36006 THE WORLD BANK GROUP WASHINGTON, D.C. TRANSPORT PAPERS TP-10 MARCH 2006 Rural Access Index: A Key Development Indicator Peter Roberts Shyam KC Cordula Rastogi TRANSPORT SECTOR BOARD RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR Peter Roberts Shyam KC Cordula Rastogi THE WORLD BANK TRANSPORT SECTOR Washington, D.C. BOARD © 2006 The International Bank for Reconstruction and Development/The World Bank 1818 H Street NW Washington, DC 20433 Telephone 202-473-1000 Internet www/worldbank.org Published March 2006 The findings, interpretations, and conclusions expressed here are those of the authors and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. This paper has been produced in part with the financial assistance of a grant from TRISP, a partnership between the UK Department for International Development and the World Bank, for learning and sharing of knowledge in the fields of transport and rural infrastructure services. To order additional copies of this publication, please send an e-mail to the Transport Help Desk transport@worldbank.org Transport publications are available on-line at http://www.worldbank.org/transport/ RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR iii TABLE OF CONTENTS Acknowledgments...................................................................................................................v Executive Summary ..............................................................................................................vii 1 Introduction ................................................................................................................... 1 1.1 Purpose of the Rural Access Index ........................................................................... 1 1.2 Poverty and Isolation ............................................................................................. 1 1.3 Rural Access Index and MDG................................................................................... 1 1.4 Equity and Development......................................................................................... 2 1.5 Impacts of Rural Access ......................................................................................... 2 1.6 Conceptual Shifts .................................................................................................. 2 1.7 Definition of RAI.................................................................................................... 2 2 Rural Access Index.......................................................................................................... 3 2.1 Values and Aggregates........................................................................................... 3 2.2 Sensitivity to Change ............................................................................................. 3 2.3 Measurement and Reporting ................................................................................... 4 2.4 Updating .............................................................................................................. 4 2.5 Costs to Establish and Maintain the RAI......................................................................... 5 2.6 Alternate Procedures to Estimate RAI............................................................................ 5 2.7 Estimating Models ................................................................................................. 5 2.8 Network Models..................................................................................................... 5 2.9 Quick Accessibility Mapping..................................................................................... 6 2.10 Regional Variations in Rural Access .......................................................................... 6 3 Household Survey of Access to Transport ........................................................................... 6 3.1 Proposed Transport Survey Module .......................................................................... 6 3.2 Essential Question ................................................................................................. 7 3.3 Measuring Time and Distance.................................................................................. 7 4 Applications of the Index.................................................................................................. 8 4.1 Regional Rural Access Targets ................................................................................. 8 4.2 Sub National Analysis............................................................................................. 8 4.3 Equity in Change of Access ..................................................................................... 8 5 Priorities for Action.......................................................................................................... 9 5.1 Ensure that the RAI Is Monitored ............................................................................10 5.2 Extended Ownership of RAI....................................................................................10 5.3 Implement Applications and Studies for RAI .............................................................10 References...........................................................................................................................11 RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR iv LIST OF ANNEXES List of Annexes··················································································································· A-1 Annex I: Rural Access Index: Summary Sheet ········································································ A-3 Annex II: Rural Access Index by Region: IDA Countries···························································· A-5 Rural Access Index by Region: IBRD Countries·························································· A-7 Annex III: Rural Access Index by Region: IDA and Non-IDA Countries········································ A-9 Annex IV: Household Level Transport and Access Module·························································A-13 Annex V: Time-Distance Survey: Sample Documents ······························································A-15 Annex VI: Africa: Transport Targets and Indicators Related to the MDGs····································A-25 Annex VII: Vietnam ­ Provincial Allocation to Improve Rural Access ··········································A-27 LIST OF FIGURES Figure 1. Rural Access Index: Aggregate Average for IDA Coountries............................................. 3 Figure 2. Rural Access: Regional Estimates................................................................................ 6 Figure 3. Households' Level of Access to Nearest Basic Services (0-30 minutes) by Consumption Quintile Nepal (1995/1996 and 2003/2004)........................................................ 9 Figure 4. Rural-Urban Variation in Level of Access to Basic Services (Nepal 2003/2004). .................. 9 LIST OF TABLES Rural Access Index by Region: IDA Countries·········································································· A-5 Rural Access Index by Region: IBRD Countries········································································ A-7 Rural Access Index by Region: IDA and Non-IDA Countries······················································· A-9 Household Level Transport and Access Module: Roads Component (Draft)··································A-13 Africa: Transport Targets and Indicators Related to the Millennium Development Goals (MDGs)·····A-25 RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR v ACKNOWLEDGMENTS This Technical Paper has been prepared as part of the World Bank's Transport Results Initiative. This is led by the central Transport Unit (TUDTR) and is supported through the trust fund provided by the United Kingdom's Department for International Development under the Transport and Rural Infrastructure Partnership (TRISP) with the Bank. Access to the results of the many household surveys which were examined from IDA and non-IDA countries was made possible through contact in the respective country teams. We are grateful to them all, too many to be individually mentioned here. However, special note must be given to the extensive assistance given by Pascal Heus and his team working in the Africa Household Survey Databank: Christophe Rockmore, Zena Angesom and Rose Mungai who undertook the major part of the analysis for the countries in Sub-Saharan Africa. Much of the analysis for other countries was carried out by Anuja Saukar. The modeling to establish estimates of the RAI on the basis of data on rural population, available land area and road network length was carried out by Shyam KC, one of the authors. This work benefited from advice given by John Hine. Peter Roberts is the Lead Specialist for Infrastructure Services in the central Transport Unit of the World Bank. He heads the Transport Results Initiative and the thematic group for Transport and Social Responsibility. Shyam KC is a doctoral candidate in the Social Science Faculty of the University of Maryland. He contributed to work on Rural Access under the Transport Results Initiative. Cordula Rastogi was a consultant for the Transport Results Initiative during FY2004. She is currently a Transport Economist in the South Asia Region transport sector team. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR vii EXECUTIVE SUMMARY This paper describes the Rural Access Index (RAI), a headline transport indicator which highlights the critical role of access and mobility in reducing poverty in poor countries. The Index is part of the Results Measurement System for IDA 14. It is defined together with the official method of measurement which is on the basis of locally representative household surveys. Current estimates of the Index indicate that some 900 million rural dwellers world wide do not have adequate access to the formal transport system. Initial values for more than 30 IDA countries show the overall level of access for their rural population to be 57 percent. Within the IDA group access is worst for the Sub-Saharan Africa countries for which the average RAI is 30 percent. For a selection of non-IDA countries the average value of access is nearly 90 percent. The preferred measurement of RAI is from household survey results. A draft transport questionnaire module is proposed for new household surveys with estimates of the relatively modest resources required to establish and update the measurement. Alternative methods of measurement and estimating techniques are described in case there is no ready prospect of undertaking a suitable household survey. The challenges for extending and updating the Index are described, together with the resources which have been developed to tackle these. Links are provided to those resources. