33907 THE WORLD BANK GROUP WASHINGTON, D.C. TRANSPORT PAPERS TP-5 JANUARY 2005 Development of a Transport Module for Multi-topic Household Surveys Judy L. Baker and William Denning TRANSPORT SECTOR BOARD © 2005 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 The findings, interpretations, and conclusions expressed here are those of the author and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. 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/ DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS iii TABLE OF CONTENTS 1 INTRODUCTION ...................................................................................................... 1 2 TRANSPORT AND MULTITOPIC HOUSEHOLD SURVEYS ........................................................... 1 3 POLICY CONCERNS AND DATA NEEDS ............................................................................ 3 3.1 Access and Mobility .................................................................................................. 3 3.2 Affordability ............................................................................................................ 3 3.3 Quality ................................................................................................................... 3 3.4 Data Needs ............................................................................................................. 4 4 DATA ANALYSIS FOR TRANSPORT ................................................................................. 5 4.1 Core Analysis........................................................................................................... 5 4.2 A Transport and Poverty Profile .................................................................................. 7 4.3 Monitoring and Evaluation ......................................................................................... 8 4.4 Rural Transport Problems, Lack Of Basic Access And Mobility .......................................... 9 4.5 Urban Transport Problems, Symptoms Versus Cause Of Poverty...................................... 9 5 SURVEY DESIGN ISSUES ......................................................................................... 10 5.1 Sampling ...............................................................................................................10 5.2 Survey Administration..............................................................................................11 5.3 Travel Diaries .........................................................................................................11 5.4 Use of GIS .............................................................................................................12 6 INTEGRATING TRANSPORT QUESTIONS INTO HOUSEHOLD SURVEYS ........................................ 12 7 REFERENCES ....................................................................................................... 15 Annex I: Sample Questions on Access, Mobility, Affordability and Quality ........................16 A. Household Survey........................................................................................................16 Annex II : Welfare Aggregates ..................................................................................22 Annex III: List of Transport Indicators and Use of Household Survey Data ......................24 Endnotes ...............................................................................................................27 DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS iv ABSTRACT This paper is aimed at providing guidance on transport issues for those involved in designing multitopic household surveys such as the Living Standards Measurement Studies (LSMS) surveys. The inclusion of a few key questions can provide critical information for better designing transport programs and policies aimed at improving access, affordability and quality services. Questions on transport access, quality, mode, distance, time, and cost can help to understand the constraints that the population may face in accessing jobs, markets, schools, health clinics and other social services. All of this can be broken down by subgroups such as income, geographic area, gender, employment, etc. further strengthening the relevance of the analysis and contribution to policy decisions. The paper covers background on transport and multitopic household surveys, key transport policy concerns and data needs, approaches to analysis, issues of survey design, and prototype questions that could be included in existing surveys. Special thanks to Kinnon Scott and participants at the World Bank seminar on this topic in November 2004 for their comments. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 1 DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 1 INTRODUCTION 2 TRANSPORT AND MULTITOPIC HOUSEHOLD SURVEYS Multitopic household surveys are now carried out in a large number of developing countries In order to design and analyze results from to analyze a range of economic and social household surveys, transport practitioners issues, and in some cases on a frequent basis. need to understand the context in which These surveys have been effectively used for household surveys are created and policy analysis, planning, poverty monitoring, administered. Generally household surveys assessing the distributional aspects of policy, have been undertaken through a national and for sector specific analysis. Key transport statistical agency, often with support from issues such as access, quality, safety and international organizations and donors. These affordability can also be covered in multitopic surveys can be for strategic purposes; surveys, sometimes with the addition of a analytical work, poverty reduction strategy minimal number of key questions. This can be papers, country assistance strategies, particularly important in better tailoring monitoring and evaluation, etc., but also for transport programs and policies to meet the routine purposes; price indices, labor force needs of the poor by understanding their travel estimates, national accounts, etc. Other needs and travel behavior. Relevant surveys are also carried out to support special information includes descriptive statistics on purpose and project related surveys, which travel patterns and time use, access to jobs, may or may not be household based. schools, services, travel demand, travel conditions, affordability, and the incidence and The Living Standards Measurement Study coverage of government programs. All of this (LSMS) household surveys have been in use by can be broken down by subgroups such as World Bank client countries since the 1980s. income, geographic area, gender, employment, They are national probability sample surveys etc., further strengthening the relevance of the conducted irregularly as required, ideally every analysis and contribution to policy decisions. three to six years. Their focus is to administer a relatively complex questionnaire to a This paper presents the tools for developing a moderate size of sample (2,000 to 3,000 country-specific transport module to be households). The questionnaire is divided into included in a household survey, particularly the sections referred to as "modules" covering a LSMS surveys. Producing a "standard" single sub-topic. Anywhere from eight to transport module for household surveys can be twelve modules may be combined to form the challenging, as with all standard survey complete questionnaire. Some modules, such modules, as each application is sufficiently as the Household Roster, Education, Health, different. Nevertheless experience provides and Economic Status, are almost always used. lessons and this paper suggests a survey Annex Two provides a listing of the standard approach and standard survey module for modules, a short description of their contents, transport, along with further notes to allow and illustrates how transport relevant each practitioner to adapt the example to the information can be extracted from the existing specific policy and planning issues to be modules. explored. Section 1 provides this introduction; Section 2 covers transport and multitopic Substantial LSMS survey resources are household surveys, Section 3 reviews policy devoted to quality control, and the tradeoff concerns and data needs; Section 4, between sampling and non-sampling errors approaches to analysis; Section 5 discusses since the focus is not purely on the issues of survey design; and Section 6 measurement of an indicator, but instead on presents sample questions that could be understanding the causes of observed included using the sample modules. outcomes. This requires a focus on ensuring DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 2 within questionnaire consistency more so than broader coverage and analysis of transport in a larger sample survey, 1 ensuring the issues. quality of response of the complex questionnaire rather than the quantity of Survey research is a part of an investigative responses. The goal is to explore in detail the process which develops over time. Results relationships of individuals and their from one step will change the approach used in households to determinants of poverty. The a later step. It will include use of qualitative use of careful sampling ensures that the methods and non-household surveys that are results can be generalized to the national level not discussed here. In transportation other for strategic policy goals. The sample design common types of survey work include (among permits disaggregation of results for ethnic others): classified traffic counts, vehicle groups or to the rural-urban level. 