation Use. Although job opportunities are not always available, she is not interested in
looking for a job outside the village. She works inside the village so she does not need transportation for work.
She needs transportation only when she goes to the market weekly to buy provisions for her family. She realizes
that it is easier for her to go to the market after the road was rehabilitated because public transportation is
available every time. Before the road was rehabilitated, the round trip fare to the market was Rp4,000. Now, it is
just Rp1,000. The other benefit from good road is it is now easier for her to sell the firewood because many
intermediaries comes from outside the village to buy the firewood at a higher price than before the road was
rehabilitated.
Conclusion. She understands that it is a better situation now after the road is rehabilitated, but it does not give
many benefits for her job as she works inside the village. She benefits from the good road only when she goes to
the market, while firewood has become more expensive due to higher demand. She does not have any idea how
to use the road to make her life better.
Box 4: A Very Poor Family—Kurunegala, Sri Lanka

Ms. Nandawatie and her husband live with their two children (7 and 12 years
old) on a small piece of land at the far extreme of the Nugannoruwa Village
bordering the tank where they have been living for 35 years. When they first
came to the area, they were given permission by her great uncle to squat on
the land, which was a state land, and were later given a license.
Nandawatie’s husband works in the area as a laborer, and she tends their
garden crops while working as a laborer when opportunities arise.
They have a small well at the bottom of the garden and grow an extensive
variety of crops for their own consumption and for sale. She grows brinjals,
which she sells for Rp20 per kilo and chilies, which cost Rp16 per kilo. They
use their coconuts for oil and she cultivates beetle, which she sells for
Rp25 per bundle. They grow banana plants and sweet potato and buy rice
from the local store. Most of their income comes from laboring wages. Her
husband is paid Rp150 per day plus food. Whenever they need extra
income, they sell vegetables from their very productive household plot. They
are wholly reliant upon their garden and laboring for their livelihood.
Nandawatie seldom travels to the nearest town at Nikaweratiya. Her husband cycles the area in search of work,
and he goes to the town twice a week to collect groceries. They have one bicycle but would like to get another as
her daughter likes to use it too. She acknowledges that goods are cheaper there than in the village. For example,
kerosene oil is Rp20 a liter in Nikaweratiya, and Rp20 a bottle in the village store. She travels once every 3–4
months to the town when she needs to buy clothes at the market or when she needs to go to the hospital with the
children. She burnt her hand badly on the kerosene lamp during the night some time ago and had to wait to catch
the bus to the hospital in the morning as her husband was working away.
Box 5: A Very Poor Family—Bengkulu, Indonesia

Mrs. Kustina lives with her husband Mr. Marin (household head) and three sons (18, 15, and 9 years old). The
living conditions of this very poor family are bad and unhealthy (4 meters x 6 meters, grass roof, bamboo wall,
and ground floor). They earn their living either by farming, collecting firewood, weaving baskets, or sometimes
catching fish using a traditional net. These means of livelihood are all temporary and incomes are unpredictable.
They always work hard to get an average net income of Rp100,000 monthly. The oldest son completed
elementary school but stopped school for two reasons: (i) the family could not pay for his schooling, and (ii) the
oldest son has to help his father earn a living.
Mrs. Kustina said her family did not benefit from the district road that crosses the village because they do not go
outside the village for work or other needs. As laborers, they benefit more from the farm road (Tree Crop
Smallholder Sector Project road). Before the road was built, they could only carry 35 kilograms of crops by pull
cart from the plantation to the farmer’s house, and this would usually take about an hour. In half a day, they made
Rp2,000 only. With the construction of the farm road, they can now bring about 100 kilograms of crops in
45 minutes. This results in about Rp6,000 for half a day’s work. The rest of the day is spent weaving baskets,
Appendix 8 111
Box 5—Continued
whereas before the Project, they would already be too tired to work. It is also easier for them to go to the farmers’
rice field and plantation for other kinds of jobs and to bring firewood from the forest.
Before the farm road was built, they were more interested in basket weaving, catching fish, and collecting
firewood than working as farm laborers. As farm laborers, they received a maximum of Rp3,000. After the farm
road was built, they preferred working as laborers since they earn about Rp6,000 per day. They do not want the
farm road to be asphalted because when the farmers use pick-ups to transport their crops, the opportunity for
carrying will be reduced and there will be less work.
The family income did not increase as the traders or farmers did. Job opportunities for farm labor have not
increased, but the competition has increased because it is easier to transport crops by pull cart. The demand for
basket production is also not as good as before because people do not use baskets anymore. The best source of
income for the family now is collecting firewood. Mrs. Kustina said that she knows of six other families who are in
a similar condition to her family.
B. Poor Households Making Use of the Road with Potential to Move Out of
Poverty
3. The following case studies show some families using the road as a means to better
their lives.
Box 6: A Household Making Use of the Road and with the Potential for Moving Out of
Poverty—Sorsogon, Philippines
Mr. Ely Bon and his wife Ms. Luz have seven children. They have lived in their current house beside the road for
3 years. They do not own the land, but have built a
simple two-room hut on vacant land. Their home has a
tin roof and wooden walls. They moved to this part of
Barangay Palale 7 years ago from Patag, a remote
part of the barangay. They thought that opportunities
would be better here closer to the center of the
barangay though they say that they did not consciously
move to be close to the road. They previously squatted
in an adjacent plot for 4 years, but were moved on and
settled here.
In Patag, both Ely and Luz had been laborers. Ely now
works periodically as a laborer, processing copra for
tenant farmers in the vicinity, but the work is not
regular. He also mines sand from the nearby river and
sells to a local storeowner for P3 per can. He had
three orders for 50 cans of sand last year. The sand is transported to Bulan, the nearest marketing center, and is
used locally in house construction. Ely is a carpenter and supplements his income by making furniture for local
people. He also labors for house construction. He had two orders last year for cabinets, for which he was paid
P1,300 each. The cabinets take him about 1 week to make, and he charges extra if he provides the materials.
The business is more profitable than unskilled laboring. Since they have no land to cultivate, they receive
vegetables from neighbors but must otherwise buy everything from stores. Ely travels at least weekly to the Bulan
market center to buy foodstuffs (rice and fish). They prefer to buy from Bulan as the prices are cheaper than in
the village, though they buy small quantities of things they need from the local store periodically.
Luz has recently started a business trading in bananas. She buys from farmers in the interior of the barangay in
Patag and carries them down to the road. From there, she hires a tricycle and transports them to Bulan where
she goes from shop to shop selling them. She started this business only 4 months ago, and goes approximately
once or twice a month when she has capital to buy bananas. She doubles her investment as she buys the
bananas at P300 and sells them in Bulan at P600. The transportation cost by tricycle is P74 for herself and the
freight. As far as she knows, she is the only woman in the barangay doing this and has been helped in
developing this business by the ready availability of tricycles in the village now. This business has potential, she
feels, as she can easily sell all the bananas she brings to Bulan. This trade is the primary reason for Luz to travel
outside the barangay; otherwise, she only goes to Bulan if the children require medical treatment.
Appendix 8 112
Box 6—Continued
Although still very poor, their household income has fluctuated significantly over the past year. They periodically
receive cash through Ely’s cabinet making and now increasingly from Luz’s banana trading. This allows them to
recover from debt and even buy some household possessions. Ely bought a karaoke machine some time ago,
although it has not often been used and is currently pawned to a neighbor in the village as they needed money
for food and medical expenses. They invest in their children’s education, aspiring for them to finish high school.
Ely finished only grade 5 and Luz grade 6.
Poverty—Kurunegala, Sri Lanka
Mr. Piyadasa, 50, his wife Karunawathie, 49, and their 18-year-old daughter make up their household. They are
one of the very few families in Nugannoruwa who have completely given up farming despite owning both paddy
and chena lands. The primary occupation of the
family is a confectionery enterprise. About
10 varieties of sweets are made and marketed by
the family. Piyadasa has the responsibility of
traveling to Nikaweratiya to purchase all the
necessary raw materials such as rice, sugar, oil,
etc. He does not have to transport firewood as it is
delivered to their house as they buy in bulk.
Karunawathie is responsible for the preparation of
the sweets. They take it to the market to sell.
As can be expected, transport is crucial to the