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 1 INTRODUCTION those with reliable access. Problems of access are particularly severe in those rural areas 1.1 Purpose of the Rural Access Index which are distant from roads that carry motorized transport services on a regular The World Bank Infrastructure Action Plan basis. Roughly half of the world population, (World Bank July 2003) emphasises the and 70 percent in the IDA countries, are rural importance of helping borrowing countries to dwellers (World Bank 2005). On the basis of strengthen the collection and use of data in current values of RAI it is estimated that about their infrastructure sectors. The World Bank 900 million rural dwellers in developing Transport Sector has responded through the countries, the great majority of them living in Transport Results Initiative . This includes 1 IDA countries, are without reliable transport identifying a small number of key diagnostic access. The large majority of these people are measures which have particular significance for poor as defined for the MDGs. The lack of the sector in contributing to the wider reliable transport systems in IDA countries, development process. These are considered to mainly in rural areas, has compelled household be headline transport indicators. members to spend significant time traveling in order to meet basic needs. Studies in two This paper describes the establishment of the areas of Tanzania, for example, revealed that a Rural Access Index, one of several Transport major portion (on average 40 to 50 hours) of Headline Indicators endorsed by the World the total weekly time available to each Bank Transport Sector Board in 2003. The household was spent on transport (World Bank index has been adopted for the Results 2003). Usually this time burden is distributed Measurement System (RMS)2 of the 14th round inequitably within the household. In this case, of the International Development Association women were particularly overburdened, taking (IDA-14) which was launched in July, 2005. responsibility for about 80 percent of the time The Index was developed in response to the cost (World Bank 2003) consensus led by borrowers that it identifies an important priority for poverty reduction 1.3 Rural Access Index and MDG strategies in view of the established links between physical isolation and poverty. This Transport is not specifically identified in provides stronger linkage to the Millennium Millennium Development Goal targets, but it Development Goals (MDG) and better assesses makes key contributions to achieving many of the contribution of IDA assistance to the the Goals--underpinning pro-poor growth and sustainable development of the beneficiary improving social inclusion. Surveys show that countries. poor people recognize isolation as a major contributor to their poverty and The Rural Access Index provides a consistent marginalization. In practice, improving access basis for estimating the proportion of the rural to transport for rural men and women in low population which has adequate access to the income countries is considered essential to transport system. It can help to inform policies promote rural development, to increase uptake and strategies which ensure that the rewards of human development services (educational of development are distributed more equitably and health), to facilitate inclusion of different to the rural population ethnic and other groups, to improve employment opportunities, and to stimulate 1.2 Poverty and Isolation growth for poverty reduction. Fan and Chan- Kang (2005) examined the factors which Physical isolation is a strong contributor to contributed to the exceptional growth and to poverty. Populations without reliable access to the reduction of poverty in China during the social and economic services are poorer than past thirty years. They concluded that, whilst the rapid introduction of the expressway network did play a part, the much shorter 1 www.worldbank.org > topics in development > lengths of low standard feeder roads made an transport > transport results measurement even more important contribution to growth 2 The RMS involves measuring a set of indicators and poverty reduction, achieving about four which capture key aspects of the development of 81 times greater benefit/cost ratio than did the countries which receive IDA concessional assistance (IDA, 2004). expressways. Similarly Gannon and Liu RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 2 (1997:11) contend that improved transport transport sector have been based on contributes not only through enabling better characteristics of the road network such as the access to services and opportunities, but also length or density of different categories of by lowering the transport costs incurred in roads (paved and gravel roads, urban, feeder delivering and/or accessing such services and roads and national highways, etc). Such opportunities. measures do not give a clear picture of the transport access level available to the rural 1.4 Equity and Development population since they do not relate the provision of transport facilities to the location The Rural Access Index also helps to address of the target population. Thus the first the issue of equity which is an increasing focus important conceptual shift has been to of the international development community. measure the accessibility of the target The World Development Report (World Bank population to the road network, rather than 2005) explores the relationship between equity simply some aspect of network size. and development strategy and makes the issue of equity central to "poverty-reducing" Corresponding to this has been the second development. A previous World Development conceptual shift, which is in the measurement Report points out that the affordable access to units used to study the impacts of changes in services is especially low in many of the transport infrastructure. Thus, rather than poorest countries, with poor people needing to focusing on an administrative unit and travel much further to reach basic services assuming that all households within the unit such as health and education, than richer have the same level of access to the transport people in the same country (World Bank 2003) network, the definition of RAI enables more As an example, the Report notes that in rural detailed measurement - that is, at the Nigeria, children from the poorest quintile of household level. This approach can capture the population need to travel more than five how access to transport services relates to times farther than the children in the richest household characteristics. quintile to reach the nearest primary school. The third shift is that the definition of the index 1.5 Impacts of Rural Access provides a common international basis for understanding rural access as it relates to Change in rural access has differing impacts on transport. However, detailed interpretation of various sections of the population. In the RAI for any country must be set in the particular, the socioeconomic impact of local context. increased access differs by gender, by age groups, by different caste/ethnic groups, and Further, by providing information on rural by income. For example, improved access residents' differential access to the all-season potentially increases men's migration, and may road network, the RAI helps planners to devise result to increased workload on women in the policies and programmes to meet specific rural farm and household. Similarly improved motor access objectives. RAI thus provides an access brings consumption goods nearer to objective basis for governments to set households, but affects rural artisans and transport sector goals, and to establish those residents whose livelihoods depend upon investment priorities for improving rural portering. For policy makers, this underlines access. that transport interventions often need to be coupled with complementary policies if all 1.7 Definition of RAI socially and economically disadvantaged households are also to reap the benefits of In practice the RAI measures the number of improved access. rural people who live within two kilometers (typically equivalent to a walk of 20-25 1.6 Conceptual Shifts minutes) of an all-season road as a proportion of the total rural population. An "all-season The Rural Access Index helps to address the road" is a road that is motorable all year round issues mentioned above by changing the way by the prevailing means of rural transport the outputs of investment in the rural transport (typically a pick-up or a truck which does not sector are measured. The shift in the form of have four-wheel-drive). Occasional measurement is threefold--what to measure, interruptions of short duration during where to measure and how to define what is inclement weather (e.g. heavy rainfall) are being measured. The indicators which have accepted, particularly on lightly trafficked been generally used to report progress in the roads. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 3 Some sections of the population (the elderly, Amongst the IDA countries there is also a the disabled or those carrying heavy burdens) significant difference between those which may find that even distances of less than one receive blended assistance (aggregate average kilometer present a significant barrier to of 65 percent) and those which do not access, particularly under extreme conditions (aggregate average of 30 percent). Thus, of terrain or climate. On the other hand, in increasing the number of rural people with many remote situations (such as the hills of sustainable access to rural transport is Nepal or remote areas of rural Africa) people expected to be very significant for the may be accustomed to walking many development of lower income countries. kilometers in order to reach formal transport services. Alternatively they may manage The Rural Access Index is designed to respond without using such transport at all. The choice to borrowers' priorities and to be measurable of two kilometers as the defining distance for for a `critical mass' of IDA countries. In `adequate access' is a compromise between establishing this indicator, therefore, priority these extremes. This establishes a consistent has been given to the countries with the definition of the Index, enabling comparison highest populations so that the 32 countries for and aggregation (on the basis of population which RAI has been established represent 88 weight) of the values for target populations. percent of the total rural population of IDA countries. Priority has also been given to those The specific emphasis given to roads in the countries with the larger land areas for which definition of the index reflects the importance rural access is likely to be particularly relevant. of road transport for improving rural access for A two page summary of the RAI is presented in the great majority of rural people in most low Annex I, and the current, endorsed values of income countries. In those situations where Rural Access Index by country, showing the another mode, such as water transport is source and year of survey with the current dominant the definition can be modified to rural population are given in Annex II. reflect that. If this is done, then the fact should be explicitly recorded against the resulting RAI value. Figure 1. Rural Access Index: Aggregate Average for IDA Countries. 