2 occupancy counts, vehicle intercept surveys, origin-destination surveys, license plate The LSMS surveys have evolved considerably surveys, and enterprise surveys of over time. In the last five years more attention infrastructure providers, transport service has been paid to issues of infrastructure and providers, and transport users. its related services and most recently service levels and quality. An additional survey Information relevant for transport can be instrument, the Community Questionnaire or obtained through the LSMS approach by Population Point Questionnaire, is also used to editing existing survey modules to add or capture information from outside households change questions, and by modifying the coding but relevant to the households of a single of responses received to include more sampling location. This generally covers the transport relevant possibilities. Transport territory of a village or neighborhood. It is information can also come from an additional administered to a small set of informed local survey module specifically on transport use. inhabitants (such as community leaders) and is Changes to the questionnaire will require an extremely useful extension of the LSMS complementary changes to the survey methods for transport purposes. administration, in particular the interview guide which provides instructions to the Income and expenditure surveys (also know as interviewer, especially on how to probe for household budget surveys) are more more meaningful responses. To date, there are commonly available than the LSMS. In most only a few existing examples of surveys that countries they are carried out infrequently, have incorporated extensive transport every 5-10 years, though in FSU, Europe and modules. Among these are Vietnam LSMS North America, they are done annually (with (repeated), Guatemala LSMS (2000), data often being provided on a quarterly basis. Azerbaijan (Sigma, 2000). These surveys have The Income and Expenditure surveys generally generated new analysis and contribute to a do not cover personal and intra-household small, but growing body of literature on socio-economic situation and relationships to transport using household survey data. the same extent as the LSMS, though they can provide income and expenditure data A second approach is to carry out a Transport (including transport expenditures), and some Survey, incorporating some of the core socio-demographic material (economic status, modules from multi-topic surveys. Using this highest education). 3 Sample sizes vary from approach, it is important to ensure that the country to country, and in some regions, such approaches for core information such as as Latin America, data are collected for urban location, household demographics, and income areas only. and expenditure are designed to be comparable to similar sample design and Other surveys which also measure aspects of questionnaire design formats in the country. living conditions include the Demographic and Including questions on other infrastructure Health Surveys (DHS), the Multiple Indicator sectors (water, communications, energy, Cluster Survey (MICS) the Core Welfare housing) can provide the basis for anlayiss of Indicators Questionnaire (CWIQ), and multi-sectoral urban, rural development, and censuses. These surveys are highly standar- infrastructure projects at a relatively small dized as they are used for international marginal cost. Where up-to-date poverty comparisons. It is generally not possible to analysis is available transport research should measure welfare in monetary terms, nor do make use of the already established welfare they typically include questions on transport. aggregate methodology and the Bank-accepted Including a few key transport questions, poverty line estimates. For maximum however, could provide an opportunity for relevance to broader poverty work, transport DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 3 surveys should incorporate the questions below. Travel forecasting or transport needed to calculate an acceptable welfare service/facility design is difficult with aggregate. household surveys because the sample sizes are too small for forecasting purposes. Household surveys can be supplemented by qualitative approaches which provide important 3.1 Access and Mobility information from respondents that may not come across in a coded response. Qualitative A fundamental role of transport is in providing approaches include techniques such as focus physical access by the population to jobs, groups and open ended interviews, and markets, education, health, and other services. generally use small sample sizes. This includes Assessing the travel behavior of individuals and subjective data on the attitudes and reasons households is thus key to understanding which underlie travel behavior and public patterns of access and mobility in both rural transport use. Ideally, a study methodology and urban areas. This is particularly important would include both a statistically for the poor, as the lack of access can deprive representative household survey, and a smaller them of job opportunities, the accumulation of number of qualitative surveys from the same human capital, and markets for trading their sample. goods. With an understanding of specific travel patterns for by mode, travel time, travel cost, 3 POLICY CONCERNS AND DATA NEEDS and reliability, appropriate policies can be designed to address the needs of the poor, and Transport plays a central role in development other subgroups which have traditionally been and poverty reduction. The role of transport in excluded to access such as the elderly and poverty reduction is linked in several ways. disabled. First, it impacts directly through providing poor people better access to jobs, markets, schools, 3.2 Affordability health clinics, and other social services. Second it contributes to economic growth by Beyond physical access is whether transport lowering transaction costs, promoting services are affordable to the poor. Fares are economies of scale, widening opportunities and often beyond the reach of the poor, making extending connection to rural hinterlands, existing services inaccessible. While we expanding trade, integrating markets, and actually know relatively little about what strengthening effective competition. Third, it households spend on transport, this kind of facilitates empowerment by providing access to information is possible to analyze with meeting places and town centers, and by household survey data. One can quantify allowing for the flow of information. Finally, a existing spending patterns for the household, reliable transportation system can enhance what is currently spent on transport (in the security by helping a country respond to aggregate and as a percent of total household economic and natural shocks, such as spending), and simulate the impact of fare transporting food during famines, or moving increases and reductions on different income people or medicines during medical groups. This can have a powerful effect on emergencies. policy decisions about fares and the impact of targeted transport subsidies for the poor. Interventions in the transport sector include a range of policy reforms and physical 3.3 Quality investments which may vary in rural and urban areas and across countries depending on need. Access to reliable, safe and secure transport is These might include physical construction of also critical to effective mobility. Bus services highways, rural roads, or ports, or institutional that do not consistently follow their routes, or reforms such as with rail systems, road provide an unsafe environment for women are maintenance, or road safety programs. inadequate to meeting the needs of the population. Issues of quality can be addressed Table 1 summarizes areas of concern in through household surveys by questions on the transport strategy and policy. Many of these conditions of existing services, as well as issues are not best or solely addressed through perceptions of these services. This information household survey based research. The areas can be disaggregated by subgroups such as that are most relevant are related to the links income or gender, further defining the nature between poverty and transport, in particular and scope of problems. Issues of traffic safety with regard to access and mobility, are difficult to capture using household surveys affordability, and quality and are discussed because the incidence in a sample population DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 4 is likely to be relatively small, not allowing for of analysis, specialized surveys are generally statistical significance in analysis. For this kind required. Table 1. Transport Policy Concerns. Policy Area Policy Goals and Interventions Relevant for Household Survey Based Analysis Access and Mobility • Improve physical access to employment, markets, schools, health * centers, and other social services • Reduce excessive time spent walking * • Promote inclusive access for poor, elderly, disabled, women * Affordability • Ensure affordable transport services for the poor through targeted * programs • Develop progressive fuel levy schemes for resource mobilization * Quality • Promote reliable, safe and secure transport services and * infrastructure • Support adequate maintenance of existing infrastructure • Implement road safety strategies Economic and Financial • Make transport more cost effective through competitive markets * Efficiency and for services and infrastructure, increased efficiency in the provision Sustainability financing and management of transport infrastructure. Provide adequate compensation if efficiency measures lead to labor redundancies • Ensure an adequate, reliable, and sustainable source of financing for maintenance, including a governance structure with user and private sector representation • Support cost effective, labor intensive approaches where appropriate for constructing and maintaining roads Environmental • Reduce threats to life and health from congestion and pollution. Sustainability • Promote an environmentally sensitive framework with standards for the provision of non-motorized transport, urban mass rapid transit, and road user charges that reflect externalities. Institutional • Introduce innovative financial and management techniques in the Development provision of transport services • Build capacity in the public and private sectors to enable both to perform their respective roles effectively and to encourage partnerships • Develop participatory mechanisms for stakeholders such as users and the private sector and users in monitoring and planning 3.4 Data Needs commonly found on transport use, mode, travel time, cost, and quality. The following The data that are needed for analyzing the matrix in Table 2 provides an overview of the issues raised above range from existing type of data needed for selected transport information on household and individual policy issues. Specific questions are discussed welfare and gender to questions that are less in section 6. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 5 Table 2. Data Needs for Transport Policy Analysis. Policy Area Data needs Household Questionnaires Community Questionnaires Access and • Individual transport use to job, school, health, • Physical access to all services Mobility market, road, water, fuel, other business by: transport use (including transport of - mode agricultural and other products), for each: - travel time/distance - frequency of trip - modes - travel time/distance • Household vehicle ownership Affordability • Travel expenditures per trip (individual and • Costs by mode/service business) • Subsidy programs availability • Transport services in kind • Participation in subsidy programs • Travel expenditures as a proportion of total expenditures Quality • Frequency/reliability of service • Frequency/reliability of services • Conditions of infrastructure (roads, closures due to flooding) • Condition of infrastructure • Perceptions of quality (e.g. reliability, safety, • Perceptions of quality (e.g., security and comfort) reliability, safety, security, comfort) 4 DATA ANALYSIS FOR TRANSPORT which are to be used in poverty analysis must obtain the data required for each 4.1 Core Analysis step described above. A standard reference for more information on welfare Different types of analysis can be carried aggregates of is "Guidelines for out for the above issues using household Constructing Consumption Aggregates for data. Since much of the work is focused on Welfare Analysis", Angus Deaton and the distributional aspects of transport Salman Zaidi, LSMS Working Paper No. policies, the analysis usually begins with 135, May 2002. An excerpt from this the construction of a welfare aggregate. 4 publication summarizing some of the considerations in constructing a welfare The welfare aggregate is normally aggregate is provided as Annex Two. constructed from information on a household's patterns of consumption rather Descriptive analysis is used to characterize than income. Constructing a consumption- individuals and households, and group based measure of individual welfare them according to their socio-economic typically requires three steps: (i) situation. Quintiles or deciles are often aggregation of different components of used rather than poor and non-poor household consumption to construct a groupings to allow for distributional nominal consumption aggregate, (ii) analysis. Cross tabulations can then link construction of price indices to adjust for subgroups with the transport variables of differences in prices faced by households, interest. Comparisons can be made and (iii) adjustment of the real between groups of individuals or consumption aggregate for differences in households and, if more than one survey is household composition. Household surveys available, over time. For example, DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 6 rural/urban or regional breakdowns might and households put on reductions in time be particularly useful for transport and money costs, by income group. analysis. Comparing changes in access and travel mode by the poor over time might The example in Table 3 looks at the be another type of analysis relevant for determinants of household spending levels transport policy. in Cote d’Ivoire. Specific to transport, it indicates that access to roads has The household survey data can also feed substantial predictive power – households into more sophisticated analysis to located in villages that are nearer to both "decompose" poverty and examine poverty paved roads and public markets are better correlates using regression and inequality off, as are households located in areas with analysis. Regression analysis can quantify higher wage levels. This points to the the effect of specific household or importance of rural roads in helping individual characteristics on poverty while families out of poverty. It also indicates holding constant all other characteristics. that education plays a different role in Transport variables may shed light on the urban and rural areas (where it does not linkages between transport issues (e.g. seem to significantly influence access) and poverty. Models of travel consumption), as do different types of demand and residential location choice can assets. This raises policy questions about be used to assess the value that individuals the quality of education in rural areas. 5 Table 3. Determinants of Household Spending Levels in Cote d’Ivoire. Variable Urban Rural Distance to the nearest: Paved road - -0.04 (-2.9) Market - -0.09 (-3.3) Education level of most educated male Elementary .38 (5.3) 0.04 (0.6) Junior Secondary .62 (8.6) 0.08 (0.9) Senior Secondary .80 (9.6) 0.05 (0.4) University .93 (9.4) - Education level of most educated female Elementary .11 (1.7) 0.07 (1.0) Junior Secondary .24 (3.1) 0.27 (2.2) Senior Secondary .34 (3.4) - University .52 (4.1) - Value of selected household assts Home .06 (5.3) - Business assets .04 (3.3) 0.16 (4.9) Savings .08 (4.7) - Hectares of agricultural land Cocoa trees - 0.17 (4.3) Coffee trees - 0.04 (1.3) Unskilled wages (males) - 0.37 (6.4) - Not applicable. Note: T statistics are in parenthesis Sources: Poverty Reduction Sourcebook, Chapter 1, and Grosh and Munoz, 1996 (p.169) . Among the subgroups that can be used for ethnicity, education level, household size, analysis, depending on the objective of the family composition, migration status, research, are: location, gender, age, employment, sources of income, quintile, DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 7 education outcomes, health outcomes and by mode of travel used, by purpose of trip, social outcomes (participation, access to and by time of travel. Additionally, information, social contacts). estimates of distance and time are likely to be skewed, particularly in rural areas The potentially relevant transport variables where people may not be accustomed to include: mode(s) used, trip rates, average wearing watches, and the estimates do not travel distance, average travel time, account for terrain. A pilot has been variability of travel time, out-of-pocket initiated in Albania by the World Bank to costs, etc. and can then be compared explore measurement comparing relative to group averages from within the perceptions of time and distance with transport survey (including by objective measures. This research has welfare/income group) and to examples or used pedometers and timers by norms established with reference to other respondents, though thus far the results work. are inconclusive. This is an area requiring additional research. Surveys should In surveys, it is important to develop generally include both measures so that questions and ways of coding responses later analysis can cross-check the results that capture poverty related behavior, to eliminate gross discrepancies and use usually specific to the country or society. the best (consistent, reasonable) results as Trip purpose is often classified as; school- required. Adding the use of GPS by related, work-related, social-related, and enumerators who can measure the other personal business-related. Better coordinates of particular places, understanding will result from particularly for the community surveys, is distinguishing travel undertaken for an effective solution for measuring household use (e.g. getting