functioning of the business. Piyadasa makes over
four trips a week to Nikaweratiya to buy raw
materials. Karunawathie and Piyadasa together
make three trips a week to market the produce. As
such, the single most important factor in making the business viable is access to transport. Their house is
situated just by the road, which helps the process of moving raw material and finished products. For all trips to
Nikaweratiya, Piyadasa uses the public bus. Though there is no official bus stop by his house, the bus driver
always stops by his compound to enable easy loading and unloading. An extra payment of Rs20 is made on days
that he carried more than one gunny bag of goods. When traveling to the market with the products, they prefer
the use of a 3-wheeler to Nikaweratiya and Anamaduwa markets. They spend a total of Rs400 for the two trips.
This is a special rate fixed with their regular 3-wheeler man. The return trips are made by bus as they have no
cargo. This costs them about Rs75. The trip to Puttlam is done purely by bus as it is a 45-kilometer trip. They
take the 6 am bus to Nikaweratiya and then take the Puttlam bus. The fact that the bus leaves early and is very
reliable is a critical factor in their accessing the larger market in Puttlam.
The family is currently thinking of expanding the business by employing village women to make the sweets and
visiting more markets. A major constraint to further expansion is Karunawathie’s health. Two months ago, she
was diagnosed with a serious spinal problem. However, they hope to cope with it by hiring women to make the
sweets.
Their clay and cadjan house is in a very poor condition. There are no doors or closed-off spaces in the house.
There is minimal furniture in the household: 12 plastic chairs are stacked in a corner of the front space and the
single bed is in a back space of the house. If judged visually, it would seem that the quality of life of this
household was lower than the village average. However, Karunawathie felt very strongly that it was actually
better than the neighbors’. Despite the fact that they have so far been unsuccessful in improving their housing
conditions, there is a constant flow of cash in the household. While it was acknowledged that they had to buy all
consumer goods including rice, they had the advantage of being able to get bargains at the end of the market
day. Hence, they spend less than if they had to buy from retail shops. Crucially, they have the knowledge, assets,
business networks, and the motivation to be sustained by a livelihood strategy that is not dependent on rainwater
nor threatened by elephants. Karunawathie considers the ability to move away from nonviable farming and to
have the motivation to take advantage of opportunities available as a critical. “I tell the women, let’s do it. I’ll find
you buyers. If enough of us make sweets, buyers will come to the village as they do in Aluthgama. But they don’t.
Even my sister who is a widow will not do it. We have to help her financially.”
Appendix 8 113
Poverty—Negros, Philippines
Mr. Ricardo Valenzuela and his wife Mariethel have one child. They have been living in their current house, which
is made of grass thatch walls, packed clay floor, and galvanized iron sheets on the roof, about 50 meters from the
project road, for the last 4 years. Prior to their transfer, they lived in Valderrama, Antique, his home place after
their marriage in 1995. He was then helping his parents in their farm and doing carpentry works in his hometown.
His mother-in-law encouraged him to move his family to Barangay Magallon Cadre in 1998, as at that time, the
local government gave free access to land for those with a family who wanted to relocate in the village. The
family of his wife is from Magallon Cadre.
As they have no land to farm in Magallon Cadre, he works as a carpenter, accepting wages of P100 per day
provided the food is free. He gives priority to carpentry work than wage laboring in a hacienda as the wages for
the former are much better than the latter. If carpentry work is not available, he does not mind laboring in a
hacienda. As a farm laborer, he only earns P70–P100 per day without food, depending on the nature of the job.
Recently, he has established a network with local contractors engaged in building and repairing houses not only
in the village but also in other villages of Moises Padilla. This expands substantially his earning capacity. He is
known to be a reliable carpenter in the village.
Though he was not in the village before the road rehabilitation, he observed that transport services along the
project road have been expanding every year. As cement, lumber, and roofing materials are sourced as far away
as Bacolod City, he knows that a good road network has made a substantial contribution to decreasing the cost
of materials and making delivery faster; this also facilitates his repair or construction activities. Availability of
motorized transport in the village has reduced his travel time, leaving more time for carpentry. In the last few
months, he traveled mainly locally on foot as most of his job orders were within the village. He and his wife travel
weekly to the market center of Moises Padilla mainly to buy foodstuffs such as rice and fish. The family has to
buy most of their foodstuffs, as they have no land to grow crops or vegetables of their own.
Due to lack of capital, Mariethel is sometimes engaged in peddling various foodstuffs such as salted or boiled
peanuts and salad jackfruit since their arrival in the village, which she herself prepared or processed. All she
needs is a capital of P1,000 to start a peanut business. She usually engages in this business from October to
December in the town proper and surrounding areas, coinciding with peak supply of peanuts. During these
months, she can easily earn a net income of P300–P400 in a day from a capital of P1,000. When there is no
cheaper supply of peanuts, she shifts to making and selling jackfruit salad. She can also earn as much as 40%
on her capital, but she can only do this in a much reduced scale. Her transport costs from the village to the town
proper and vice versa of P10 is affordable as long as she has a business. For the whole village, there are only
two of them doing this kind of business. The availability of a tricycle has enabled her to move to other barangays
of Moises Padilla. She feels the business of selling peanuts has more potential than jackfruit salad as she can
easily sell all the peanuts within the village and surrounding villages. This gives her the reason to travel outside
the village; otherwise, she only goes to the town proper with her husband to buy essential household items.
Although still poor, they are slowly finding ways to increase their household income. They feel they would have
enough resources for their 5-year old daughter to obtain a college degree if they can expand their business
connections in the village. Mariethel is now pregnant so they are thinking more about increasing their household
income. For them, good roads and better transport services contribute to the overall development of their
community that ultimately expands their livelihood capacities.
Poverty—Yogyakarta, Indonesia
Mrs. Yuliana Sukarta lives in Candirejo Village with her husband and three sons of 17, 15, and 5 years old. She
has been a teacher at the elementary school of Kakung Village, 14 kilometers from her residence, since 1989.
She has used the study road before and after it was rehabilitated.
Situation Before Road Rehabilitation. From 1989 to 1995 before the road was rehabilitated, she used two
methods of public transportation to go to school everyday. She left for school at 6 am using the open pick-up to
Semanu and arrive there at 6:45 am. At Semanu, she needed to wait for about 20–35 minutes for other route to
Kakung by the same kind of transportation. From Semanu to Kakung, she took about 30 minutes which means
she needed to travel about 95 to 110 minutes everyday.
She often arrived at school late or could not go to school due to difficulty with transportation. If the first open pick-
up at 6 am was fully loaded or she came to the street late, she had to wait for the second one at 8 am, and would
Appendix 8 114
Box 9—Continued
arrive at school at 9:50 am; making her 2 hours and 20 minutes late. The third and last pick up would come at
10 am. The open pick-ups from Semanu came back at 11 am, 1 pm, and 3 pm.
The school ends at 1 pm so she had to use the last pick up from Semanu to her home, and she would arrive at
home at 3:45 pm, therefore, she spent 9 hours and 45 minutes per day for her job as a teacher. She spent a total
of Rp7,000 per day for her fare.
Situation After Road Rehabilitation. Since the road was rehabilitated in 1996, it has been very easy for her.
There are so many public vehicles everyday. She leaves for school at 7 am and arrives at 7:30 am. After school,
she arrives home at 1:30 pm. She spends Rp2,000 per day as fare. She teaches 24 days per month on average.
Comparing before and after road rehabilitation since 1996, she saves Rp5,000 per day, and 2 hours and 15
minutes. Since 1997, she has done makeup for bridal couples as her second job in the afternoon. She does the
new job inside and outside the village, and she has no transportation constraints. She gets orders for about 5 to
11 days per month with an average net income of Rp2,200,000 per month. The other additional job is doing
farming on the sharecropping system. She works with farmers and gets an average yearly net income of
Rp1,500,000.
Conclusion. Mrs. Sukarta uses the road very well. She has received many benefits from the road and
transportation improvements, specifically in encouraging her in her new job and improving her income.
C. Cooperative Society as a Safety Net for Poverty
Box 10: A Cooperative Society Helping Poor Households to Cope with Crisis—
Kurunegala, Sri Lanka
Weeragama is the neighboring village to Nugannoruwa,
(the project site), and has a milk-collecting center for the
milk cooperative society. Villagers from Nugannoruwa
belong to the cooperative, as well as people from other
surrounding villages. The cooperative is state
sponsored, and started here in the village on 1 March
2001. It grew out of a project called “Dairy Milk” which
operated in the area for 5 years previously. Once that
project finished, villagers decided to start a cooperative
themselves. It is part of a regional network, with a sector
officer based in the nearby town of Nikaweratiya. The
cooperative is open to all villagers. They simply need to
sign up and there is no joining fee. For farmers, a key
purpose in starting the cooperative was to collect and
market milk for themselves, cutting out the
intermediaries. Whatever the cooperative gets for the milk is passed on to the farmers. The milk-collecting center
is in the compound of Mr. Subasingha (pictured), who is the local secretary of the coop. His house/shop is beside
the road, and people bring their milk to him in containers on foot or by bicycle. He receives a commission of
25 cents per liter collected. Every morning, a milk truck hired by the central cooperative visits the collection point
and collects the milk. Weeragama is the last stop of the collection truck, which continues to Kuliyapitiya where the
milk is sold to a factory that processes flavored milk, yogurt, etc., and also supplies Nestle.
The peak season for the cooperative is January–April, the wet season, when cows produce 50% more milk.
Members are paid Rp25 per liter for buffalo milk and Rp13 a liter for cow’s milk depending upon the fat content
determined by Mr. Subasingha using a simple centrifuge. The other main outlets for milk in the neighborhood
villages are the motorscooter milk traders who ply their trade along the road. They will pay Rp22–Rp23 per liter
for buffalo milk and Rp10–Rp12 per liter for cow’s milk. There is an economic incentive, then, for farmers to use
the cooperative but the coop will also take all the milk that they produce so they have a guaranteed sale.
Motorcycle intermediaries do not guarantee that they will purchase everyday nor pay on time. The cooperative
currently produces approximately 300 liters per day.
The improved road has resulted in an increase in the milk trade, with more motorscooter collectors, and more
people joining the cooperative every year. The improved road has meant that the milk truck can come to collect
the milk every day of the year. Previously, the road was treacherous during the rainy season, sometimes
inaccessible, and the surface was in a poor condition, resulting in slower transport times and damage to the
vehicle. Given the necessity of transporting fresh milk quickly, the road is important in making the whole
Appendix 8 115
Box 10—Continued
enterprise viable. Better road has also given farmers the confidence to invest time and resources in developing
cattle holdings because the truck will come whenever they have milk to sell. During the current long dry spell too,
the development of the milk cooperative and the guarantee of good access for the truck means that villagers
have an important alternative source of income while their lands remain dry and unproductive.
D. Households Successfully Moving Out of Poverty
4. The following case studies show households that have moved out of poverty using
their length of residence and asset base.
Box 11: A Household Successfully Moving Out of Poverty—Negros, Philippines

