36208 Transport and Poverty in Guatemala: A Profile using Data from the ENCOVI 2000 Technical Paper No. 8 Jyotsna Puri* I. OBJECTIVES Transport services are widely viewed as critical to reducing poverty and promoting growth by providing access to basic services. An efficient and cheap transport system empowers people by providing them political and physical access. Fast and easy access to basic transport and communication also reduces vulnerability to natural and man-made disasters. This study examines the extent to which provision of transport and infrastructure services are correlated with poverty in Guatemala. By judging the relationship between roads and transport services, and poverty, it aims to identify geographical areas and demographic groups in Guatemala, which will most benefit from transportation-related interventions. A companion study quantifies the impact of transport related interventions on welfare and poverty. The objective of this study is to construct a detailed and informative poverty-access profile that can simultaneously inform Guatemala's poverty reduction and rural transport strategies. In this respect, the presence of an extensive household and community level dataset on transport that can be linked to a living standards survey at the household level­the ENCOVI 2000 (Encuesta Nacional de Condiciones de Vida or The Living Standard Measurement Survey, LSMS)­presents a rare opportunity. The ENCOVI 2000 was collected by Instituto Nacional de Estadistica (INE), Guatemala, and contains a special module on transport and road construction which highlights its importance in discussions of poverty in the country. Transport and road data in the ENCVOI 2000 was collected at both levels, the household level and at the Primary Sampling Unit (PSU) level. The structure of the data is discussed in detail below. The specific objectives of the transport-poverty profile are to examine several things: 1. Measuring the extent to which access to basic services is already available in Guatemala. This includes obtaining a geographical, demographic and spatial overview of the extent of physical access to critical services or `merit goods' such as health and education; 2. Determining the linkages between provision of transport infrastructure and services and poverty, vulnerability and exclusion; 3. Measuring the extent to which physical basic access is a limitation in getting access to services in Guatemala; 4. Recommending a set of policy instruments for the Government of Guatemala for increased provision of physical access to transport services. This study provides a detailed overview of the correlation between the provision of transport *Author was consultant with LCSFT during the preparation of this study. Comments are welcome and should be directed to jyotsna.puri@undp.org. The study benefited immeasurably from comments by Guillermo Ruan (LCSFT) and Kathy Lindert ( LCSHD). Valuable comments were also provided by Roberto Armijo (Consultant, LCSFT), Guillermo Duarte (BIOENERG, Guatemala) and discussion participants at SEGEPLAN (Secretaria de Planificacion y Programacion), Guatemala and DGC (General Directorate of Roads), Guatemala.. 1 infrastructure and services on one side, and access to other services such as health, education, markets and employment on the other. However as the QER (2001) (Quality Enhancement Review) cautions, one cannot infer causality from cross-tabs and correlation's. This is because roads are not randomly placed and it is highly likely that factors that have led to road placement, also impact outcomes of interest. Also, impacts of roads are primarily indirect and depend on interactions with other investments, availability of social and physical infrastructure, and, geographic, community and household characteristics. Finally, rarely do roads have only local impacts. It thus becomes difficult to measure the impact of roads and transport infrastructure and services, and past studies have expounded on this matter in detail (See for example, Howe et al. 1989). For analytical purposes this means that it is difficult to control for heterogeneity of factors that interact with roads to produce impacts associated with roads, since a lot of factors that could help us control heterogeneity, may themselves have been determined by past road investments and networks (QER, 2001). However a correlation analysis does help to describe and explore key questions of policy. An overview of the state of physical access that is constrained/enabled by transport infrastructure and services, can aid in narrowing policy options by focusing them at more vulnerable and less endowed sections of society. Correlation's are also helpful because they help to narrow the list of possible target areas/groups. In the context of the Bank, a study of this kind is useful because it provides an opportunity to establish a baseline for future evaluations of the Bank's project and other planned interventions and can aid in providing a best-practice prototype for poverty assessments and other transport work. II. BACKGROUND Guatemala is amongst the poorest countries in Latin America. Almost 16% of the population is extremely poor (i.e. has a per capital annual income of less than $243.6) and more than half the population is poor (has a per capita annual income less than $550.2) (Table 2). This implies that a large part of the population is dependent on publicly provided infrastructure and services. Since in Guatemala, the poor largely inhabit rural areas which are characterized by low levels of investment in public goods, a vicious cycle assails the poor: The poor have much poor access to public services