2 RURAL ACCESS INDEX 100% 93% 90% 2.1 Values and Aggregates 80% 65% Results for 32 IDA countries,3 representing 88 70% percent of the total rural population in all IDA 57% countries, show that on average 57 percent of 60% rural dwellers has access to the transport 50% network. The graph also shows a significant 30% difference between `IDA' and `IBRD' countries. 40% For ten non-IDA countries (for which suitable 30% household survey results are available) the RAI value is much higher at 93 percent (Figure 1). 20% 10% For sixteen IDA countries in the African region, representing 58 percent of the total rural IDA-eligible countries (32 countries) IDA-only countries (25 countries) 0% 1 population of the region, the aggregate RAI is Blend countries (7 countries) Non-IDA countries (10 countries)* 30 percent. For four IDA countries in the South *Including estimate of 97 percent for RAI from Asia region, representing 96 percent of the Government of People's Republic of China. total rural population of the region, access is 65 percent. 2.2 Sensitivity to Change The RAI is designed to capture changes in rural 3 Albania, Azerbaijan, Bangladesh, Benin, Burkina access as a result of development in the Faso, Burundi, Cambodia, Cameroon, Chad, Ethiopia, transport sector. Relatively small Gambia, Ghana, India, Indonesia, Kenya, Lao PDR, improvements in access to an all-season road Madagascar, Malawi, Mali, Mongolia, Nepal, have marked impacts in terms of improving Nicaragua, Niger, Nigeria, Pakistan, Papua New the overall indicator results. However, to date Guinea, Tajikistan, Tanzania, Uzbekistan, Vietnam, there are only a few countries for which we Yemen, Zambia. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 4 have for more than one set of household data Income/Expenditure Household Survey (IES), results to determine in detail how the RAI Poverty Survey (PS) and Core Welfare changes over time. Indicators Questionnaires (CWIQ). The surveys are designed to produce high-quality data on a An analysis of preliminary measures for variety of key topics and be representative for Vietnam with time-series data (during the the main segments of population. Thus, the period of 1997-2002) indicates the sensitivity main subgroup `rural population' is adequately of the indicator to change over time. In covered. Vietnam, RAI has increased from 73 percent in 1997 to 76 percent in 2002. Both figures are With the establishment of the Rural Access based on data from the Vietnam Living Index, there are several key challenges for Standards Survey (VLSS) which is modeled on extending and maintaining measurement of the LSMS. They indicate a clear increase in this headline indicator over the coming two to rural access equivalent to an average rate of three years. improvement of about half a percent per year. · Updating the value of the index at This reflects the development activity which intervals of not more than three years responded to the government's priority. or so in each country for which the RAI Limited time series data from two other has been established; countries shows slower rates of change. · Establishing a current value for those However, relatively small improvements in countries for which the index has not access to an all-season road have marked yet been derived; impacts in terms of improving the overall · Complementing the index with related indicator results. information which enhances under- standing of its significance. 2.3 Measurement and Reporting 2.4 Updating The preferred approach to measuring this indicator is by analysis of household surveys Updates of the indicator will largely depend on that include appropriate questions about the frequency of household surveys. However, access to transport. The aim is to integrate this only a few of the surveys are carried out on a with the measurement of household regular basis. Of the 30 countries where LSMS characteristics such as income and access to surveys had been carried out by 2003, for services such as education, health and clean example, 19 countries had only one survey water supply. All of these surveys completed completed until then (World Bank 2003). Even to date which include a suitable question on with the emphasis which the IDA-14 RMS and access to rural transport have been analyzed other initiatives place on strengthening data for the current values of RAI. The design and collection in low income countries it will be a conduct of such surveys is costly and time slow process initially to increase the number of consuming. Thus, although this is actively countries for which the RAI can be based on encouraged as an important requirement for all household surveys. borrowing countries to monitor their progress in poverty reduction, it will take at least two or For this to be done on a sustainable basis, three years before the results of any new national institutions must be involved and survey will be available. However, there are a encouraged to take ownership of the processes number of surveys already established in which to measure RAI. In this regard, the Bank is a question or short module on access to raising awareness in beneficiary countries of transport could be incorporated at a modest the significance of the index for capturing marginal cost. differences in performance between countries. The central Transport Unit is contributing Where suitable surveys of households and guidance on relevant rural access questions for individuals exist, these are the most cost- household surveys to strengthen the effective way of obtaining information on rural effectiveness of the responses. Three access (IDA 2004). The majority of the 32 consultation workshops on Transport Results available IDA country indicators have been held in Nairobi (2004), Washington DC (2004) established using this approach. Information and Colombo (2005) were able to draw from Living Standard Measurement Surveys valuable participation from a wide range of (LSMS) and similar household surveys carried governmental, civil and bilateral organizations. out between 1994 to 2003 has been used for These workshops reviewed current experience the calculation. Specifically these surveys are in collecting and applying core measures and Living Standard Measurement Survey (LSMS), exchanged interim findings in collecting data RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 5 and formulating indicators. The overall supplemented by estimates for some other approach was endorsed in these workshops, countries which have been derived by different including the value of headline transport methods. These essentially involve working indicators in general and of the RAI in with representative sample area data with particular. layers of the road network map and the population map to determine how many people Along with involving stakeholders, there is the live within the specified catchment of the road need to "intensify support within Country network. Sometimes the data is available Assistance Strategies (CASs) and IDA projects within a GIS and may be processed for improving the statistical capacity of electronically. In other cases all or some of the member countries and to work in partnership data (usually the location of the roads) may with other agencies to strengthen the only be available from graphic maps or aerial international statistics system" (IDA 2004). photographs. This includes helping beneficiary countries to develop the capacity to conduct, analyze and 2.7 Estimating Models report such surveys on a regular basis. It is planned that the Rural Access Index will be In the absence of such data a `best estimate' of reported in the 2006 edition of World the Index can be made in consultation with Development Indicators. specialists having the necessary country experience. In this regard, two methods, 2.5 Costs to Establish and Maintain the RAI namely `Network Models' and `Quick Accessibility Mapping', are proposed to arrive For countries with national household surveys at some broad estimates of the Rural Access which include questions that permit RAI to be Index for a given rural population, land area calculated, the marginal cost of producing the and length of road network. These estimates indicator is generally up to one day of are provisional and must be replaced once experienced statistical input. Where there is a measures are calculated based on household suitable survey which does not include a survey results (or GIS data). relevant question, there will be a one-off cost to negotiate, design, test and incorporate such 2.8 Network Models a question. This cost is estimated to be up to a week of technical input (which will probably be Two network models--namely the Random spread over a much longer period of elapsed Road Network Model (Hine 1984) and the time) in addition to the cost of analysis. Where Square-Grid Model--are proposed to arrive at there is not yet a suitable household survey it broad estimates for Rural Access Index. For will be necessary to estimate the RAI in the those countries where we do not know first instance by a sample mapping technique precisely how roads are located in relation to as described above. The input for this is the people, we need to make assumptions estimated to be about one month of analysis about the road network and population for each estimate of the indicator, provided distribution patterns. Both models calculate that the necessary data on the location of access as a function of the length of road roads and of the rural population is available. network, habitable land area, distribution The latter costs may be significantly reduced if pattern of roads and the distribution of suitable data is available in a Geographic population as follows: Information System. 2.6 Alternate Procedures to Estimate RAI Access= f [(length of road network) * (habitable land area) * (distribution of As the Rural Access Index is incorporated in the IDA 14 Results Measurement System it is roads) * (distribution of population)] important to establish initial values of the RAI for all IDA countries for which no relevant The models utilize two of the variables which household survey based data are yet available. are most readily available in time series for Where the necessary geographic information many countries--the road length and the system (GIS) or other map data is available arable land area. Some assumptions are made this can provide a good estimate of the RAI. about the habitable areas of a country and distribution of roads within the habitable and The values of RAI estimated from national non-habitable areas. The results of calculations household survey results have been based on these models are given in Annex III. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 6 2.9 Quick Accessibility Mapping on less than a quarter of the region's rural population. South Asia (at 58 percent) and This is a simplified version of a full GIS-based LAC (at 54 percent) have nearly twice the level measure. The "quick map" is prepared by of rural access as does AFR. Rural access is overlaying the road map on the topographical even better for Europe and Central Asia (75 country map. The RAI is determined on the percent) and for East Asia and the Pacific (94 basis of an informed approximation for the percent) ­ the latter is reduced to 86 percent if population distribution of the country in the figure of 97 percent provided by the question. One example of such an exercise was Peoples Republic of China is omitted. carried out in Nepal by overlaying, the all- weather Road Map and Village Development Figure 2. Rural Access: Regional Estimates Committee4 (VDC) map on the Topographical Map of Nepal (Shah, 2004). The population of Rural Access: Regional Estimates (no. of countries included) a whole VDC was assumed to be concentrated at its center on the basis that more than 50 Sub-Saharan 30 percent of the population is concentrated at Africa (24) the VDC center. Obstacles in the form of hills Middle East & North Africa (5) 34 and rivers, necessary relief for contour and Latin America & detour distance for river crossings were taken 54 Caribbean (7) into consideration. South Asia (4) 58 Thus was calculated the percentage of Europe & Central Asia (16) 75 population having accessibility to nearest all- weather road in terms of walking time. Based East Asia (8) 94 on this procedure, it was estimated for 2004 0 20 40 60 80 100 that on average 41 percent of the overall (%) population had access to the nearest all-season road within 30 minutes. This compares reasonably with LSMS (2003/2004) estimates of 37 percent of the overall population having 3 HOUSEHOLD SURVEY OF ACCESS TO access within 30 minutes of paved roads. The TRANSPORT quick mapping procedure also showed variation between the access level in the 3.1 Proposed Transport Survey Module Mountains/Hilly and Terai (flat) regions where this ratio was found to be 28 percent and 56 While some information on household percent respectively. accessibility is available, review of existing multi-topic household surveys reveals the lack 2.10 Regional Variations in Rural Access of a consistent transportation questionnaire module in these surveys. Often any questions The RAI values calculated for 32 IDA countries related to rural access are included under the (tabulated in Annex II) together with values community module. Such inquiries treat the calculated or estimated for a further 32 community being surveyed as a single unit, countries by the methods described in this which does not allow measurements of chapter give a good basis for assessing the variations among households. This generally variations between regions (tabulated in Annex results in an overestimation of access. There is III). Overall these countries account for some also a lack of uniformity in rural access 82% of the total rural population (2003 data) questions across various surveys. of the six regions. The regions of Middle East and North Africa (MNA) and Latin America and A Transport and Access Module has been the Caribbean (LAC) are rather weakly prepared for field testing to be included in represented but values for the other four multi-topic household surveys of the LSMS regions cover more than two-thirds of the type. This module builds on substantial regional population in each case. background work by Baker and Denning The regions are compared in Figure 2. Sub- (2005), as well as on the community transport Saharan Africa (AFR) has the poorest level of module developed by Frankenberg (2000), and rural access at only 30 percent. MNA is only on some specific transport modules such as slightly better at 34 percent but this is based those included in the Guatemala Living Standard Survey (2002). The proposed module 4The VDC is the lowest political unit in Nepal with a is based on an accessibility framework adopted population numbering about 4000. by ILO (ILO/HMGN 2003). It will bring RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 7 transport and access related questions into a and distance and the relationship between single module which is currently planned to these are likely to be influenced by individual have three sections: `Road infrastructure', and cultural factors. This may be particularly `Transport services' and `Other'. The current significant in those areas of developing draft of the `Road infrastructure' section is countries where time measurement devices presented in Annex IV. and measures of distance are not commonly used or comprehended by rural dwellers. 3.2 Essential Question A field survey exercise has been designed to There is often strong resistance to improve the understanding of people's incorporating an additional module or several responses to questions about the time or questions on account of the practical survey distance of their travel. This will contribute to constraints and resource costs. At the very practical guidance for the designers of least all IDA countries should be willing to household and similar surveys who wish to include a single question about the distance or include questions on individuals' travel time time to walk to an all-season road. This will and/or distance. enable them to meet their responsibility to the IDA-14 RMS. A field survey compares subjective reports by individuals of their travel time and distance If at all possible, a second question should be with the corresponding objective measures. included to ask about access to a reliable, all- Initial surveys for this exercise have been season `transport service'. In the medium-term carried out in rural Albania (in cooperation with it is likely that the focus will shift to assessing the government survey department INSTAT) the level of access to transport services rather and Tanzania. than access to the road network, since not all sections of the network necessarily carry The initial field work verified the survey satisfactory services. procedure. It showed that, with practical calibration and supervision processes, Field studies described in this paper indicate pedometers could be used to measure the that questions about access to a road or other actual distance and time between the facility are probably best based on the `time household and the destination, and to gain taken' ­ unless there are strong reasons locally some information on the quality of data to base them on `distance'. Where possible collected through household surveys as it respondents' replies should be checked against relates to time and distance traveled. objective measures. The increasing user- friendliness and reduced cost of global The results from Albania showed quite close positioning system equipment can offer correlation between the reports of time and practical ways of doing this. distance as well as between the reports and objective measurements of both. The expected 3.3 Measuring Time and Distance variations in relation to age and terrain were confirmed to a limited extent. In household surveys, data on travel time and distance is usually collected through face-to- A second time and distance survey was face interviews of randomly selected completed in Bukoba, Tanzania in July 2005. households (or individual members of The results of this survey indicate a weak households). This data is influenced by a correlation between estimated time and variety of factors, such as travel purpose, measured time, but a very weak correlation in transport mode used, trip route, characteristics case of distance. In addition, a high degree of of terrain, etc. Moreover, the data collected is over-estimating was observed for both time susceptible to subjective variation, as it is and distance. This over-estimating was based on reports of perceptions rather than on particularly pronounced for distance, with over observations of events. 60 percent of the estimates being more than twice the actual distance. Furthermore the accuracy of these surveys depends on the memory of the respondents. In both Albania and Tanzania, the measured There are various considerations such as time and measured distance correlated deterioration of recall over time, the tendency strongly, indicating the reliability of the to `round up' responses and variation in recall pedometers used to measure the time and with personal characteristics of respondents distance in these surveys. Similar surveys are (such as age). Additionally, estimates of time planned for other countries (including in RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 8 Vietnam, Laos and Cambodia with the financial halving poverty and hunger, increasing access support of DFID). The sample terms of to education (especially for girls), reducing reference for the survey, together with maternal mortality and improving child health. examples of the procedures and survey forms are presented in Annex V. 4.2 Sub National Analysis Initial indications from these surveys are that The Vietnam Living Standards Survey (VLSS) questions about `time' rather than `distance' was carried out in 1998 and repeated in 2002. may generally be more reliable and informative The General Office of Statistics is now required in respect of physical access. However, in each to repeat the VLSS every two years to monitor case this should be verified in relation to local progress in implementing the national experience. Preference should be given to Comprehensive Poverty Reduction and Growth established local practice where this is proven Strategy (CPRGS). The VLSS was conducted to be working well. again in 2004 and the results have recently become available. The survey is designed to be Similar surveys to compare time and distance representative at the Provincial level ­ there perceptions are planned for other countries are 67 provinces in Vietnam. On the basis of (including Vietnam, Laos and Cambodia with the 2002 survey the Rural Access Index has the financial support of DFID). When designing been estimated for each Province. These a new survey or if improvement may be estimates show a close correlation between the required, reference should be made to the lack of rural access and the level of poverty in format of the sample questions for the each province. Tanzania study, in Annex V. The Government is working with the World Bank and DFID to establish a protocol