The family of Mr. Rolando Labarga is the most enterprising in the whole village in terms of finding ways to
maximize the opportunities that improved roads can bring. They are the only family that has continually met the
fish requirement for the village for more than 25 years. Through the joint efforts of Rolando and his wife Sarianita,
they were able to raise and provide education to all their children. They have 11 children with ages ranging from
5 to 28. The eldest is married with two children, with his house just behind Rolando’s. As the eldest, he actively
helped in their business to raise enough money for the education of his younger brothers and sisters.
Consequently, he only completed first year of college education (11 years). Three children completed college
degrees; two are employed in Manila, and one has just completed a degree in education. Out of the 6 children of
school age, only 5 are currently studying: 2 are in high school and 3 in elementary. The one not attending school
completed his high school last year but had to postpone studying as the family decided to first let his sister finish
her degree. As his sister completed her degree in March 2002, he will start attending college this coming school
year. Their youngest son is not yet of school age.
Their current house is situated along the barangay road 75 meters from the project road and is made of
cemented floor, wooden walls, and galvanized iron roof, which was built some time in 1990 in one of the
hacienda’s vacant lots. This is their second house on the same piece of land. Their first house was built in 1974
after their marriage. As it was just a small hut and after saving much money, Rolando felt the need to have a
bigger house as his family grew bigger. They are one of the few families in the village with no member of their
household working in haciendas.
Fish trading was his wife’s idea as she came from a fishing village in Himamaylan. Sarianita started peddling fish
in a basket around the village right after their marriage in 1974, with Rolando accompanying her in buying and
transporting fish from Barangay Sarait, Himamaylan. As fish demand in the village increased, so did Sarianita’s
volume of sales. Then she had difficulty physically peddling the fish so she started scouting for a location for a
fish stand. In 1986, Sarianita and her husband constructed a makeshift fish stand with just a table and temporary
roofing along the project road in the village. Together they left the village for barangay Sarait, Himamaylan at 9
am everyday by renting a tricycle at P120 per return trip and back to the village at about 1 pm to 2 pm, depending
on fish availability. As they have established credit lines, they get the fish on a consignment basis. They usually
have 2–3 boxes of fish daily of 40–50 kilograms (kg) per box. They start setting up the fish at their stand at 3 pm
and close it at about 6 pm. In days that they have a fish surplus, Sarianita peddles the fish in the village or simply
dries the fish for sale over the following days. The price of fish that they sell ranges from P50 to P120 per kg.
They usually have 20–30% marketing margin for every kg of fish. At the end of the day, they have earned at least
P350. Their fish stand is becoming the center of village information as people usually exchange information while
buying fish.
With over 25 years of using the road, Rolando and his wife have seen how the road contributes to the
development of their village and is the only family that fully utilizes the road to the extent of graduating into a
nonpoor household in the village. They observe that more people in the village are investing in transport facilities
as their source of livelihood.
Before project road rehabilitation, there were only two to three tricycle owners in their village; now there are about
25. They can now leave home at mid-morning instead of early morning to be at the fish-landing center. About
30% of their customers are project road users. People traveling to barangay Cabacungan, La Castellana, and
Canla-on City now and then stop at their fish stand to buy fish.
Appendix 8 116
Box 12: A Household Successfully Moving Out of Poverty—Sorsogon, Philippines