water, getting distance (use of GPS discussed in next fuel, child-care trips), and economic section). purposes (e.g. sales of home production goods, obtaining of inputs to home 4.2 A Transport and Poverty Profile production, looking for work). It is also important to identify when trips are taken To determine how the transport sector can by combinations of individuals and why, best help to reduce poverty, it is first and the presence of cargo and how it is necessary to assess the needs of poor transported. people and how much the poor require transport to meet those needs. Household Mode of travel needs to be recorded using surveys are ideal for generating such as broad as possible a list of responses in information. Descriptive analysis can order to capture the full range of coping provide a snapshot of access by the poor mechanisms employed by the poor in to essential services, by mode, cost, travel transport. Rather than looking at travel time, and select measures and perceptions mode as a physical characteristic it should of quality. Information can be collected on be thought of as "how is the trip the availability, frequency, and usage of accomplished" which includes a broader public transport services, travel conditions, consideration of the means employed. For expenditures, costs of services, and vehicle example, travel by “automobile” may be ownership. further broken down by "passenger in an automobile" and "auto driver". In In preparation of the Guatemala Poverty measuring travel costs, it is important to Assessment (GUAPA), World Bank, 2003, capture information per trip, and by mode. and extensive transport module was including in the 2000 Guatemala LSMS. 6 Travel time and travel distance are With this data, a transport and poverty measures of "transport difficulty" or profile was generated and included as a "transport effort". There is not a clear core part of the study. The results were consensus in the literature which measure then included as a fundamental part of the is most accurately collected in a household poverty reduction strategy. The specific survey. There is reference to the problems objectives of the transport-poverty profile of accuracy of recall or estimation varying were: DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 8 “Measuring the extent to which access to Recommending a set of policy basic services is already available in instruments for the Government of Guatemala. This includes obtaining a Guatemala for increased provision.” geographical, demographic and spatial Among the main issues identified in the overview of the extent of physical access study were that road quality was found to to critical services or ‘merit goods’ such as be a critical bottleneck for economic health and education; progress in Guatemala, impacting access to markets, employment and merit Determining the linkages between services (Table 4). The complex provision of transport infrastructure and topography and inadequate road network services and poverty, vulnerability and contributed to the geographic isolation of exclusion; the rural poor. This analysis was subsequently fed into the design of a rural Measuring the extent to which physical roads project which addressed the major basic access is a limitation in getting findings of the analysis. access to services in Guatemala; Table 4. Percentage of households with access to roads, by type of road. Quintiles Surfaced Unsurfaced Non- Paths Overall but motorable (non- motorable track motorable) 1 (poorest) 21 70 43 86 22 2 30 72 44 78 16 3 37 74 40 61 11 4 48 71 34 52 9 5 (richest) 61 62 29 38 9 Source: World Bank, Guatemala Poverty Assessment, 2003 4.3 Monitoring and Evaluation often not collected with sufficient frequency, and may not be at a level With the increased focus on measuring and sufficiently dissagregated to allow for monitoring results, there is a need to capturing changes specifically related to a develop better approaches to monitoring project. Some work in transport has been and evaluating transport interventions. done on evaluating project impact, with Household surveys can provide the basis most focusing on the impacts of rural road for specific kinds of indicators, as well as construction. the basis for evaluating the impact a project may have on beneficiaries. Impact evaluation relies on household and community survey data, and measures An extensive list of performance measures impact looking at direct and indirect in transport has been developed by the effects. The direct effects relate to changes World Bank and is attached in Annex 2. Of in travel time (to work, schools, etc.), this list, several can be captured using savings in fuel and other direct transport household surveys. The key is to ensure costs, and changes in seasonal or other that these data are collected on a periodic weather related road closures. Indirect basis to measure changes over time. effects consist of increases in income and other dimensions of well being such as Monitoring at the project level is difficult health, education social interaction and with household surveys because data are political participation brought about by the DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 9 infrastructure. Carrying out this work include the lack of income generating requires a carefully designed sample to opportunities outside of agriculture and account for ‘treatment’ groups which inadequate access to economic and social received the intervention, and ‘control’ services. In many cases the character and groups which did not. Methods are extent of these problems are largely a discussed in Grootaert, “Socioeconomic function of the inadequate provision of Impact of Rural Roads: Methodology and rural travel and transport infrastructure Questionnaires”, 2002 and in Baker, and services, especially at the village and ‘Evaluating the Impact of Development community level. 8 Projects on Poverty,’ 2000. 7 Among the examples are studies in Peru, Vietnam, Survey research is used to understand the Morocco, Brazil and Tanzania. Impact time and effort spent on transport in the analysis for transport can present context of overall household labor substantial methodological challenges and allocation and of the outputs achieved from thus careful design in designing such work the inputs to transport. This can mean is necessary. accessing supplies and getting goods to market in a timely and cost-efficient 4.4 Rural Transport Problems, Lack manner. Research from Azerbaijan Of Basic Access And Mobility examines local level rural transport as a factor and constraint in agricultural Rural welfare is strongly influenced by development and in the use of essential transport. Key features of rural poverty services (Table 5). Table 5. Percentage of households indicating they would start or expand production if roads were to be improved, by type of agricultural activity. HH Status Corn Potato Livestock Poultry Overall Very poor 93 87 88 43 81 Less poor 71 76 59 62 67 Non poor 71 57 70 66 62 Source: Poverty Assessment and Poverty Alleviation in Azerbaijan's Central Corridor, 2001. 4.5 Urban Transport Problems, distance, and travel mode, in an attempt Symptoms Versus Cause Of to minimize the social exclusion associated Poverty with low earning potential. The "income poor" may in fact have chosen poorly Urban poverty may be reduced through the served peripheral locations precisely contribution which transport makes to because they are the places where their increasing the efficiency of the urban overall welfare (in terms of availability of economy and thus to overall economic shelter, access to activities, etc) is best growth. Urban transport policies can also served. High transport cost is then a be focused more specifically on meeting symptom of their poverty rather than its the needs of the poor particularly with fundamental cause. regard to helping them access jobs, ser- vices and social networks. Sprawling urban Household survey information is needed to development can make the journey to understand this complex set of work excessively long and costly, particu- relationships and activities. Related to this, larly for the poor who often cannot afford specific aspects of individual and household the cost of living close to the central busi- behavior in urban areas that have drawn ness district where work may be available. attention include: wide variety of trip rates (trips per capita per time unit) among the The urban poor exhibit a complex trade-off poor, wide variation of transport between residential location, travel consumption/expenditure in the household DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 10 budget (can be zero if walking - to very The type of sample design generally expensive (25-30% of income) with the recommended for LSMS surveys (a upper bound income-constrained), wide nationally representative random sample of variation in travel time to work. There is a households drawn in two stages) is need to understand the trade-off between appropriate for transport applications. In housing cost and transport cost. Finally as policy related work a whole national overall incomes increase there appears to sample may be the most appropriate. In be a two stage transformation as project related work a similar sampling households progress first from non- design may be used, but focused motorized to