for 4 APPLICATIONS OF THE INDEX allocating national resources amongst the provinces with the poorest basic access in Some countries and regions have started to support of the national CPRGS. The algorithm extend use of the RAI beyond its function as a for this protocol is shown below. headline indicator which focuses on a key development issue and being incorporated in allocation = base + f(poverty rate)* the IDA-14 RMS. Three cases are described below. (rural population)*(lack of access) 4.1 Regional Rural Access Targets In Annex VII there is an illustration of the protocol being applied to the 33 provinces in The Rural Access Index has been used, Vietnam having poor levels of basic access. together with other headline transport indicators to strengthen the relationship 4.3 Equity in Change of Access between the Millennium Development Goals and the broad response of the Transport An example from Nepal demonstrates how Sector in Africa. This is set out in a report improvement in access to basic facilities over prepared for a meeting of the Africa Transport the past decade has favored the non-poor. Ministers (Africa Union April 2005) and is Although households' overall access to basic summarized in the Table which is reproduced services in terms of time required to reach the in Annex VI. The headline indicators constitute service-points has improved for all services high level targets for the transport sector between 1996 and 2003, the proportion of which link the direct interventions of the sector households in the richest quintile with access to their significant impacts on the MDGs. to key facilities within 30 minutes exceeded those in the poorest quintile by from 11 The Table shows that, for Africa, improvement percent to 28 percent in 1996, and from 10 in rural access is a key requirement for percent to 47 percent in 2003 (Figure 3, Nepal achieving a number of the Goalsin particular: CBS 2005). RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 9 Figure 3. Households' Level of Access to Nearest Basic Services (0-30 minutes*) by Consumption Quintile Nepal (1995/1996 and 2003/2004). 120.0 100.0 80.0 elitniuQfo Poorest Quintile 60.0 Richest Quintile % 40.0 20.0 0.0 1995/96 2003/04 1995/96 2003/04 1995/96 2003/04 1995/96 2003/04 Primary School Health Post Bus Stop Paved Road *The threshold of 30 minutes walking time is longer than that currently indicated in the definition of the RAI but is appropriate for the difficult access conditions in Nepal. Source. Nepal Census Bureau. Similarly, urban dwellers have far better It is evident from these observations that there access to basic services than do rural is considerable inequality in the level of access households, with the difference ranging from to the transport system amongst different 10 percent to 60 percent (Figure 4). groups of people, and that a large portion of the world's population is still excluded from Figure 4. Rural-Urban Variation in Level of such services and opportunities. In this Access to Basic Services (Nepal context, the RAI shows the very low levels of 2003/2004). effective access to the transport system for rural residents in the poorest countries and regions. By doing so, it points to the need to 120 ensure that the rewards of development are Urban Rural more equitably available to the rural 100 population. )leveLseccA( 80 5 PRIORITIES FOR ACTION 60 Now the primary challenge is for countries to % mainstream and sustain the Rural Access Index in routine monitoring at the national and 40 sub-national levels. Some priorities are summarized below. TUDTR is ready to assist 20 such activity through the Transport Results Initiative. Specialists supporting country programmes or their partners may make 0 contact through the appropriate World Bank Primary School Health Post Bus Stop Paved Road regional Transport Sector Manager or directly with the Transport Results Initiative Source. Nepal Census Bureau. (proberts@worldbank.org). RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 10 5.1 Ensure that the RAI Is Monitored 5.2 Extended Ownership of RAI Countries which receive IDA assistance will be Those non-IDA countries which recognise that required to report changes in the value of the improving physical access and mobility is an Rural Access Index as part of the IDA-14 Re- important priority for the rural population can sults Measurement System. The responsibility also use the RAI as a tool for informing policy for this in each country will fall to the National and guiding resource allocation. This should be Statistics Office which has to ensure that there encouraged. The Index offers the advantages is an up to date value of the Index by including of having an internationally established the appropriate question in a national LSMS- definition, which is supported by a standard type survey. Where it is impractical for this to measurement process with related protocols be done before the end of 2006, a `current and benchmark values. value' should be estimated by one of the methods described in this Paper on the basis of 5.3 Implement Applications and Studies available data for the all-season road network for RAI and for the location of the rural population. International reporting of the Index needs to The RAI value or estimate should be be complemented locally by established determined in consultation with the relevant applications for analysis with related measures departments of the Ministries of Transport and to set it in the context of prevailing national of Agriculture and/or Rural Development to and sub-national priorities and constraints. ensure that it is integrated with the plans of those ministries. The aim should be to establish a data framework which is sufficiently Further studies of individual perceptions of detailed to determine the Index for different walking time and distance will enable improved income categories of the population and for design of all household survey modules which sub-national planning units of the country. are concerned with access (eg. to health or education services). This will help to improve Measurement of the RAI value should be the interpretation of the data and will updated at least every three years, together contribute to better understanding of local with those other indices which are based on priorities for household travel and mobility. household survey data. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR 11 REFERENCES Africa Union, UNECA, AfDB, WB etc. 2005. Nepal Census Bureau of Statistics (CBS). 2005. "Africa Transport and the Millennium Nepal Living Standard Survey 2003/2004 Development Goals" Report. Accessed on August 22, 2005 at Baker, Judy L and William Denning. 2005. "Development of a Transport Module for Multi- Nepal Census Bureau of Statistics (CBS). 2005. topic Surveys." TP-5 Transport Papers. World Nepal Living Standard Survey 1995/1996 Bank, Washington, DC. Report. Accessed on August 22, 2005 at Binswanger, Hans P, Shahidur R. Khandker and Mark R. Rosenzweig. 1993. How Infrastructure Shah, Neeraj. 2005. "Transport Sector and Financial Institutions affect Agricultural Performance and Impact Indicators- A Output and Investment in India." Journal of Nepalese Case Study." A Paper presented in Development Economics 41: 337-366. the Regional Consultation Workshop in Colombo, Sri Lanka. Frankenberg, Elizabeth. 2000. "Community and Price Data." Pp. 315-338 in Grosh, United Nations 2000. "Millennium Declaration." Margaret and Paul Glewwe (eds.) Designing New York: United Nations. Accessed August Household Survey Questionnaires for 22, 2005. Developing Countries, Volume I and "Community Data." Pp. 253-255 in ___ Volume III. World Bank, Washington, DC. World Bank. 2003. "Implementation of LSMS by Country, Data Information and Gannon, Colin and Zhi Liu. 1997. Poverty and Questionnaire Used" Accessed on August 22, Transport, a World Bank Discussion Paper. 2005. World Bank, Washington, DC. Hine, John L. 1984. "Some Limitations to the Opportunities for Road Investment to Promote World Bank. 2003a. Technical Note on Rural Rural Development." International Conference Transport in Multi-sectoral and Community on Roads and Development, Paris 22-25 May, Driven Projects. World Bank, Washington, DC. 1985. World Bank. 2003b. World Development Report International Development Association (IDA), 2004. World Bank, Washington, DC. 2004. Measuring Results: Improving National Statistics in IDA countries. IDA, Washington, World Bank. 2005. The GDF & WDI Central DC. (April 2005) database. Accessed August 22, 2005 at International Labor Organization/His Majesty's Government of Nepal. 2003. Integrated Rural Accessibility Planning Guidelines. Bangkok: ILO ASIST Asia Pacific. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-1 LIST OF ANNEXES Annex I: Rural Access Index: Summary Sheet Annex II: Rural Access Index by Region: IDA Countries Rural Access Index by Region: IBRD Countries Annex III: Rural Access Index by Region: IDA and non-IDA countries (including estimates for 32 countries) Annex IV: Household Level Transport and Access Module Annex V: Time Distance Survey: Sample Documents Annex VI: AFRICA: Transport Targets and Indicators related to the Millennium Development Goals (MDGs) Annex VII: Vietnam - Provincial Allocation to Improve Rural Access RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-3 ANNEX I: RURAL ACCESS INDEX: SUMMARY SHEET Definition `Rural Access Index' is the percentage of rural people 100% 93% who live within 2 km (typically equivalent to a walk of 20 Rural Access Index: Aggregate Average 90% minutes) of an all-season road as a proportion of the 80% total rural population. An "all-season road" is a road that 72% is motorable all year round by the prevailing means of 70% 61% rural transport (often a pick-up or a truck which does not 60% have four-wheel-drive). Predictable interruptions of short 50% duration during inclement weather (e.g. heavy rainfall) 37% are accepted, particularly on low volume roads. 