Mr. Jesus Gripon is a tenant farmer of coconut in the
Patag area of Barangay Palale, with 6 hectares (ha) of
land. His house and land are situated at the end of a
track running up from the project road. It takes
approximately 20–30 minutes to walk one way, and up
to 1 hour when transporting goods. Patag is the
agricultural hinterland of the barangay, accounting for
70% of the agricultural output of the area. Jesus lives
with his wife and five children, and has a strong family
network in the area. His father and brother are also
tenant farmers. His father has been farming the same
12 ha for 22 years and Jesus for 6 years.
Jesus is required to sell his coconut to his landlord in

Palale and has other lands and tenants throughout the area. The landlord buys the product and sets the price.
He receives P10 per sack. The price in Bulan is P10.60. However, he is only required to transport the produce
down the track to Palale. The owner transports it on for further processing in San Francisco. He feels that it is a
fair arrangement as there are no transportation costs for him. The standard division of the crop is a 50/50 split
between the landlord and tenant, though in Jesus’ case, a 20% reduction is made by the landlord for “moisture
content,” so the split is 60/40 in the landlord’s favor. Other tenants are free to sell their copra to other buyers, but
a
landlords provide good credit to tenants before harvesting against the next crop.
From the money he has made from copra, Jesus invested in a chainsaw in 1998. He paid P30,000 and now
operates a business in cutting down old coconut tress for timber for house construction. He makes timber planks
and sells them for P100 per slab. He also started a business 8 years ago, trapping and selling birds from the
forest. He does this every December and sells to a shop in Bulan from where the birds are sold as pets. His
connection with the shop is now so good that he catches birds to order only though during December. He spent
P200 on traps initially and now sells the birds for P100 each. Jesus is in no doubt that the better road between
Bulan and Palale has made the development of these small enterprises easier, as transportation is quicker and
more readily available. He travels more regularly, and people come to the barangay more regularly too so
communication has improved in both directions.
Patag does not have a road link to the main barangay center and the project road. However, there was a road
constructed over 2 kilometers in October–November 1986. The road was constructed for better access of a
government official to his land in the area and a jeep operated along the route for a few months in 1987. This
service soon stopped, however, as the road deteriorated. Local people had undertaken maintenance for some
time, but landslides affected the road, fewer people were able to contribute labor, and they lacked basic tools and
machinery to undertake the work. Jesus feels that a road simply was not a priority for people. If they spent time
maintaining the road everyday, livelihood activities would suffer and the benefits from the road were not large
enough to make this worthwhile.
Jesus has been offered 3 ha more elsewhere to farm on a tenant basis, but is unsure of whether he can legally
farm more land. He has no doubt, however, that he could manage the additional land as his brother and family
would help. Family networks and connections, both in times of difficulty and in providing agricultural labor, are
clearly critical here in helping the poor overcome their poverty and in offering a social safety net in times of
hardship.
_____________________________
a
Coconut is harvested approximately every 2 months, so six crops in a year is possible.
Box 13: A Household Successfully Moving Out of Poverty—Matara, Sri Lanka

Mr. Nimal’s family has been living in Heegoda for generations. He does not have any land himself, but his family
has 2.5 acres of tea land. In 1993, he had just left school after completing his advanced level examination when
he was chosen to be trained as an Assistant Tea Inspector by the Tea Smallholder Development Authority
(TSDA). The program was one of the many inaugurated by the State in the aftermath of the political unrest of
1989. Along with 18 others from the Pasgoda Divisional Secretary’s Division, Nimal was trained for 9 months in
every aspect of the tea industry. As part of the practical training, the trainees were expected to set up small tea
plant nurseries in their home plots. They were provided with watering cans, plastic for making baskets for the
plants, and 4 kilograms of fertilizer.
Once the training was completed, Nimal continued with the tea plant nursery and developed it with the aid of the
officials he had met during the training. A partnership was formed with two of his friends based on labor sharing.
Appendix 8 117
Box 13—Continued
However, the location of the nursery in Nimal’s home garden on the slopes of Thannehena was not conducive for
expansion. Customers were limited to those who had tea holdings in the area and did not mind headloading the
saplings. They decided to relocate the nursery
closer to the road where better access would
enable customers to load saplings on to trucks
and other vehicles. They rented 30 perches of
state land just by the road for the very nominal
cost of Rs300 per month. “Sales have gone up
dramatically and we now have over 115,000
plants. We don’t even need the help of TSDA
to find buyers,” he says. In addition to the new
location by the roadside, the constant
encouragement and transfer of information by
the extension officers of TSDA has been the
main factor in the success of Nimal’s
enterprise. He frequently visits the TSDA
office in Matara to keep in touch with the
officers as well as be updated on new
techniques, assistance schemes for tea sector
enterprises, etc. “I got to know these officers
during the training. I have since then kept in close contact with them. The extension officer visits the nursery at
least once a month. He will sometimes stop if he is passing on his bike and sees us in the nursery. I go to the
office to meet the other ladies and gentlemen. If they feel you are working hard and you listen to them, they are
very good to you and love to help you.”
In addition to being a good source of income, the tea plant nursery needs extensive labor only during certain
periods of the month. This leaves Nimal with time to engage in other income-generating activities. Transporting
locally produced treacle to town is one such activity. During the period when treacle is very cheap, he buys about
70 kilograms which is the maximum he can transport by bus and sells it to shops in Ambalangoda town where the
price remains high throughout the year. When the price is good, he makes as much as 75% profit on the trade.
E. Vehicle Operators
5. The following case studies indicate that nonpoor households have invested in
vehicles to take advantage of the road.
Box 14: A Vehicle Operator—Bengkulu, Indonesia