motorized public transport geographically on only the project area and and then from motorized public transport a suitable non-project control area. A to motorized private transport. major difficulty in sampling for LSMS-type surveys is having an adequate sample A general constraint with most household frame (comprehensive, up to date, and surveys with relation to urban areas is the relatively error free) from which to draw sample size. Data is usually available at a the sample. The sample frame is a listing city level or for all urban areas, though of "all" households from which the cannot be disaggregated to analyze households to be surveyed are randomly differences within cities. Physical transport selected. The sample frame is usually improvements (infrastructure, services divided (stratified) into two or more strata, provided) can be focused on where poor such as rural and urban. Sampling beyond people live and work, with the caveat that the basic requirement for overall statistical improvements can also lead to rising land reliability (oversampling) may be done to prices and the exclusion of poor residents. ensure specific population groups are The household characteristics derived from adequately covered (e.g regions, ethnic the survey can be used with more groups, etc.). Countries with a national comprehensive statistical sources, such as census within the last three to five years a census, to allow for better poverty should make use of the census mapping of a city. Censuses themselves enumeration for the sample frame. lack the necessary household consumption Otherwise resources may have to be information needed to estimate a welfare allocated to generating a sample frame aggregate. Care must be taken when from other sources. merging the benefits of multi-topic household surveys and geographically The sample design must be statistically comprehensive census data to prevent rigorous and specified in the survey errors arising from over-generalization of documentation with sufficient clarity and the results. detail that the same approach can be used time after time to draw a sample with the 5 SURVEY DESIGN ISSUES same statistical characteristics. A robust and well documented sample design is There are a number of issues specific to needed so that post-project impact the incorporation of transport questions in evaluations can later use the same design the design of household surveys. Among and so be statistically comparable. them are sampling, survey administration, travel diaries, and the use of GIS, all Sample design must match the needs of discussed below. the survey. In a country-wide survey such as the LSMS, this is designed to produce a 5.1 Sampling nationally representative sample. In the case of a transport project the sample Sample design is critical for the usefulness design must provide for an adequate "with- of the survey: if this is done wrong, all project" and an adequate "without-project" other work to ensure data quality is (or control group) sample. The two wasted. Sampling is a very technical samples should be comparable in socio- discipline and unrecoverable errors can economic characteristics so that a post- easily be made. Proposals for sample project evaluation can be done with a designs made by client ministries or minimum complexity due to changes in consultants should be subject to competent other factors. review. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 11 5.2 Survey Administration what an individual does for a given length of time. Three types of diaries are The LSMS-type surveys are based on face- commonly found: (i) travel diaries where to-face personal interviews that take place only trips are recorded, (ii) activity diaries in the household over several visits. In where mostly out-of-home activities are most developing countries this is generally recorded, and (iii) time-use diaries where the only feasible method (rather than mail- most of what is done throughout the day is back or telephone based approaches). recorded. There is a well developed set of approaches from the LSMS work used to Completing a diary is much more facilitate data collection which is also demanding on the respondent and may applicable for transport questions. This involve higher non-response rates. Diaries includes the design and physical layout of are more difficult to comprehend and the questionnaire itself, the organization of require more effort in terms of preparation, fieldwork, pre-test, interviewer training, pre-testing, and administration. They are supervision while in the field, data coding, also labor intensive in terms of data and data cleaning. Because field problems coding. Diaries are have been criticised on can arise at any time, it is critical to ensure the grounds of panel conditioning, in that that the interviewers and supervisors are as respondents are required to keep log- well trained. A key reference is, "A Manual books of their behaviour, they are made for Planning and Implementing the Living aware of their habits which subsequently Standards Measurement Survey", Margaret may result in a change in those habits. Grosh and Juan Munoz, LSMS Working Paper No. 126, The World Bank, 1996. Despite these costs diaries have become more frequently used in travel research as In a recent survey of Urban Transport and new forecasting models have focused on the Poor in Mumbai, the World Bank used activities as opposed to trips. The activity- Handheld PDAs to collect the data. 9 Data based approach can consider the entry software was designed with the substitution of non-travel alternatives, questions and coded responses. better use of non-motorized modes, and Interviewers then recorded the responses trip-chaining. The focus on activities should directly into the PDAs. This allowed for a also permit the examination of high degree of accuracy in the data interpersonal dependencies among collection as error messages were built into household members, time constraints the software to account for inappropriate related to hours of operation of activities, or non-coded responses. The data were work schedules, expected activity duration, then downloaded every other day and and multi-day scheduling of activities. To checked for accuracy. A GPS unit was also develop such understanding the diaries attached to the PDA to mark the exact need to be collected from all household location of the household. members over several days. 5.3 Travel Diaries The paper, "Effects of Data Collection Methods in Travel and Activity Research", Conventional travel questionnaires may Ettema, Timmermans, and Veghel, collect data on only a portion of individual European Institute of Retailing and mobility (work trips, school trips, etc.). Services Studies, 1996 provides a good Questions related to travel may not be discussion of the trade-offs between recalled accurately, private motorized trips conventional travel surveys and travel often underestimate the time taken, public diaries. Among the rules-of-thumb they transport trips often overestimate travel develop are (everything else being equal): time. Also infrequent trips, small unimportant trips, and locations close to • If the aim of the study is to describe home and work may be overlooked. To activity patterns then a conventional some extent these problems can be survey questionnaire is sufficient. If reduced by specifying actual travel from a the aim is to examine the various recently preceding day rather than "usual" inter-dependencies between travel and travel. Another solution is a travel diary, activities then diary data should be which provides detailed information about collected. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 12 • Incentives, surveillance in the form of With the advent of the Global Positioning an appointment to pick up the diary or System (GPS) adding a geocode to provide repeated visits, and user-friendly the location of a surveyed household is design are essential to successfully inexpensive and very beneficial for collect diary data. analysis. During the conduct of a Community Questionnaire the geocodes of • Use a maximum of a 48 hour diary community facilities such as schools, unless the application (modelling) clinics, markets, bus terminals, village requires a longer frequency. "gateway" locations, etc. should also be • Use a designated day diary (to allow routinely collected. For rural households every day of the week to be specified). where barns and other outbuildings may be some distance from the main household • Use a 24 hour recall period for an these other locations should also be activity diary and a maximum of a 48 collected. In order to establish the location hour recall period for a trip diary. with a better degree of precision a number • Use an open interval diary over a fixed of GPS readings (3 to 5) should be taken at interval diary. the same location at different times and the results averaged. In household surveys • Include questions about timing, with more than one visit to the household duration, location, and mode. this could be done once per visit. The use Dependent on the aim of the study of GPS can present challenges in include questions about travel/ implementation – the technology does not activities companions, motivations, and always work in urban areas with tall planned/unplanned nature of activity. buildings, mountainous areas, or