40% 30% Aggregation 20% 10% Results for 32 IDA countries (representing 83% of the total rural population in all IDA countries) show that on 0% IDA-eligible countries (31 countries) 1 average 61% of rural dwellers have access to the IDA-only countries (24 countries) Blend countries (7 countries) transport network. The graph also shows a significant Non-IDA countries (8 countries) difference between `IDA only' and `blend' countries. A sample of 10 IBRD countries has much better access (93%) on average. Individual country data was weighted by population. IDA Countries Included: Albania, Azerbaijan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Cameroon, Chad, Ethiopia, Gambia, Ghana, India, Indonesia, Kenya, Lao PDR, Madagascar, Malawi, Mali, Mongolia, Nepal, Nicaragua, Niger, Nigeria, Pakistan, Papua New Guinea, Tajikistan, Tanzania, Uzbekistan, Vietnam, Yemen, Zambia. IBRD Countries: Belarus, Brazil, Bulgaria, China, Latvia, Namibia, Peru, Romania, Russia, South Africa Relevance to Poverty Reduction and MDGs Physical isolation is a strong contributor to poverty. Populations without reliable access to social and economic services are poorer than those with reliable access. Problems of access are particularly severe in those rural areas which are distant from roads that carry motorized transport services on a regular basis. It is estimated that about 700 million rural dwellers in developing countries (the great majority of them living in IDA countries) are without reliable access. The large majority of these people are poor and they include most of those who are extremely. Sustainable access to rural transport is not an indicator for the Millennium Development Goals, but it is a key contribution to achieving many of the Goals - underpinning pro-poor growth and improving social inclusion. Surveys have shown that poor people view isolation as a major contributor to their poverty and marginalization. Therefore, improving access to roads for rural dwellers is considered essential to promote rural development, to improve access to human development services, to raise incomes and to stimulate growth for poverty reduction. While sustainable access to rural transport is a good indication of the shortfall in rural access, the full picture requires more detailed information. Before upgrading tracks or paths to motorable condition, the availability of affordable transport services that will be operated on the new road should be confirmed. Sensitivity to Change The Rural Access Index adequately captures changes in performance over time ­ usually a period of three years or so. An analysis of preliminary measures for two countries with time-series data (during the period of 1997-2002) confirms that the rate of change in the index is likely to be less than 1% in a year for a large country. In small countries or sub-national areas widespread investment in or neglect of the rural network can result in more rapid change. Relatively small improvements in access to an all-season road have marked impact in terms of improving the Index value on a local basis. Current measures of this indicator show a significant difference between IDA (aggregate average of 64%) and IBRD countries (aggregate average of 93%). Amongst the IDA countries there is also a significant difference between those which receive blended assistance (aggregate average of 72%) and those which do not (aggregate average of 45%). Thus, increasing the number of rural people with sustainable access to rural transport is expected to have high and sustainable impact. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-4 Measurability and Reporting There are two main approaches to the measurement of this index, by using: (a) household surveys that include information about access to transport, or (b) map data to determine how many people live within the specified catchments of the road network. Including appropriate questions in a broader survey of households is usually the most informative and cost-effective way of obtaining information on rural access. The majority of the 32 RAIs which are available for IDA countries have been established using the results of household surveys which already included questions on access. These surveys are designed to produce high-quality data and be representative for the main segments of population (thus, the `rural population' is adequately covered). In establishing this new index, priority has been given to the countries with the highest populations so that the 32 countries already covered represent 83% of the total rural population of IDA countries. Priority has also been given to those countries with larger land areas for which rural access is likely to be particularly relevant. Updates of the index will essentially depend on the frequency of household surveys. These are usually expected to be carried out on a three-year cycle. In collaboration with partner organizations, the World Bank is developing a work program to ensure the sustainability of this and other IDA indicators through regular updating of surveys and expansion of coverage to additional countries. It is planned to establish this indicator for an additional 20 IDA countries so it will be available for over 50 IDA countries. A time series will be established for several countries and there will be further benchmarking against `non-IDA' countries. Costs (implication for IDA borrowers and the Bank) The Bank is raising awareness of the significance of the indicator for capturing differences in performance between countries. It is expected that this initiative will help to `harmonize' rural access questions across household surveys (both for government and donor sponsored surveys) to encourage efficient use of resources by linking with the proposed global household survey network. The inclusion of the rural access index in the 2006 edition of World Development Indicators also is being considered. About 38 IDA countries already have completed national household surveys which include questions that permit this indicator to be estimated. In these cases the marginal cost of producing the indicator is generally up to one day of experienced statistical input. Where there is a suitable survey which does not include a relevant question, there will be a one-off cost to negotiate, design, test and incorporate the question. This cost is estimated to be in the order of three days in addition to the cost of analysis. Where there is no household survey it will be necessary to estimate the indicator by the mapping technique described above. The input for this is estimated to be about one month of analysis for each estimate of the indicator, provided that the necessary data on the location of roads and of the rural population is available. Application of the Rural Access Index Currently the primary application of the Index is as a component of the IDA -14 Results Measurement System on the lines described above. Various countries are also extending the application of the Index in different ways. In Vietnam the index is being updated every two years on the basis of data from the Vietnam Living Standards Survey. The VLSS is representative at the Provincial level so the index provides a basis for comparing the degree of rural accessibility enjoyed by the Rural population in each province. Procedures are being examined for using the Rural Access index together with the Poverty Rate for allocating central government resources between provinces in response to the national Comprehensive Poverty Reduction and Growth Strategy. As a headline transport indicator the Rural Access Index also has significance at both the national and the wider Regional level. The Africa Transport Ministers prepared a Declaration in 2005 showing the key targets, such as improving rural accessibility and urban mobility, for the transport sector to contribute to achieving the Millennium Development Goals. This note was updated by Peter Roberts (TUDTR): September 2005. ANNEX II: RURAL ACCESS INDEX BY REGION: IDA COUNTRIES RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-5 ANNEX II (continued) RURAL ACCESS INDEX BY REGION: IDA COUNTRIES RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-6 ANNEX II (CONTINUED) RURAL ACCESS INDEX BY REGION: IBRD COUNTRIES RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-7 ANNEX III: RURAL ACCESS INDEX BY REGION: IDA AND NON-IDA COUNTRIES (INCLUDING ESTIMATES FOR 32 COUNTRIES) RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-9 ANNEX III (continued) RURAL ACCESS INDEX BY REGION: IDA AND NON-IDA COUNTRIES (INCLUDING ESTIMATES FOR 32 COUNTRIES) RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-10 ANNEX III (continued) RURAL ACCESS INDEX BY REGION: IDA AND NON-IDA COUNTRIES (INCLUDING ESTIMATES FOR 32 COUNTRIES) RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-11 ANNEX III (continued) RURAL ACCESS INDEX BY REGION: IDA AND NON-IDA COUNTRIES (INCLUDING ESTIMATES FOR 32 COUNTRIES) RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-12 ANNEX IV: HOUSEHOLD LEVEL TRANSPORT AND ACCESS MODULE ROADS COMPONENT (DRAFT) RURAL ACCESS INDEX :A KEY D EVE LOPMENT IND ICATOR A-13 RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-15 ANNEX V: TIME-DISTANCE SURVEY: SAMPLE DOCUMENTS 1. Questionnaire administered 2. Enumerator instructions 3. Information for respondents RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-17 1. QUESTIONNAIRE ADMINISTERED TIME-DISTANCE MODULE - SCHOOL HOUSHOLD ID (1) HOW MANY MEMBERS IN THE HOUSEHOLD WALK TO SCHOOL? (2) HOUSEHOLD MEMBER CHOSEN FOR THIS MODULE PERSON ID FIRST NAME LAST NAME (3) NAME OF SCHOOL CHILD IS CURRENTLY ATTENDING ____________________________ (4) IS THIS THE SAME SCHOOL AS REPORTED IN KHDS QUESTIONNAIRE (1=yes, 2=no) (5) SEX (1= Male, 2= female) (6) AGE (years) (7) IS THIS PERSON 1. THE DESIGNATED RESPONDENT 2. A REPLACEMENT (8) RESPONDENT'S ESTIMATE OF DISTANCE TO SCHOOL FROM HOME* . Km (also equal to distance estimated back to home) (9) RESPONDENT'S ESTIMATE OF TIME TO WALK TO SCHOOL FROM HOME* HRS MINS (10) RESPONDENT'S ESTIMATE OF TIME TO WALK TO HOME FROM SCHOOL* HRS MINS * new estimates, DO NOT copy from questionnaire CALCULATING AVERAGE STEP LENGTH IF AN "ACCEPTABLE" READING, COPY ENUMERATOR PEDOMETER THE NUMBER OF PEDOMETER STEPS COUNTED STEPS STEPS MINUS 1 BELOW. (11) FIRST WALK SECOND THIRD FOURTH FIFTH SIXTH STEPS (12) TOTAL OF THREE ACCEPTABLE READINGS (13) CALCULATE THE AVERAGE STEP LENGTH: 15000/TOTAL IN QUESTION 15 CM ROUND TO NEAREST WHOLE NUMBER INSERT THIS AVERAGE STEP LENGTH IN THE PEDOMETER RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-18 RECORDING THE DATA FROM THE RESPONDENT AFTER THE PEDOMETER USE (14) DATE OF WALK FROM THE RESPONDENT SHEET DAY MONTH YEAR (15) To school Back to home Line 1. STEP (steps) Line 2. DIST (distance) . Kms . Kms Line 3. TIMER __ . __ __ . __ __ __ . __ __ . __ __ [HR. MIN.SEC] (16) CHECKING THE PEDOMETER AFTER USE FROM PEDOMETER AFTER USE: WITH MODE ARROW ON DIST, PRESS THE "SET" BUTTON AND RECORD THE NUMBER ON THE SCREEN (17) DESCRIPTION OF THE TERRAIN OF THE PATH WALKED (PLAINS=1, UPHILL/DOWNHILL =2) RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-19 2: TIME-DISTANCE MODULE: RECORDING DISTANCE AND TIME Enumerator Instructions Choosing a respondent from the household members (TO BE DONE BEFORE ARRIVING AT HOUSEHOLD): Here we describe the way sampling of respondents was done in the Tanzanian time-distance survey. There respondents who traveled to the school, and who collected water were sampled for the survey. Here is the sampling procedure adopted in the survey. (1) The Kagera Health and Development Survey (KHDS) was used to sample 9 clusters. Clusters were chosen purposively to provide geographical and ethnographical mix of data observations, as well as maximum number of questionnaires