Mr. Supardi has been living in Talang Kabu Village for a long time. He started working as a driver in 1993. He is a
head of household with a wife and two sons. He looks like a successful driver with a good house. During 1993–
2000, he drove a closed pick-up with a capacity of 11 passengers. Since the end of 2000, he has been driving an
open pick-up for freight transportation. The frequency of trips did not change before and after asphalting. He did
one round trip a day from Penago to Bengkulu. The differences before and after asphalting are given below.
Situation Before Asphalting. He explained that before the asphalt road he did not have competition. The pick-
ups were very limited, and it was easy for him to get passengers. His car was fully loaded everyday with more
than 17 passengers. For one round trip per day, he gets around Rp51,000 (2 x 17 x @ Rp1,500) and from the
passengers’ freight around Rp25,000, totaling Rp76,000 (with $1 equivalent to Rp2,300). The average driver’s
income is Rp15,000 per day. Everyday he needed 14 hours for driving (7:00 am–9:00 pm). He needed to repair
his vehicle every week and spent 1–3 days on each. Consequently, he drove for around 12–18 days a month and
the other days were used for vehicle repair and rest for himself.
Situation After Asphalting. After the road was asphalted in mid-1997, he has had a lot of competition with many
other pick-ups, trucks, and other kinds of vehicles for passenger and freight transportation. He gets about
13 passengers a day, and they do not want to pay for freight under 30 kg per person. For one round trip, he gets
around Rp260,000 (2 x 13 x @ Rp10.000) and from passengers’ freight he gets around Rp5,000, totaling
Rp265,000 (with $1 equivalent to Rp10,000). The average driver’s income is Rp50,000 per day. Everyday, he
needs 8.5 hours for working (7:45 am–4:15 pm). The condition of the road is advantageous for the vehicle. He
needs to repair his vehicle only twice a month and spends 1 day on each. Consequently, he drives for around
24–28 days a month. The other days are used for vehicle repair and taking rest.
Appendix 8 118
Box 14—Continued
Conclusion. After asphalting, he can save 4 hours per day for taking rest or other activities. The other benefits
are that the vehicle is more durable, he does not need to do repairs every week so he can work for about 28 days
a month, and the expenses for vehicle repairs have decreased by around 200%. Income has increased by
around 300% for the driver.
Box 15: A Vehicle Operator—Matara, Sri Lanka

Gnanatilake operated the only form of public transport on the Katuwana Road since it was first broadened and
tarred in 1994 until the road was rehabilitated in 2001. This was a 5,000-kilogram Elf truck which had been
converted to an 18-seater dual-purpose vehicle by adding parallel seating, steps, a metal-framed roof, etc. This
a
bus ran on a permit and timetable issued by the provincial road authority. As the only form of public passenger
and cargo transport available, the operation made a good profit until 1998. From 1998, profits started falling as
the cost of maintenance of the truck increased substantially due to the condition of the road aggravating the
repair cost that increased as the truck aged.
The takings of the truck also dropped as the passengers on the main roads preferred to use the regular buses—
which over a period of time had increased on the main roads—than climb on to the truck. Hence, the core of
passengers came from the villages along the Katuwana Road. When the condition of the road deteriorated to the
extent that he had trouble running his truck, he took the step of personally spending over Rs40,000 to surface the
impassable sections of the road. He carried out routine maintenance such as shifting rocks on the road.
When in 2001 the road was improved to the extent that it made possible the running of a bus, Gnanatilake sold
the modified truck and bought a single-door minibus to ply the same route, using the same permit and timetable.
His fleet now includes the truck he has hired out to the Kiruwanaganga estate to carry laborers from the villages
along the Katuwana Road to the estate. This truck is modified to carry passengers and is insured at a rate of
Rs50,000 per passenger for 24 passengers. This cost, as well as all running and maintenance costs, have to be
borne by Gnanatilake. He earns a rate of Rs1,450 per day from the tea factory. While the improvement of the
road has opened up opportunities for him to expand his vehicle-based business, he admits that the competition in
terms of more public and private buses operating on the road will be a challenge that needs to be faced. As long
as the road was in a deplorable condition, no other operators threatened his monopoly. Now, however, both the
private and state sectors have moved in within 4 months of the road improvements, and two more buses and
about 10 more vans/trucks that service the tea wage laborers have started operation. Gnanatilake foresees at
least one or two more state and/or private buses plying the route. He is currently thinking of ways to meet this
challenge.
By 2000, Gnanatilake was running into serious problems with regard to the profitability of the operation due to the
excessive costs of vehicle and road maintenance borne by him. It was during this time that he heard that a road
improvement project was being launched in the Southern Province and the selection of roads was in progress.
Gnanatilake immediately mobilized all resources at his command to ensure the Katuwana Road was selected for
the project. He approached the village monk to help in drafting the petition and accessing the relevant authorities.
Mobilizing support from the villagers did not pose any problems: “All the houses near the road, as well as on the
slopes, were very willing to participate in the signature campaign. They all felt the need for the road.” The road
was selected for the project and work was completed in 2001. While Gnanatilake is very pleased with the
dramatic improvement in the condition of the road, he still remains vigilant about its maintenance.
_____________________________
a
Carrying passengers in modified trucks is possible only under permits for special travel requirements such as
pilgrimages. These permits are issued for a specified period of time. The Southern Provincial Council has
made an exception and issues permits for modified vehicles to carry passengers.
Box 16: A Vehicle Operator—Kurunegala, Sri Lanka