and have lesser means available, to extricate them out of their poverty. Poor transportation infrastructure Poor access to markets health and and services education High poverty Figure 1: The Vicious Cycle 2 Correlation between poverty and absence of roads is also brought out vividly by overlaying a poverty map on a road map for the country. The poverty map (Map 1) shows the percentage of poor by municipio.1,2 The road map shows the classified network of paved roads and unpaved roads in Guatemala (surfaced roads).3 The map clearly shows that regions with high levels of poverty (Northern and North-western Guatemala) also have the lowest motorable road density in the country. Not incidentally, these regions have also been the most affected by the pre-treaty conflict and civil war. On the other hand, the metropolitan region where less than 1% of the population is extremely poor (and less than 20% are poor), is amongst the best provided regions of the country (along with the Central region). Poverty in Guatemala also has a distinct ethnic face. The indigenous population of Guatemala is on average poorer than the non-indigenous population. Amongst the indigenous groups, the Mam and Qeqchi are the poorest, while the Kaqchikel and the Kiche are the most well off. Not surprisingly access to transport infrastructure and services, and thus opportunity and vulnerability are also different across different ethnic groups, in tandem with their poverty characteristics. II.A. Transport Infrastructure and Services and Poverty Guatemala's road network is small and inadequate. Guatemala's total road network is relatively small equaling 26,000 km. Of this 14,000 km is `classified' while the rest is still in the process of being classified4. The classified road network comprises nearly 4,000 km of main and secondary roads (mostly paved) and 10,000 km of tertiary and rural roads (unpaved). Although Guatemala is a small country, this 5 network is inadequate. This is brought out vividly if we compare the size of road network in Guatemala, with that of Costa Rica­a country half the size of Guatemala. Costa Rica has a road network two and half times that of Guatemala (equal to 35,600 kilometers) although the population of Costa Rica is one-third the population of Guatemala. The road service index in Guatemala is one-tenth that of Costa Rica. This means that in Costa Rica, every 1000 people have 11 kilometers of road, while in Guatemala, they have little more than a kilometer. Till recently, road works in Guatemala have had an urban bias. Most road works, which include road rehabilitation, improvement and repair,have concentrated on the paved road network of Guatemala which predominantly connect prominent cities and towns. Consequently when its percentage of paved road network is compared with that of other countries in Latin America, Guatemala fares well (Table 1);6 In 1998, Guatemala ranked second (along with Panama) in the percentage of its road network which has been paved (Table 1). 7 One third of its road network is paved.8 However, since paved roads usually coincide with urban areas, the predilection of road interventions to occur in these areas implies that 1The poverty map was constructed by a multi-agency technical team consisting of representatives from SEGEPLAN-INE-URL with technical assistance provided by the World Bank. It combines data from the 1994 census and the 1998-99 household budget survey (ENIGFAM). It will be updated with data from the ENCOVI 2000. 2The road map only shows the `classified' network. 3The map does not show rural roads ­ which are essentially single person cleared tracks which may be traversed by humans and animals but not vehicles. 4One of the byproducts of the San Marcos Rural Roads Pilot Project financed under Loan 4260-GU, was the classification of roads, in which both DGC (General Directorate of Roads) and INFOM played a very important role. 5 These usually cover the paved roads and unpaved roads. Most classified roads have been constructed and are maintained by the DGC. Unclassified roads can also be paved or unpaved (more often the latter) and may be motorable but may not be all-season (may only be open in the dry season). Information about these "un-official" roads is scarce. Additionally there are dust roads and paths which are not motorable. The improvement of these tracks and paths is the goal of a new project being motivated via a pilot project on tracks and trails in San Marcos. 6Paved roads are roads that are surfaced with asphalt or concrete. They usually sustain high volumes of traffic and large passenger and freight loads. 7The absence of data on road and transport infrastructure has been decried elsewhere (WDR; Various Issues). Even the data that is present, is not strictly comparable across countries because the International Road Federation, which collects international data, obtains it from different sources which are country specific. In countries where National road associations do not exist or do not respond, other agencies such as Road Directorates, Ministries of Transport or Public Works or Central Statistical Offices are contacted. As a result, compiled data are of uneven quality. 8Data on other correlates such as the quantity of goods and passengers carried by the road network which may indicate the importance of the road network, is not available. 