within the given time-frame. (2) Within these clusters all households with at least one non-boarding, school-going child above 12 years old (at the time of survey) were selected. (3) For households with multiple respondents fitting the criteria simple random sampling was used to select one respondent that should be interviewed. If this respondent was found to be no longer living in the household, deceased, traveling, or unavailable for other reasons, then a randomly (but still pre- determined) replacement was interviewed if available. A second respondent within these households was then chosen to conduct the water-source survey with, provided they were at least 12 years of age. Now we will prepare the pedometer to give to the respondent. Operating instructions (SPORTLINE 350 PROLINE PEDOMETER) What is a pedometer? The basic function of pedometer is to count steps. A liver which is inside the pedometer gets activated with the walker's every hip movements, counting each movement as one step. When you wear the pedometer and walk from one place to another, it counts the number of steps that you have taken in traveling this distance. The total number of steps that you have taken is then multiplied by the average length of your step to arrive at the overall distance between any two points. The machine also records the time taken by you in traveling the distance. General information before you begin 1. To open the cover of the pedometer, gently lift and push the plastic lip that we have painted red. 2. There are 5 blue buttons across the bottom part of the pedometer. The larger blue button on the right (marked E in the sketch) is the MODE button. Pressing the mode button moves a small black arrow in the screen to point to one of 5 functions. 3. Press the MODE button and watch the arrow point to the functions: STEP (counts the steps), DIST (distance ­ calculates the distance), KCAL (we will not use), CLOCK (we will not use) and TIMER (measures the elapsed time). 4. Push the MODE button until the small arrow points to DIST. On the left part of the screen you should see written KM (meaning kilometers). If you see written ML, you must hold down the RESET key (marked A in the sketch) for 5 seconds until the "ML" disappears and "KM" is displayed. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-20 5. Press the TOTAL button (marked D on the sketch) several times. Each time the TOTAL button is pressed, you will see a small word "TOTAL" come on or disappear at the top right of the screen above the numbers. You want the "TOTAL" word to not be displayed, so if it is written on the screen press the TOTAL button (marked D) to make it disappear. To calculate the average step length of the respondent In order for the pedometer to correctly calculate distance, we have to figure out the average length of the respondent's stride (step) and then enter it in the pedometer. 1. Use the 10 meter rope to mark out a length of track of 50 meters (5 lengths of rope) on a fairly straight and level road or path. Mark the start line and the finish line. 2. You are going to have the respondent walk the length of track several times to count the number of steps the respondent takes and compare it to the pedometer. You will continue until you have enough acceptable readings to calculate the respondent's step length. 3. Hold down the RESET button for two seconds to put all the values for STEP, DIST and TIMER back to zero. 4. Position the respondent on the starting line, with the toe of his or her right foot just behind and touching the start line. Close the pedometer and have the respondent put it in the correct place on the belt or waistband, near the hip but still on the front. The pedometer must be level. 5. Ask the respondent to walk the length of the track in his or her normal walking manner, and to take one step beyond the finish line, then stop and stand still. You should count the steps as the respondent walks to the finish line. 6. Record the number of steps you counted: write it in Question 7 on the first line, called "First walk" in the column for "ENUMERATOR COUNTED STEPS". 7. Go to the end of the track and show the respondent how to open the pedometer (preferably while the respondent is still wearing the pedometer) and show the respondent how to look at the STEP, DIST and TIMER. 8. Look at the number of steps recorded by the pedometer (press the MODE button until the arrow points to STEP). Record this on Line 1 below under "PEDOMETER STEPS minus 1", but subtract one from the number because of the extra step taken after the finish line. 9. Press the RESET button for two seconds to put the pedometer count back to zero, and have the respondent walk the track again, with you counting. Record the number of steps you counted and the number of steps (with one subtracted) from the pedometer count on the second line. 10. Press the RESET button and have the respondent walk for a third time and put the number of steps you count and the pedometer count (with one subtracted) on the third line. 11. Look at the first line: called "first walk". If the difference between your enumerator count and the pedometer steps is less than 4, the count is "acceptable". If the value is acceptable, put the pedometer step number in Column 3. If it is not "acceptable" do not write anything in Col 3. 12. Look at the second line. If the difference between your enumerator count and the pedometer steps is less than 4, the count is "acceptable". If the value is "acceptable", put the pedometer step number in Column 3. If it is not "acceptable" do not write anything in Col 3. 13. Look at the third line. If the difference between your enumerator count and the pedometer steps is less than 4, the count is "acceptable". If the value is acceptable, put the pedometer step number in Column 3. If it is not "acceptable" do not write anything in Col 3. 14. If you have three acceptable counts, you do not have to have the respondent walk the track again. If you do not have three acceptable counts, adjust the pedometer by shifting it slightly right or left of by placing it on the opposite side near the hip, and try some more times. You must have three "acceptable" measurements in all. 15. Once you have three acceptable measurements, you have to add them up and put the total as indicated. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-21 16. Now we have to calculate the average step length. The respondent has walked 150 meters for the three acceptable measurements. This is 15000 centimeters. We divide 15000 by the total steps calculated in 14, and this is the average step length. Round this number to the nearest whole number. This step length should be between approximately 50 and 100 centimeters. If it is not, check all your calculations. EXAMPLE: CALCULATING AVERAGE STEP LENGTH IF AN "ACCEPTABLE" READING , CO PY ENUM ERATO R PEDO M ETER THE "PEDO M ETER STEPS m inus 1" CO UNTED STEPS STEPS m inus 1 BELO W . (7)FIRST W ALK 78 84 SECO ND 79 81 81 THIRD 78 78 78 FO URTH 80 84 84 FIFTH SIXTH STEPS (8) TO TAL O F THREE ACCEPTABLE READING S= 243 CALCULATIO N (9)CALCULATE THE AVERAG E STEP LENG TH: 15000/TO TAL IN Q UESTIO N 8= 62 CM 15000/243=61.73 RO UND TO NEAREST W HO LE NUM BER (INSERT THIS AVERAG E STEP LENG TH IN THE PEDO M ETER) In the above example, in the first walk of 50 meters, the enumerator counted 78 steps and the pedometer recorded 85 steps. We subtract one from the pedometer reading because we have asked the respondent to take one step beyond the end, so we put in 84 for the pedometer steps. If we take the difference of the two readings: 84-78 the answer is 6, which makes these readings not "acceptable" since there must be a difference of 4 or less to be acceptable. So we cannot put the pedometer count in the third column, and we leave it blank. But the next three readings are acceptable, and for each acceptable reading we put the pedometer step count in column 3. Once we have three acceptable readings we add the three counts in Column 3, so we add 81+78+84 to get a total of 243. Now we make the calculation of 15000 divided by 243: 15000/243=61.73. This is the average step length in centimeters. We round it to the nearest whole number which is 62. This is the number we must put in the pedometer. (This is quite a small step length, probably associated with a person who is short. The step length of the average person is about 75 cm.) To put the average step length into the pedometer: 1. Press the `MODE' button until the arrow reaches the `DIST' display and check that this shows `KM' on the left of the screen (if not, then refer above to "general information before you begin", number 4 to see how to change to KM.) 2. Press the `SET' button (marked B on the sketch) once. The reading which flashes is the current `step length'. If this has to be increased a little that can be done one digit at a time by repeatedly pressing the `SET' button. For instance if the current step length is 75 and you wish to set a step length of 88 cm you can press the SET button once to get to 76, again to get to 77, again to get to 78, etc. until you get to 88. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-22 3. If you have to set a number that is LESS than the current `step length' you must hold down the SET button to make the numbers go by more quickly ­ the reading will go up until 240 and then will return to the minimum of 30 and begin going up from 30. For example, if the current step length is 75 and you wish to record 62, you hold the SET button down while it goes all the way up to 240 then once it reaches 240 it goes down to 30 and begins going up from there. Once you get to about 55 you should stop holding down the SET button, and instead just press it time by time to go to 56, 57, 58, and so on to 62. 4. Once you are at the correct number you want to input for the step length, press the MODE button to get out of the average step setting. (If you want to check that the step length is properly input, put the MODE arrow on DIST, press the set button once, and you should see the number you have put in flashing. Press MODE again to make sure you do not change the step length.) Giving instructions to the respondent 1. Give the two page information sheet/form to the respondent. 2. Put in the Issued Panel Household ID number and the name of the respondent on the form. 3. Go over the instructions on the information sheet to ensure the respondent understands what to do. 4. Arrange for a time to return to pick up the information sheet/form and the pedometer. 