Lalith owns the only vehicles in Nugannoruwa. He has a truck, which he uses for collecting cow dung, and a
minivan, which he bought after selling his tractor 2 years ago. Lalith’s is one of the better-off families in the
village. He has a land of his own and also married into another landowning family. He has about 15 acres of
paddy land, but like others in the village, has not farmed for a couple of years because of the severe water
shortage. He currently only farms his chena (dry) lands for household food. In the meantime, he has developed
both of his transportation businesses to compensate for the lack of farming, and his income now is at least equal,
if not better, than previously.
Appendix 8 119
Box 16—Continued
Lalith bought the second hand truck from his savings from farming and now uses it to collect dry cow dung from
Nugannoruwa and other villages along the road. This activity has
been taking place in the village for at least 10 years, but Lalith
started 2 years ago. Villagers were well used to selling the dung by
this stage, and he has no problem with supply. He employs three
people in this business to drive the truck and collect the cow dung,
and he collects a truckload from along the road everyday. Once
collected, the dung is sold to farmers along the roads in Kalpitiya, a
vegetable-growing region where the demand for manure is high.
Lalith says that he can easily sell the manure, but that he sometimes
has to wait for payment. He plans to expand the business to bring
back vegetables from Kalpitiya for sale in the vegetable wholesaling
area of Dambulla.
He pays about Rp500 for one truckload of dung, and makes about
Rp1,000 after deducting his costs. For villagers who have about
25 cows, it may take up to 2 weeks to collect half a truckload. For
families with only a few cows, they get together with others to fill a
truck. He started the business once the road had already been
rehabilitated, so he says he cannot comment on the road improving
his business, but feels that the completion of the bridge further along
the road (undertaken by the project) has had an enormous benefit in improving access along the road. He faces
no difficulties in driving along the road any time of the year, and feels the surface is now much better than before.
Lalith bought his van about 1 year ago. He does not use it for business daily, but hires it out fairly frequently to
villagers. The most popular trip is to Anuradapura for pilgrimage for which he charges Rp2,000 for hire. He
makes this trip about once a month. He also takes villagers to the south and to the coast for which he charges
Rp5,000. He made this trip three times last year. He has a driver to whom he pays 20% of the revenue. The van
also plays an important social function as the informal “village ambulance.” Villagers know that they can rely upon
Lalith to take them to the hospital in an emergency, day or night. This happens two or three times a month on
average, and although he does not ask directly for payment, there is an accepted charge of Rp250 to the
Kotawehera Hospital, and Rp400 to Nikaweratiya. Villagers always pay as promptly as they can.
F. Case Studies of Traders Using the Road Successfully
Box 17: A Successful Trader—Yogyakarta, Indonesia

Mrs. Peti, 40 years old, has been living in the village since 1990 with her husband. She has a 10-year-old
daughter and a 4-year-old son. Her mother is 65 years old. She has been successful with her kiosk and truck.
Situation Before Road Rehabilitation. This family was a poor family. She started to build her family business
with her husband in 1990 when the road was still bad. She did not work everyday and just sold meals once a
week at a weekly market inside the village and was unemployed 6 days a week. Her husband works as a
motorcycle (ojek) operator at the route of Candirejo to Semanu (7 kilometers). The motorcycle was the main
family income earner. The family income ranged from Rp25,000 to Rp100,000 a month.
Situation After Road Rehabilitation. The family started a business in 1997 selling sugar, coffee, rice, and other
villagers’ daily needs. A good road and a lot of public transportation has been an advantage for her to go to
Semanu (capital of the subdistrict) and Wonosari (capital of the district) to buy things those needed by villagers.
The capital to start the business was limited, and came only from her husband’s earnings as a motorcycle driver.
Nevertheless, the business grew very well and quickly, and in 1999 she started to multiply her business supplying
agricultural input for the farmers inside the village. She provided fertilizer and pesticide for the farmers. Since
2000, she has been supplying the three villages of Candirejo, Giri Panggung, and Balong. To give better services
to customers, she decided to buy a truck in 2001 for transportation to remote areas with big orders. She is
supplying many farmers’ groups in those three villages. Another business of the family is buying crops from the
farmers in the three villages, for instance rice, groundnut, and soybean. She sells the crops to Yogyakarta and/or
Semarang.
At present, the family income is estimated at around Rp2,500,000 per month, a large difference from before
1996. She explained that it was very difficult to develop business activity before 1996, because of transportation
Appendix 8 120
Box 17—Continued
constraints. It was difficult to go to Semanu and Wonosari to buy and difficult to give a good service to a village
that has a transportation problem. The family is still building the house and extending the store. The family is very
successful in using the road and taking advantage of better transportation. She understands that the road has
played an important part in the rapid development of her business. She hopes that the community can also take
advantage of the good condition of the road and public transportation.
Box 18: A Small Trader—Bengkulu, Indonesia

Mr. Anwar is an intermediary who collects paddy, coffee, and rubber from farmers. He is a head of family with five
members. His business started 15 years ago. He does not have capital for his business so he is very dependent
on traders (moneylenders) from outside the village who lend the money to him. His bargaining position is very
weak in negotiating the crop prices because he has to sell the crops to the moneylender. The family seems not to
be very successful in the business. It has a simple house, but not as bad as the poorest families.
Situation with a Bad Road. He said that both the main road and the farm road (Tree Crop Smallholder Sector
Project [TCSSP] Road) give benefits to him in managing his business. Before the TCSSP road was built, they
had to carry the crops on their shoulders; this was expensive and very slow. To collect 400 kilograms (kg) of
crops, he needed 5 days. At harvest time, it was very difficult to find two or three workers because most of them
were more interested in doing harvest activity than carrying crops. The main road before 1996 was not asphalted.
There were very few pick-ups that came to the village and trucks were very rare. He sent the crops for selling to
Bengkulu generally once a month in the dry season when the truck came and did not do it for about 5 months in
the rainy season. He temporarily stopped to work as a trader in the rainy season when he worked as a farmer.
His monthly income during the dry season averaged Rp150,000 and in the rainy season Rp90,000.
Situation with a Good Road. Since the farm road was built by TCSSP, he has been able to bring crops by pull
cart or bicycle. He can collect 500 kg of crops in 2 days. Real benefits for him are (i) the cost is 25% lower, and
(ii) he can save time because the work is quicker. Since the main road was asphalted at the end of 1996, he has
felt that there are no constraints on transportation. He can send the crops to Bengkulu everyday if the crop is
available. Normally, at the moment, he sends the crops for selling twice a week during the year. He gets a
monthly income of Rp650,000. It is difficult for him to send crops more than eight times a month because he is in
competition with another trader from outside the village and some new traders from the village who have started
in the same business.
The problem for him is lack of capital in extending his business. He got the capital from a moneylender at
25% interest per month, and he has to sell the crops to the moneylender from a very weak bargaining position.
He tried to borrow money from a formal bank, but was unsuccessful. The bank needs collateral and has difficult
administration requirements. He buys paddy at Rp950 per kg from the farmers and sells it at Rp1,100 per kg to
the moneylender. If the farmers sell to the market, the price is Rp1,000 per kg. The rubber he buys from the
farmers at Rp1,750 per kg and sells it at Rp1,950 per kg to the moneylender. If the farmers sell to the KUD
(Village Unit Cooperation), the price is Rp1,850 per kg.
Mr. Anwar said that the present condition of both the main road and the farm road is good enough for his
business. Both roads help him carry out his business. Although he cannot improve his business to the level of the
bigger trader, he thinks his business and family income have been better since both roads were improved. His
main constraint is lack of capital.
Box 19: A Successful Trader—Bengkulu, Indonesia