areas • Be sure the sample is spatially with heavy tree cover. representative. In addition to locations where the geocode 5.4 Use of GIS can be collected using a hand-held GPS receiver there are locations of interest A Geographic Information System (GIS) is which may not be visited in the field a software application that combines a survey. For example, a place of work or a database of statistical information with a market in another community not covered database of geographic co-ordinates so in the sample. Attaching a geocode to the that the statistical information can be household record for these other locations further processed in spatial ways and can requires either a database of street be viewed through various types spatial addresses or reference points displays, including maps. In order to use a ("monuments" such as a school, clinic, bus GIS the statistical information must have station etc.) linked to the geographic attached to each record a "geocode" which coordinate system or digital maps on which could be, for example, the location of the locations can be manually coded. These statistical record in terms of latitude, types of databases or maps may not longitude, and altitude, or its UTM always be available in developing (Universal Transverse Mercator) co- countries. Survey resources need to be ordinates. allocated for geocoding important non- household locations for the usefulness of Geocoded data is particularly important in GIS approaches is to be realized. transport applications. Geocoding of travel survey results is now commonly done in 6 INTEGRATING TRANSPORT developed country household travel QUESTIONS INTO HOUSEHOLD surveys. Geocoding allows the use of any SURVEYS or various zone systems to aggregate the data for further analysis or display. It As discussed above, it is possible to permits the calculation spatial statistics incorporate specific questions into existing that cannot be calculated otherwise and it core modules, or include a separate allows additional data checking routines to module to get at more detailed issues, be performed on the survey data so that depending on the country circumstances. inconsistent data are removed. Table 6 summarizes the core modules typically found in multitopic household DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 13 surveys, and what kinds of data can be survey. Qualitative approaches include collected related to transport in both rural techniques such as focus groups and open and urban areas. Specific questions are ended interviews generally use small included in Annex I for (a) household sample sizes. This includes subjective data surveys, (b) community surveys, and (c) on the attitudes and reasons which travel diaries. underlie travel behavior and public transport use. Ideally, a study approach In addition to the household survey, would include both a statistically qualitative approaches can also provide representative household survey, and a important information from respondents smaller number of qualitative surveys from that may not come across in a coded the same sample. Table 6. Transport information that can be obtained through household surveys by standard modules. Standard Modules Specific information for transport Recall Period Unit of data sector analysis collection Metadata on household • Geographic location, sample frame Current Household • Composition of household (demographic Household Roster breakdown) Current Individual Consumption Food expenditure • n.a. Last 7 days, Non-Food • All transport related expenditures should 30 days, 1 expenditures be detailed: year bus services, taxis, boats, other public Household transport modes, tolls, gasoline for motor vehicle); vehicle (car, motorcycle, bicycle) parts and repairs (tubes, tires, spare parts), vehicle insurance, vehicle registration, drivers licenses, parking fees) • Can include travel to collect drinking water, firewood, laundry, bathing (or in water module) (by mode, time, Housing Current Household distance, frequency, cost). • Basic information on housing relevant for location models. • Should include ownership of all vehicles (car, bus, truck, motorcycle, bicycle, Durable goods boat, pushcart, etc.). Also should 1 year Household include year of purchase, amount paid, estimated value today if sold. Other Assets • Animals for transport 1 year Household • Travel to economic activity (by mode, Labor income time, distance, frequency, cost ) Last 7 days Individual (primary and secondary employment) Transfers and other non • Transport services in-kind, subsidized Last 7 days, Individual labor income services 30 days DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 14 Table 6. Transport information that can be obtained through household surveys by standard modules. Standard Modules Specific information for transport Recall Period Unit of data sector analysis collection Current, Savings and Credit • n.a. Individual 1 year Prices • Average prices of transport Current Community Agriculture Agro-pastoral • Transport of agricultural inputs or Last 7 days, activities products to markets, processing, etc.) 30 days, 1 (ask mode, time, distance, frequency, year cost) Household • Collection of firewood, water (also could Last 7 days Non-farm self be collected under housing) employment • Travel/transport use Last 7 days • For those in school, travel to schools Education (travel mode, time, distance, frequency, Current Individual cost ) • For those using health services, travel to Health clinics, hospitals, health centers, (travel Last 30 days Individual mode, time, distance, frequency, cost ) • For those using social services such as nutrition programs, workfare programs, Last 7 days, Social Services Individual etc. travel to benefit, (travel mode, 30 days time, distance, frequency, cost ) Participation in • Social visits, festivals, ceremonies, Last 7 days, community affairs (non- religious events, (travel mode, time, 30 days, 1 Individual core) distance, frequency, cost) year • Physical access: Foot path, unpaved road, paved road, highway, river, lake, train, airport. • Access of other services: transport services, schools, health clinics, post, Community markets, church etc. (and travel to Infrastructure and Current and Community these services) Transport Seasonal • Affordability: Standard costs of services • Quality/barriers to mobility: -Condition of roads, closures due to flooding, reliability of public and private transport services, security, comfort DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 15 REFERENCES Barwell, I. 1996. Transport and the Village, Findings from African Village-level Travel and Transport Surveys and Related Studies. SSATP Working Paper No. 23. World Bank, Washington, DC. Banjo, G. and S. Robinson. Developing Rural Transport Policies and Strategies. (In progress) Deaton, A., and S. Zaidi. 2002. Guidelines for Constructing Consumption Aggregates for Welfare Analysis. LSMS Working Paper No. 135. World Bank, Washington, DC. Ettema, D., H. Timmermans, and L. Veghel. 1996. Effects of Data Collection Methods in Travel and Activity Research. European Institute of Retailing and Services Studies. Grosh, M. and P. Glewwe. 2000. Designing Household Survey Questionnaires for Developing Countries. World Bank, Washington, DC. Grosh, M. and J. Munoz. 1996. A Manual for Planning and Implementing the Living Standards Measurement Survey. LSMS Working Paper No. 126. World Bank, Washington, DC. Kolev, A. and R. Nitti. 2001. Constructing Multi-Sectoral Monitoring and Evaluation Indicators for Poverty-focused Interventions: The Case of Infrastructure. World Bank, Washington, DC. Lanjouw, P., J. Olson, and C. Elbers. 2002. Micro-Level Estimation of Welfare. World Bank Policy Research Working Paper No. 2911. World Bank, Washington, DC. Starkey, P. 2001. Local Transport Solutions - People, Paradoxes and Progress. SSATP Working paper No. 56. World Bank, Washington, DC. World Bank. 2002. Cities on the Move. A World Bank Urban Transport Strategy Review. ISBN 0-8213-5148-6. World Bank, Washington, DC. World Bank. 2001. Sourcebook for Poverty Reduction Strategies Vols 1 and 2. ISBN 0-8213- 4978-3. World Bank, Washington, DC. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 16 ANNEX I: SAMPLE QUESTIONS ON ACCESS, MOBILITY, AFFORDABILITY AND QUALITY A. Household Survey Basic Access Questions (which can be included in sector modules - education, health, employment, etc.): 1. How far from your home is […] (depending on module: place where you work, school, market (for selling produce, or shopping), nearest health clinic, nearest all weather road) ? Fill in ______ (distance)____(code) Distance Codes (meters 1, kilometers 2, ). Or precoded responses: 1. 0-1 km 2. 1-3 km 3. 3-5 km 4. 5-10 km 5. 10-20 km 6. >20 km 2. How long does it take you to travel from your home to [….] (destination above depending on module, e.g. school, work, etc.) ? ____ hours ____ minutes (total time one way) Or precoded responses (which may need to be adapted for rural/urban distances): 1. < 5 minutes 2. 5-10 minutes 3. 10-20 minutes 4. 20-30 minutes 5. 