5. Once you pick up the pedometer, press the MODE button until the arrow is on DIST, and press the SET key once. Record the number on the screen in Question 12 on the Module 14 form. RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-23 3: TIME-DISTANCE MODULE: RECORDING DISTANCE AND TIME Pedometer information for respondents Thank you for agreeing to record the distance you walk to school or water source. What is a pedometer? The basic function of pedometer is to count steps. A liver which is inside the pedometer gets activated with the walker's every hip movements, counting each movement as one step. When you wear the pedometer and walk from one place to another, it counts the number of steps that you have taken in traveling this distance. The total number of steps that you have taken is then multiplied by the average length of your step to arrive at the overall distance between any two points. The machine also records the time taken by you in traveling the distance. General information before you begin (Sportline 350 Model) 1. To open the cover of the pedometer, gently lift and push the plastic lip that we have painted red. 2. There are 5 blue buttons across the bottom part of the pedometer. The larger blue button on the right (marked E in the sketch) is the MODE button. Pressing the mode button moves a small black arrow in the screen to point to one of 5 functions. 3. Press the MODE button and watch the arrow point to the functions: STEP (counts the steps), DIST (distance ­ calculates the distance), KCAL (we will not use), CLOCK (we will not use) and TIMER (measures the elapsed time). 4. Push the MODE button until the small arrow points to DIST. On the left part of the screen you should see written KM (meaning kilometers). If you see written ML, you must hold down the RESET key (marked A in the sketch) for 5 seconds until the "ML" disappears and "KM" is displayed. 5. Press the TOTAL button (marked D on the sketch) several times. Each time the TOTAL button is pressed, you will see a small word "TOTAL" come on or disappear at the top right of the screen above the numbers. You want the "TOTAL" word to not be displayed, so if it is written on the screen press the TOTAL button (marked D) to make it disappear. 6. We will not use the SET button (marked B on the sketch) or the SCAN button (marked C on the sketch). When you are ready to record your walk: 1. Press the MODE button (marked E) until the small arrow is pointing to DIST. If the number on the screen is not zero, press the RESET button (marked A) for two seconds until the number resets to zero. 2. Check that KM is displayed on the left side of the screen. (if it says ML, refer to number 4 above.) 3. Close the pedometer and use the hinge to put it on your belt or waistband in the correct position shown to you, in line with the crease on the left or right front, and level. 4. Place the pedometer on your belt, always in a vertical position. Do not hang it on the side of the belt or waist band. Keep in mind that movements such as pulling up pants, bending over to tie your shoe can register a step. 5. Walk to school or waters source trying to keep the same regular pace you used when the enumerator measured your steps. You can stop on the way and the pedometer will stop RURAL ACCESS INDEX: A KEY DEVELOPMENT INDICATOR A-24 recording. It will start recording again when you start walking again. If you want to check the readings on the way, you can open the cover and see the readings, but you must close the cover again before continuing to walk, or the machine will not record. 6. When you get to school or waters source, you must stop, stand still, open the pedometer cover and press the MODE button to show the readings: · First press the MODE button until the arrow is on STEP and record the number of steps on Line 1. · Second, press the MODE button again until the arrow is on DIST and record the distance in kilometers. Please write all 4 or 5 digits on Line 2. · Third, press the MODE button twice more until the arrow is on TIMER, and record the time, writing all the 5 digits displayed, on Line 3. 7. The pedometer cannot be turned off. Just keep it in a safe place until you are ready to return home. 8. When you are ready to return home check that you have this recording sheet. 9. Open the pedometer cover and press the MODE button until the small arrow is pointing to DIST. Press the RESET button (marked A) for two seconds until the number resets to zero. 10. Close the cover and put the pedometer in the correct position on the belt or waistband. 11. Return by the same route that you took to get to school or waters source. 12. When you get home, stop, open the cover and record the readings "to home" in the same way as described in Number 5 above. Do not worry if the numbers are different. 13. Put the recoding sheet and the pedometer together for pick-up by the enumerator. ISSUED PANEL HOUSEHOLD ID_________________________________ NAME:____________________________Date of walk:____________________________ Day Month Year WALKING TO (underline one): 1. WATERS SOURCE or 2. SCHOOL To waters source or school Back to home Line 1. STEP (steps) ___________ ___________ Line 2. DIST (distance) _____ . __ __ __ _____ . __ __ __ Line 3. TIMER (minutes) __ . __ __ . __ __ __ . __ __ . __ __ ANNEX VI: AFRICA: TRANSPORT TARGETS AND INDICATORS RELATED TO THE MILLENNIUM DEVELOPMENT GOALS (MDGS) MDG Targets Indicators MDG 1 Access to inputs and markets, and generation of employment Proportion of rural population within 2 km of an all season road Eradication of extreme poverty opportunities, improved by halving the proportion of rural and hunger population living beyond 2 km of an all-season road % Reduction of travel and vehicle turnaround time % Increased productivity in agriculture and economic activities % Increase in employment opportunities and income generation from transport related activities RURAL The difference in average transport cost between Africa and % Reduction in passenger fares (passenger kilometer) Asia narrowed down by 50% % Reduction in unit goods transport cost (ton kilometer) ACCESS Level of affordability of transport cost by the urban and rural poor % Increase in the use of intermediate means of transport (IMT) INDEX Existence of sustainable financing mechanisms like Road Funds... % Increase in the proportion of roads in good and fair condition :A MDG 2 + 3 Rural access and urban mobility improved to eliminate % of schools which have reliable access KEY Universal primary education and constraints on the time which all children have to participate % of households which report constraints on education due to: D EVE gender equality in education and to enable effective education to be delivered Lack of girls time for school and reached safely Difficulty (cost) of access Poor quality of education service Lack of safe access to school LOPMENT IND MDG 4 + 5 Rural access and urban mobility improved for reliable % Health centers, clinics etc with reliable rural access. Child Health and Maternal supply of inputs to health facilities, to provide affordable % of households reporting constraints on access to health services Mortality access for all households and to enable cost effective outreach because of: ICATOR health activities Distance Cost / difficulty of travel A-25 Poor quality health service Unit cost immunization / capita Unit cost / coverage of outreach services / capita Emergency transport response for medical crisis in rural % Emergency patients unable to reach health care in time: communities improved through community communications Expectant or postnatal mothers facilities linked to improved transport services Children under 5 years ANNEX VI (continued) AFRICA: TRANSPORT TARGETS AND INDICATORS RELATED TO THE MILLENNIUM DEVELOPMENT GOALS (MDGS) MDG Targets Indicators MDG 6 Ensure transport sector ceases to be an agent for spreading HIV/AIDS Prevalence among transport sector workers (public and HIV/AIDS, malaria and other HIV/AIDS private) diseases HIV/AIDS prevalence rate in transport affected communities Inter-country coordination of actions relating to AIDS in transport sector Rate of road accident fatalities reduced by half by 2015 Rate of fatality (per million vehicles-km) RURAL Number of countries adopting road safety strategies MDG 7 Share of urban residents for whom mobility problems % of households (in the various urban living environments) which Environmental sustainability severely constrain access to employment and essential report transport costs and time as major obstacles to employment ACCESS services halved % of households which report access as a major obstacle for essential INDEX services Environmental sustainability promoted in all transport Environmental impact identified by audits of programs undertaken :A operations and development programs KEY Production of leaded petrol ceased by 2010 Number of countries banning sale of leaded petrol D MDG 8 Transport cost for landlocked countries reduced by half and Percentage reduction of missing links of the Trans-African Highways EVE Global partnership for their access to global markets improved, all TAH missing (TAH) network and transit corridors. development links completed and existing portions of regional transport % reduction in transport cost for landlocked countries corridors maintained by 2015 LOPMENT All non-physical transport barriers that increase journey time, Proportion of countries that have reduced checkpoints along their main IND customs clearance, border delay and impede the flow of goods transit corridors to a maximum of 3 (between port and border of land- and services dismantled by 2015 locked country). Proportion of countries that have reduced their border crossing time to ICATOR OECD average. Proportion of countries that have reduced their port clearing time to A-26 OECD average. Axle load limits, vehicle and road technical standards Proportion of RECs with harmonized axle load limits harmonized between RECs by 2015 Proportion of RECs with harmonized standards for vehicles Proportion of RECs that have harmonized road design standards Air transport services improved fares reduced, and Number of new connections between African countries established. movement of goods and services facilitated in all African Number of products and volume of traffic of products transported by countries by 2015 air. Percentage reduction in air transport fares. Source: AU / AfDB / UNECA / ECOWAS / WB, April 2005. ANNEX VII: VIETNAM ­ PROVINCIAL ALLOCATION TO IMPROVE RURAL ACCESS 7 ) 6 mill $SU( allocation = base + fn (poverty rate) . (population) . (lack of access) 5 RURAL cen 4 ACCESS INDEX ovi pr/ 3 :A KEY on 2 D EVE ocati 1 all LOPMENT IND 0 Son Lao nh i AnBangGianga BinHa h g HueDinh Ye n g h Ca HoaCh au BienCan Sonn Bai ai ho m ucTay Binh ong ICATOR La a Namai Binhc Ninh Bac Ye o an an Na uanTT uyenang TriGianm Dinng Yenh Ph c Ha i DuHa Th LaiDien Lang NgheCa Ha Ho Qu Qu Tinhg Binhg NgPhu Tangen Q BinhPhu ai NgQu Na Hu Th Ba Qu Ba Vin Ninh Ha A-27 Tuy Th province - ranked by rural access need Note: This illustrates a possible allocation of US$ 90 million between the 33 Provinces identified to most require support for improving Basic Rural Access. Source: Peter Roberts's working draft of March 2005