This is a rich family that has been successful in business; it started in 1997 after the main road to the village was
asphalted. The family’s businesses are crop trading and rice milling. The other sources of family income are rice
field, coffee plantation, and rubber plantation. In supporting the business, the family also provides the pick-up.
Situation Before Asphalting. Before he started his business at this village, the family left the village for 4 years
in 1993–1996 to do business outside the village. Mr. Niran said that Talang Kabu Village was bad for business
before 1996. A lot of agricultural production could not be sold outside the village. They had to transport the crops
on their shoulders or by pull carts. Because of difficulties of transportation, he lacks information, communication,
and business opportunities. That time, the family stayed at Benuang Galing Village, North Bengkulu, about
150 kilometers from Talang Kabu Village. The family worked as small traders who collected and sold many kinds
of crops. The business was not growing well because of competition with many other traders at North Bengkulu.
Appendix 8 121
Box 19—Continued
Situation After Asphalting. At the end of 1996 when the Talang Kabu Road had already been asphalted, the
family came back to Talang Kabu to live and started a business in the village. The business started with a small
capital of Rp1,500,000 (around $562 at the rate of Rp2,300 per $1 in 1996). The family bought coffee, rubber,
paddy, corn, and other crops from farmers in the village. He took advantage of the better transportation
conditions. He rented the pick-up or a truck for crop transportation from the village to Bengkulu once a week at
the end of 1996 with the family making a net income of Rp500,000 per month. In 1997, the volume of his trading
increased as did the frequency of crop transportation to an average of twice a week, and he made a net income
of around Rp1,250,000 per month.
Despite the monetary crisis of 1997, production of agriculture increased. He took the opportunity to multiply his
trading volume so that from 1998 to 2002 the frequency of his trading transportation increased to three or four
times a week; he had an average income of Rp3,000,000 per month. In 2000, the family bought two pick-ups and
one rice milling machine for extending the business. He said that he is not in competition with the other trader
inside the village because he is a bigger trader and he helps the other traders with capital to expand their
business. He buys paddy at Rp950 per kg from the farmers and sells it at Rp1,200 per kg to the main trader at
Bengkulu. If the farmers sell to the market, the price is Rp1,000 per kg. The rubber he buys from the farmers at
Rp1,750 per kg and sells it at Rp2,100 per kg to the main trader at Bengkulu. If the farmers sell to the KUD
(Village Unit Cooperation), the price is Rp1,850/kg.
The family uses the road facilities very well, both the main road built by the Public Work Services and the farm
road built by TCSSP. Mr. Niran and his wife said that their pick-up brought paddy from the rice field to the rice
miller, and coffee from the plantation to their house. It was impossible to do this before the farm road was built.
They said that they and other traders would benefit more if the Government improved the farm road and main
road; they are ready to contribute to it and to persuade the other traders to do likewise.
G. Case Studies of Professionals Using the Road Effectively
Box 20: A Teacher—Yogyakarta, Indonesia
Mr. Sidik is a teacher who has lived in the village with his family for about 15 years. He also farms for a
secondary income.
Situation with the Bad Road. He said that the farm road has no impact on his job as a teacher as he does not
use it to go to school. But this road does have an impact on his secondary job as a farmer. Before the farm road
was built, he was too lazy to go to his plantation in the afternoon after school. Normally, he went to the farm once
a week on Sunday. Farming was not done well, crop production was very low, and he got just a small additional
income from this. His job as a teacher is helped by the main road, which crosses the village. Before the road was
asphalted, it was time consuming to go to and from school. He had to walk about 5 kilometers from his residence
to school along the very bad road because there was no public transport. A few pick-ups and trucks that used the
road did not serve passengers for short trips. They only served passengers who went to and from Bengkulu.
It took him 90 minutes going to school and 90 minutes coming back everyday. School begins at 7:30 am and
closes at 1 pm so that he left his home at 6 am and come back home at 2:30 pm. He struggled to go to school for
6 days a week in the dry season. But in the rainy season, he often did not go to school 3 days a week, and
sometimes because of heavy rains he would not go to school for a week. The head of school and other teachers
knew of the conditions in the area and understood his constraints.
Situation with the Good Road. Since the farm road was built in 1994, he has gone to his farming everyday after
school. His wife and his oldest son go to the farm in the morning. They planted coffee trees and dryland paddy
between two lines of coffee trees. His income as a farmer is higher than his income as a teacher. As a teacher,
he has gained a significant benefit from the main road. Since the road was asphalted, he has had no problems in
going to school. He does not have to walk to and from school; he used public transportation everyday. Many pick-
ups from the village go to and from Bengkulu. It takes only 15 minutes each to go to and from school. Everyday,
he leaves his house at 7:15 am and after school he arrives home at 1:15 pm. He goes to school 6 days a week at
the dry season and rainy season. He pays just Rp2,000 per day for transportation, but he has more time for his
farming activities with his family. He thinks he teaches better than before the road was asphalted.
The farm road has no impact on his main job as a teacher, but it helps him improve his economic condition
through farming. The main road gives him a significant benefit, specifically for his main job as a teacher.
Appendix 8 122
Box 21: A Nurse—Yogyakarta, Indonesia