30-60 minutes 6. > 60 minutes 3. What modes of transport do you use to go to [….] (work, school, sell goods, shop) ? Primary mode ________ Secondary mode ________ Country specific (select applicable, can be adapted for country circumstances) 1. Public Bus 9. Bicycle 2. Private Bus, Van 10. Ferry 3. Own private car/truck 11. Boat 4. Other private car/truck 12. Train 5. School bus 13. Subway 6. Taxi 14. Walking 7. Motorbike 15. Ox/Horse Cart 8. Auto-rickshaw 16. Other (specify) 4. How much do you pay for this transportation ? Total Cost _____ One way trip (if more than one mode include total cost) DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 17 Consumption Module (transport expenditure data should be included in consumption module) Under the standard non-food consumption module, information on the following transport expenditures should be collected for the past 7, 30 days and the past 12 months (depending on item): 1. How much have you spent during the past __ days on: a. Past 7 days: 1. Fares for public transport within the community/town (specify by mode) 2. Taxis 3. Parking 4. Gasoline, Petrol, or Diesel b. Past 30 days and past 12 months: 1. Inter-city transport (specify by mode, train, bus, air) 2. International travel (transport costs only) 3. Purchases of Bicycle 4. Routine bicycle maintenance, parts 5. Purchase of Car / Truck/ Motorcycle (should be asked separately) (include car payments, etc.) 6. Routine vehicle maintenance, parts (do NOT include gasoline) 7. Motor vehicle repair after vehicle accident 8. Motor vehicle registration/ license 9. Insurance (auto, other motor vehicle) 2. Under the standard Ownership of Durable Goods include the following transport items: Bicycle Motorcycle/scooter Car Van Truck Cargo trailer for vehicle Boat Specific Transport Services (might be added to a separate transport module) 1. How far is the nearest _____ (bus stop, train station, subway station, etc.) from your home? 1. Less than one kilometer 2. Between one and two kilometers 3. Between two and four kilometers 4. More than four kilometers 5. Don’t know [DO NOT READ.] DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 18 2. How long would it take a person to walk to this […] (bus stop, train station, subway station, etc.) from your home? 1. Less than 5 minutes 2. 5-10 minutes 3. 10-20 minutes 4. 20-30 minutes 5. More than 30 minutes 6. Don’t Know [DO NOT READ] 3. How reliable is the [….] (bus, train, subway, etc.) service? Is it …. (Read list below to respondent.) 1. Very reliable 2. Somewhat reliable 3. Unreliable 4. Don’t Know [DO NOT READ] 4. How do you feel about the amount of crowding on [….] (buses, trains, etc.)? (Read list below to respondent.) 1. Satisfied 2. Neutral (neither satisfied nor dissatisfied) 3. Dissatisfied 5. How safe do you feel when riding [….] (buses, trains, etc.)? (Read list below to respondent.) 1. Very safe 2. Neutral (neither very safe nor very unsafe) 3. Unsafe 6. How convenient are the [….] (bus, train, etc.) routes for you; do they go to the places you wish to go? (Read list below to respondent.) 1. Very convenient 2. Neither convenient nor inconvenient 3. Not very convenient 7. How frequently do the [….] (bus, train, etc.) service come ? ___ times per hour ___ times per day 8. How satisfied are you with the frequency of service ? (Read list to respondent) 1. Satisfied 2. Neutral (neither satisfied nor dissatisfied) 3. Dissatisfied B. Community Survey The prototype survey below can be asked at the community level, typically from a community leader or knowledgeable person and will provide substantial information on transport issues. This information can then be merged with household level data for distributional analysis. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 19 Now I would like to know about transportation in this community to places that community residents sometimes use, such as bus terminals,markets and post offices. 1. 2. 3* 4* 5. 6. 7. Is the […] located within or outside the boundaries of the Approximately Is it possible How often is public What is the most How much does it How long does it community? how far is the to travel from transportation common mode of cost to travel to take to travel […] from the the community available for travel transportation to the […] from the from the center of the center to the to the […]? travel from the community center community community, in […] using community center to using this mode of center to the […] FIRST ASK QUESTION 1 FOR EACH LINE, THEN kilometers? public HOUR.......2 the [...]? transportation? using this mode ASK QUESTIONS 2-7 FOR EACH LINE BEFORE transportation? DAY........3 of GOING TO THE NEXT LINE. WEEK.......4 transportation? (Note if Question 2 is less than 1 km then skip to FORTNIGHT..5 next line) MONTH......6 INSERT CODE QUARTER....7 FOR MODE OF IF < 1 KM, HALF-YEAR..8 TRAVEL MOST GO TO YEAR.......9 COMMONLY NEXT LINE USED YES..1 WITHIN...1 NO...2 TIME OUTSIDE..2 KILOMETERS (»5) TIMES UNIT HOURS MINUTES 1 Nearest supply of drinking water 2 Nearest supply of cooking fuel 3 Nearest supply of heating fuel 4 Nearest primary school 5 Nearest scondary school 6 Nearest post-secondary school 7 Nearest health clinic 8 Nearest hospital 9 Nearest pharmacy 10 Nearest intercity bus terminal 11 Nearest passenger railway station 12 Nearest railway freight depot or siding 13 Nearest daily market 14 Nearest periodic market 15 Nearest cinema, theatre 16 Nearest library 17 Nearest place to use a telephone 18 Nearest post office 19 Nearest bank 20 Nearest car rental 21 Nearest tractor rental 22 Nearest fertilizer depot 23 Nearest improved Nearest all road weather road 24 Nearest Regional / District Road 25 Nearest National / Trunk Road Distance from the center of the community to 26 the farthest edge point of the community 27 Nearest community of the same size 28 NAME OF COMMUNITY: __________________________ 29 Nearest community of at least twice the size 30 NAME OF COMMUNITY: __________________________ 31 Administrative Capital (Level 1, National Capital) 32 NAME OF ADMIN CAPITAL __________________________ 33 Administrative Capital (Level 2, Regional or District Capital) DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 20 C. Trip Diary The travel diary can be left and completed by the respondent, or can be recorded by the interviewer. It is important to ensure a random selection of days. For each member of the household NAME: _________________________________ ID: ___________ HH ID: ____________ TRAVEL DIARY DATE ________________ At the beginning of my travel day (4:00 a.m.) I was at HOME or SOMEPLACE ELSE _________________________________________ (Circle HOME or write the name of someplace else in the space above) TRIP 1 Where did you go next? _______ Purpose of this trip? _______ 1. Work 2. Shopping 3. School 4. Social Visit 5. Entertainment 6. Doctor / Hospital 7. Personal Business [This could be Bank, Paying Bills etc.] 8. Return to Home 9. Other ____________ [Non Blank if selected] Where was this located? _________________________________________________________ (Street Address, Neighbourhood, Code (PIN, Postal or other) What time did you leave? ________ AM / PM Arrive? ___________AM / PM How You Went Waiting Time Travel Time Cost (Rs.) (Bus, Train, etc.) Mode A Mode B Mode C DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 21 Codes for Modes: 1. On foot 2. By bicycle 3. By train 4. By bus 5. By auto-rickshaw 6. By taxi 7. By two-wheeler (own vehicle) 8. By own car/jeep/van 9. In someone else’s car/jeep/van 10. Other _________ (Non Blank) Did you use a bus pass for this trip? YES / NO (Circle the correct one) If so, Duration of Pass: ________ and Cost of Pass (Rs.): _____ Did you use a train pass for this trip? YES / NO (Circle the correct one) If so, Duration of Pass: __________ and Cost of Pass (Rs.): _____ TRIP 2 Repeat pattern for each additional trip loop for the entire day. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 22 ANNEX II : WELFARE AGGREGATES The following is an excerpt from Chapter 2 "Theory of the Measurement of Welfare", in "Guidelines for Constructing Consumption Aggregates for Welfare Analysis, LSMS Working Paper No. 135, May 2002. Box 1. Summary of Theoretical Issues and Recommendations Issue Recommendations Money Metric Utility (MMU) vs. Welfare Ratio (WR) MMU is the amount required to sustain a level of living and requires that consumption Attempt should be made to use Money be adjusted by a Paasche price index that reflects the prices the household faces and Metric Utility and to calculate the whose weights are different for each household. Paasche price indices with individual household weights. WR is an indication of how much better or worse off a household is than a reference household (usually at the poverty line) and requires consumption to be adjusted by a Laspeyres price index that reflects the prices faced by the reference household but whose weights are the same for all households. The use of MMU can cause difficulties in analyzing the impact of redistributive policy but, on the other hand, WR does not necessarily represent welfare correctly. The latter is the more serious drawback in practice. Income vs. Consumption Consumption is a theoretically more satisfactory measure of well-being. In most developing countries where LSMS and / or household expenditure Income is used in industrial countries where self-employment is relatively rare so that surveys are available, consumption is the most household income comes from a few sources, where annual income variation is appropriate measure to use. low, and consumption data are relatively costly to gather. Consumption