3 development will be pro-urban and pro-rich. In a country with the dubious honor of having one of the highest inequality indices in the world, this has served to increase the isolation perceived by rural areas. Guatemala's unpaved road network is of variable quality. Less than half (45%) of the country's classified unpaved network (which includes unpaved secondary, departmental and municipal roads) is in maintainable condition. (In contrast, 75% of the paved road network is in maintainable condition). Since rural households mostly have access to unpaved roads (other than footpaths and trails), the absence of year-round, all-weather roads means that their geographic and economic isolation is exacerbated. A country-wide assessment of the need for road works shows that the absence of all-weather roads, are responsible to a large extent, for the absence of a reliable public transport system in the rural areas. Poor road conditions lead to large repair and maintenance costs for transport operators, and poverty in these areas does not allow operators to transfer their costs to the consumers. The absence of a dependable road network also implies that people have to travel almost an hour to reach a medical center and more than an hour (one-way) to procure wood; Most households in San Marcos and Huehuetenango are more than an hour away from a marketplace, where they can buy and sell their produce. There is a dire need for curative road-related interventions in the country. Road closures caused by landslides and flooding, and, bad surface quality, especially assail unpaved roads. There is also evidence of a large number of communities in the north and the north-west that can only access the rest of the country via tracks and trails.9 There is also no cheap alternative to road transport. Both these facts underscore the need for two kinds of road-related interventions in Guatemala: Firstly, rehabilitating and regularly maintaining rural and unpaved roads in Guatemala; and secondly, improving and regularly maintaining tracks, trails and footbridges used by remote communities. These are also the focus of an on- going project in San Marcos and the aim of a follow-on project, soon to be implemented in Huehuetenango and other departments of the country. In Guatemala, transport policy must be directed at improving the quality of existing roads and providing cheap, easily available and good quality public transportation services. This recommendation is drawn from a number of facts: First, large parts of the country have limited accessibility (defined as not all-year-round and/or not all-weather) because of bad road quality and road closures. This makes it difficult to mobilize and provide basic services and provisions, especially to the poorer sections of the population. Second, Guatemala is a mountainous country characterized by large inter-city distances. In the absence of a rail network, there are few economic alternatives to road transport for most freight and passenger traffic. Presence of good quality all-weather roads and easily available public transport is critical for providing even basic amenities to many sections of the population. Third: More than half the population in Guatemala lives below the poverty line which means that a large part of the population is dependent on publicly provided transport infrastructure. This means that available means of transport should be cheap and affordable. II.B. Sectoral Issues The road and transport sector in Guatemala is characterized by the following: 1. Centralized decision making which has spawned several jurisdictional agencies at the national level leading to a variety of standards. 2. Poor road surface quality caused mainly due to lack of maintenance and improvement works. 9As part of ENCOVI 2000, 481 Primary Sampling Units (PSUs) were sampled, covering a total of 4,844 households, more than two-thirds of which were rural, and 57% of which were poor. 4 There are three types of roads in Guatemala depending on jurisdiction­National, departmental and rural. The DGC (Direccion General de Caminos) is responsible for constructing, maintaining and rehabilitating a large part of the classified road network which is 14,000 kms in length, and which consists mainly of either `national' roads or `departmental' roads. The DGC also directs and allocates funds to its regional offices (Zonas Viales) for road works. Rural roads are maintained variously by the community, DGC, finca owners or others (See Box 1).10 However these roads receive very little central government assistance for their upkeep and maintenance. A diverse set of agencies are responsible for road maintenance. The outcome of the multiple agency involvement is an array of ad hoc, often unsustainable interventions. The rural roads sub-sector especially (including the municipal roads), is characterized by a large number of financing and implementing agencies. This issue of multiple agency involvement in rural roads and its consequences was discussed during the Rural Transport Workshop supported by the Bank and held in Guatemala in May 2001. One of the conclusions of this discussion was that there is an urgent need to define the role of each agency and to coordinate their interventions. A great deal has been achieved though the ongoing Bank operation (Loan GU-4260) but the issue will be addressed more in depth during the development of the Rural Transport Strategy. The Rural Transport Strategy that is being prepared under the ongoing loan will contribute to the development of a more coherent set-up for the management and financing of rural roads which will also result in increased coordination among the International Financing Institutions, other agencies and the main stakeholders involved in rural roads. The problem of non-standard road works is worsened because most road works are sub-contracted to private contractors. Private contractors perform 100% of maintenance works on paved roads and 40% of