Mrs. Suryatini is 40 years old, living in the village with her husband and two daughters, 12 and 7 years old. She is
a nurse in Candirejo Village consisting of 20 subvillages. She has been working in this job for about 14 years.
Situation with the Bad Road. For her job, she has to meet people in 20 subvillages and she has to attend a
weekly meeting at the subdistrict center. From 1988 to 1995 before the road was rehabilitated, she had difficulties
in going to Pacarejo to attend the weekly meeting with the doctor and other health center officials. She needed
2 hours for a return trip as she had to walk. In servicing the 20 subvillages, she spent more than 3 hours going
back and forth from each remote subvillage as there too she had to walk. There are 20 community group health
services, one for each subvillage. Because of the bad road, the villagers also found it difficult to meet the nurse at
the village health center to get extension services. She had a schedule of visiting the 20 subvillages 20 days a
month. In the dry season, it works fine for about 7 months a year, but in the wet season it does not work well at
all, as the road becomes very bad then. Often, a housewife would have to give birth without a nurse because the
nurse could not get there in time. She did not regularly attend the meetings at the district health center because
of the road and transportation problems, especially in the rainy season.
Situation with the Good Road. Since 1996 when the road was rehabilitated, it had become easier for the nurse
to visit all the subvillages. She can use a motorcycle and spend only 15–30 minutes on the trip. Each day she
can visit three subvillages for health extension services, and preventive care of children and pregnant women.
The program can be managed well in both the dry and rainy seasons. People can also come to the village center
to meet the nurse, with the result that she can give a better service with access to better tools and facilities. At
present, all pregnant women give birth at the village health center. Normally, the pregnant woman comes 1 day
before the last day of pregnancy. The nurse can always attend her weekly meeting at the district health center.
The road condition has helped the nurse do her job well, and there are better health facilities available for the
villagers.
Box 22: An Extension Service Officer—Yogyakarta, Indonesia
Mr. Sumianto is a Javanese who lives in Penago-2 village with his family. He is an agricultural extension services
worker who works in four villages at Talo subdistrict, including Talang Kabu Village. His work area is very
extensive because most of the families in the villages are farmers.
Situation with the Bad Road. Before the farm road was built, it was very difficult for him to visit farmers in the
rice field or plantation area, and he had to walk far from his home to the field. The farm areas of the four villages
are divided into 20 areas, which are dispersed at the various locations. Generally, he could visit each area once
every 2 months in the dry season. In the rainy season, he did not visit any farm area for about 4 months; instead
he met them at their settlements at some subvillages, which were more easily accessible. Others he could not
visit at all. He was able to achieve only 40% of his annual target. The farmers did not get adequate information on
agriculture technology. When the farm road was built but the asphalt road was not, the extension services were
done through their group at each subvillage, but the problems of the farmers were not adequately solved. Hence,
agricultural production remained low because the farmers did not apply external agricultural inputs, e.g., fertilizer
or pesticide. The villages had a very limited supply of these inputs and the prices of them were very high (150%
of the regular price) because of transportation costs. The extension service officer often did not attend important
meetings at subdistrict and district levels because of the transportation problem.
Situation with the Better Road. After farm roads were built by the Tree Crop Smallholder Sector Project at
Talang Kabu and Penago villages, it helped him do his job. It is very easy for him to visit farmers in the rice field,
plantation, and/or their settlement. He goes to one or two locations per day by motorcycle, so he visits each of
the 20 areas twice a month in the dry and rainy seasons. In terms of his annual program, he met more than 90%
of his target, so that the farmers get information they need about agriculture technology, networking, and
marketing. Since 1997 when the main road was built, the farmers have been able to access agricultural input
supplies from Bengkulu. They implement agriculture technology based on the technical guidance from extension
services, and apply fertilizer, pesticide, etc. Production has increased and they sell the crops at a better price at
Bengkulu and/or other markets outside the village.
The farm road and main road have given him benefits. Both roads help him carry out his annual program. He can
serve the farmers easier in the field and/or at their settlements, and he can manage his time for attending
important meeting at subdistrict and district levels.

Rural Roads PDF

Uploaded by

Copyright:

Available Formats

Rural Roads PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Rural Roads PDF

Uploaded by

Copyright:

Available Formats

ASIAN DEVELOPMENT BANK IES: REG 2002-15

IMPACT OF RURAL ROADS

A CASE STUDY-BASED ANALYSIS

ADB − Asian Development Bank

In this report, “$” refers to US dollars.

Operations Evaluation Department, IE-68

II. CASE STUDY PROJECTS AND SITES 4

III. UNDERSTANDING POVERTY 8

A. Who Are the Poor? 8

IV. CONSIDERATIONS FOR PROJECT DESIGN AND IMPLEMENTATION 11

A. Considerations for Project Design 11

V. KEY IMPACTS AND RELATED ISSUES 16

A. Changes in Transportation Services 16

VI. CONCLUSIONS AND RECOMMENDATIONS 33

Participatory poverty assessments have long identified remoteness and isolation as

Competition is determined by potential demand for travel, users’ proximity to a

Problems of maintenance are compounded when it is unclear where the institutional

In addition to the recommendations above, the study provides a range of other

B. Study Objectives and Scope

II. CASE STUDY PROJECTS AND SITES

Table 1: Project Objective and Components

Source: Asian Development Bank.

Table 2: Description of Case Study Roads and Their Impacts

Integrated Projects Transport Sector Projects

Sorsogon/SIADP (PHI) Barangay Palale Barangay Bical

Negros/FRIP (PHI) Barangay Magallon Cadre Barangay Macagahay

Kurunegala NWP-WRDP (SRI) Nugannoruwa Walahinikalla and Wembuwa

Matara/SPRIP (SRI) Heegoda Makiliyathanne

Bengkulu/TCSSP (INO) Talang Kabu-Talaatan Talang Padang

Study Area/Project Project Site Control Site

Yogyakarta/TLRP (INO) Candirejo Village Sumber Wungu-Kelayu

TLRP = Third Local Roads Project.

III. UNDERSTANDING POVERTY

A. Who Are the Poor? Figure 1: Relationship of Livelihood Assets to Structural

life, and a susceptibility to periodic physical and CAPITAL ASSETS

economic shocks, and seasonal crises. It also Natural

describes a state of voicelessness and an inability to Social

influence the structures, institutions, and processes

and the relative strengths and weaknesses of these

shocks, and ability to participate in wider institutions

B. Key Poverty Themes Common to All Study Locations

3. Family and Community

IV. CONSIDERATIONS FOR PROJECT DESIGN AND IMPLEMENTATION

A. Considerations for Project Design

1. External Conditions that Cannot be Affected by Road Development

26. Remoteness and Access to Marketing Networks. Locational factors have an

2. Structural Conditions that Cannot be Affected by Road Development

3. Conditions that Can be Affected by Road Development

(i) changes in transportation services: transport modes, volume, cost, and

B. Considerations for Project Implementation

1. Labor-Based Techniques and Poverty Reduction Option

2. Governance Issues Surrounding Rural Roads

39. Neglect of Maintenance. Regular maintenance of rural roads is a critical precondition

3. Community Participation as a Key to Long-Term Sustainability

V. KEY IMPACTS AND RELATED ISSUES

A. Changes in Transportation Services

1. Transport Modes, Volume, Cost, and Competition33

2. Preconditions Necessary for the Emergence of Competition

3. Who Provides Transport Services?