is less variable over the period of a year, much more stable than income in agricultural economies and makes it more reasonable to extrapolate from two weeks to a year for a survey household. When self-employment is common, income data is at least as expensive and as difficult to collect as are consumption data. Durable Goods and Housing A measure of use-value, not purchase, of durable goods is the right measure to include Exclude expenditure s- instead calculate in the consumption aggregate from a welfare point of view. a rental equivalent / user cost for housing and durable goods owned by the household. Time and Leisure Households with more leisure time have a higher level of welfare than households with Omit time and leisure in the calculation of no leisure. However, valuing leisure for each individual is problematic. Furthermore, it is consumption. difficult to distinguish between leisure, non-market work for the household, and involuntary unemployment. Public Goods Clearly presence of public goods such as hospitals and schools improves the welfare of Do not include any valuation of public nearby households more than that of households without good access to these goods in the calculation of the household services. However, estimating the value of those services is problematic. Households consumption aggregate. may choose private services even if public services are available. Contingent valuation of services that don’t exist are sometimes but of questionable accuracy; Farm Households It is possible to consider household as consumers separately from household Treat the farm household as a business businesses or farms in economies with active markets. In subsistence economies, this selling to the household. Attempt to assumption is sometimes hard to justify; however, trying to separate the producer from value produce at “farmgate” rather than the consumer using estimates farm-date prices is the best strategy in practice. In “market” prices. countries where a large fraction of consumption comes from home production and markets are less active, the evaluation of welfare becomes sensitive to difficult decisions about imputations, and should be regarded with caution. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 23 Issue Recommendations Differences in Tastes Expenditure on regrettable necessities should, in theory, be excluded but in practice it is Include expenditure on items that may or impossible reliably to distinguish between necessities and choices. Household size may not be regrettable necessities. however, is important and affects the household welfare associated with a given level of Adjust household expenditure to reflect expenditure. household size. DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 24 ANNEX III: LIST OF TRANSPORT INDICATORS AND USE OF HOUSEHOLD SURVEY DATA 10 Use of Dimension Mode Indicator Household Survey ? ACCESS Access to all-season road by rural population (% of total rural Roads * population) Average distance to nearest transport stop for urban population Roads * (km) Average distance to nearest transport stop for rural population Roads * (km) Roads Road Density in terms of population (km/1,000 people) Roads Road Density in terms of land area (km/1,000 km2) Rail Rail Lines Density in terms of land area (route-km/1,000 km2) Rail Lines Density in terms of population (route-km/1,000 Rail people) Motorized Road Vehicle Ownership in Rural Areas: Private Cars Roads * (% of rural households) Motorized Road Vehicle Ownership in Rural Areas: Motorcycles Roads * (% of rural households) Non-Motorized Road Vehicle Ownership in Rural Areas: Bicycles Roads * (% of rural households) Motorized Road Vehicle Ownership in Urban Areas: Private Cars Urban * (% of urban households) Motorized Road Vehicle Ownership in Urban Areas: Motorcycles Urban * (% of urban households) Non-Motorized Road Vehicle Ownership in Urban Areas: Bicycles Urban * (% of urban households) Non-Motorized Road Vehicle Ownership: Bicycles (% of Roads * households) Air Aircraft Departures (thousands) AFFORDABILITY Road Motor Vehicle Fuel Prices: Gasoline (Super/Regular) (US$/ liter) Road Motor Vehicle Fuel Prices: Gas/Diesel Oil (US$/liter) Spending on Transport Services by Urban Households (% of Urban * Urban Household Expenditure) Spending on Transport Services by Rural Households (% of Rural Rural * Household Expenditure) Rail Average Rail Tariff, Passenger (US$/passenger-km) Rail Average Rail Tariff, Freight (US$/tonne-km) Roads Road User Charges as Share of Total Road Expenditure (%) Ports Port Handling Costs: containers (US$/TEU) Ports Port Handling Costs: containers (US$/ton) DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 25 Use of Dimension Mode Indicator Household Survey ? QUALITY (*Technical Dimension*) Roads Paved Roads (% of Total Road Network) Roads Roads in Fair/Good Condition (% of Total Road Network) Rail Rail Traffic Density (traffic units/km) Rail Route Length of Multi-tracked Rail Lines (% of total route-km) Rail Rail Service Frequency (passenger train-km/route-km) Fatalities in Road Motor Vehicle Accidents in terms of vehicles Roads (Fatalities/10,000 vehicles) Fatalities in Road Motor Vehicle Accidents in terms of population Roads (Fatalities/100,000 people) Urban Urban Transport Modes (% of work trips) * Ports Seaport Traffic: containers Ports Seaport Traffic: general cargo Rail Rail Share of Passenger Domestic Travel (%) Road Road Share of Passenger Domestic Travel (%) Inland and Coastal Shipping Share of Passenger Domestic Travel Water (%) Air Air Share of Passenger Domestic Travel (%) Rail Rail Share of Total Freight Domestic Carriage (%) Road Road Share of Total Freight Domestic Carriage (%) Inland and Coastal Shipping Share of Total Freight Domestic Water Carriage (%) Air Air Share of Total Freight Domestic Carriage (%) QUALITY (*Perception*) Average Total Time Traveling by Rural Households All * (minutes/days) Average Total Time Traveling by Urban Households All * (minutes/days) Urban Travel Time to Work in Main Cities (minutes/one-way work trip) * Commercial Perception of Services Delivered by Road Roads Department/ Public Works Rail Commercial Perception of Railway Services Air Commercial Perception of Air Transport Services Ports Commercial Perception of Port Facilities and Inland Waterways Ports Cargo Handling Services: Market Openness EFFICIENCY (*Cost*) Ports Shipping Costs (ratio) Rail Railway Employee Productivity (Annual Output/Employee) EFFICIENCY (*Economic*) Roads Road Transport System Technical Efficiency (US$/km) FISCAL COST DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 26 Roads Road Expenditure as share of GDP (%) Roads External Funds as Share of Total Road Expenditure (%) Roads Actual to Required Road Maintenance Expenditure (%) FINANCIAL AUTONOMY Roads Expenditure on Owning and Operating Vehicles (US$) INSTITUTIONAL DEVELOPMENT National Roads Boards (NRB) Exists and Reports (at least Roads annually) (Y/N) Roads Private Sector Representatives form majority of NRB (Y/N) Main (National) Road Agency operating with Annual Report Roads published (Y/N) Main (National) Road Agency publishing Technical and Financial Roads Audits (Y/N) Roads National Road Safety Action Plan (Y/N) Roads Social Assessment of Road Projects Mainstreamed (Y/N) All Gender assessment (Y/N) All Access for all (Y/N) All Planning (Y/N) Roads Environmental Assessment of Road Projects Mainstreamed (Y/N) Roads Communicable disease control (Y/N) Competitive Private Sector Participation in Transport Services All (Y/N) All Core labor standards (Y/N) All Health and safety (Y/N) DEVELOPMENT OF A TRANSPORT MODULE FOR MULTITOPIC HOUSEHOLD SURVEYS 27 ENDNOTES 1 See Grosh and Munoz, 1996, “A Manual for Planning and Implementing the Living Standards Measurement Study Survey”, LSMS WP 126; Scott, 2003, “Multi-topic Hosuehold Surveys” in The Impact of Economic Policies on Poverty and Income Distrbution, eds. Bourguignon and Pereira da Silva, and Scott, Steele and Temesgen, (forthcoming), “Living Measurement Study Surveys” in The Analysis of Operating Characteristics of Surveys in Developing and Transition Countries, ed. Ibrahim-Sorie Yansaneh, UN Statistical Deivision 2 See Grosh and Munoz, 1996; Grosh, A Guide to Living Standards Measuremetn Study Surveys and their Data Sets, LSMS WP 120; Scott, et. al., (forthcoming); and Grosh, M. and Glewwe, P., 2000, Designing Household Survey Questionnaires for Developing Countries. 3 See Scott, 2003 for a discussion on measuring welfare with Income and Expenditure Surveys. 4 See Chapter 1 "Poverty Measurement and Analysis" of the "Sourcebook for Poverty Reduction Strategies", 2001. 5 Ibid. 6 For the full report and background paper on the Transport and Poverty Profile see http://lnweb18.worldbank.org/external/lac/lac.nsf/Countries/Guatemala/90CF6C5A035DCEF58 5256CE90049E66C?OpenDocument (Jyotsna Puri, 2002) 7 These resources can be found at: http://www.worldbank.org/transport/ and http://www.worldbank.org/poverty/library/impact.htm 8 See Banjo and Robinson, “Developing Rural transport Policies and Strategies” (in progress) 9 See Baker, Basu, Cropper, Lall and Tekeuchi, “Urban Poverty and Transport; The Case of Mumbai,” 2005, draft. 10 Source: http://intranet.worldbank.org/WBSITE/INTRANET/SECTORS/INTTRANSPORT/INTTRM/0,,conte ntMDK:20283507~pagePK:210082~piPK:210098~theSitePK:514794,00.html