maintenance work on unpaved/earth roads. A pilot project in the department of San Marcos, co-financed by the World Bank is attempting to overcome this shortcoming by delegating responsibility for the management of the municipal road network to an association of mayors (ADIMAM). Box1: Agencies responsible for Road Works in Guatemala, 2000 Agency Areas of Jurisdiction General Directorate of Roads (DGC) All Guatemala except for building rural roads in Peten. Social Investment Fund (FIS) All Guatemala National Fund for Peace (FONAPAZ) Alta Verapaz, Quiche, Huehuetenango and Baja Verapaz. Municipal Development Institute (INFOM) San Marcos (But soon to be expanded under the new project). Integral Development of Rural Communities (DICOR) San Marcos, Jutiapa, Chiquimula, Quiche, Santa Rosa, Jalapa, El Progreso and Huehuetenango. Integral Development Funds for Guatemala (FODIGUA) All Guatemala Others (NGOs, Coffee plantation owners, MAGA) Since 1998, the World Bank has provided several loans for maintaining and rehabilitating roads and also to build technical capacity. The ongoing Loan 4260-GU has assisted the Government of 10 An interesting anecdote is related by local experts: A recently constructed road which connects Guatemala to Honduras was commissioned at the beginning of 1995 to increase the mobility of goods and services. However a short stretch of the new road went through a banana plantation and the plantation owner would charge most freight vehicles a tariff for road use, despite the national legislation. This had managed to significantly pervert the free flow of traffic, underscoring the power that finca owners have. (The issue has since been resolved). 5 Guatemala in eliminating the rehabilitation backlog on the country's road network. A follow-on project which is in preparation, is aimed at devolving power to communities and involving them in road-related decision making by setting up associations of municipalities. These municipal associations will be aided financially and technically by the World Bank before they become self-sustaining. The in-preparation follow-on project also contains several pilot projects: One is aimed at setting up road maintenance micro- enterprises in the San Marcos department. Another pilot is directed at enabling non-motorized access via the improvement of tracks, trails and footbridges. The Bank has also assisted the Government of Guatemala in outsourcing road maintenance activities to the private sector and in the implementation of COVIAL (The National Road Fund) to secure timely, stable and sustainable funds for the maintenance of the rehabilitated network. Funding for road maintenance is provided from COVIAL. As part of the ongoing loan, the Bank is also providing COVIAL with technical assistance (TA) to improve its organizational capacity. COVIAL was created in 1996 and is financed by a fuel surcharge of Q.1 per gallon of fuel. Approximately US$70 million are collected from this surcharge (The World Bank, 2001). Before the peace accords were signed in 1996, the country witnessed very little road maintenance. It is believed that with road rehabilitation activities currently in progress and that are variously funded by the World Bank, the IDB and other multilateral agencies, the entire paved road network and the remaining part of the unpaved classified road network will be maintainable by the end of 2005. According to the ENCOVI 2000 survey, roads in Guatemala can be divided into four types depending on road conditions and materials used for road construction. These four types are: surfaced roads, unsurfaced roads or earth roads, dirt roads and paths (See Box below) (Figures 1-5). Surfaced roads are roads which have been surfaced with either asphalt, concrete or gravel. Unsurfaced roads are essentially earth roads that have not had any imported material used on their surface. Obviously their surface quality is not as good as that of surfaced roads and vehicles experience much greater wear and tear while plying on these. However, survey respondents to the ENCOVI were not always aware of the differences in definitions of various road types. For example, it is believed that all respondents of the ENCOVI 2000 may not have understood `earth roads' or unsurfaced roads to be different from unpaved surfaced roads. (Also, unpaved roads may `seem' like earth roads during the rainy season (Figure 3)). For this reason, in this study, we club surfaced roads and unsurfaced roads into one category and refer to them as "motorable roads" or roads on which motorized vehicles can ply. The third and the fourth type of roads namely, dirt roads and paths, are clubbed together as "non-motorable" roads, since these roads can at best be used by beasts or for pedestrian traffic. Box 2: Road types outlined in ENCOVI 2000, Guatemala. Type of Road Description in ENCOVI Questionnaire Description of Road Are asphalted/surfaced roads. Surfaced Roads Carretera pavimentada o balastrada Motorized transport can ply on these. Unsurfaced roads. Motorized Earth Roads Carretera de tierra o terraceria transport can ply but may be seasonal. Mainly for beasts of Dirt roads/Tracks Caminos de herradura sin balastre burden/horses. Footpaths. Mainly for Paths Veredas pedestrians. Source: The ENCOVI 2000 and communications with road engineers at the World Bank and the DGC, Guatemala. 6 II.C. Scheme of the Study Presented with this background, it is important to assess if road quality relatively poor transportation services have inhibited `access' to markets, health services and schools, contributing in turn to low growth and poor development. (Access is defined to be a function of availability and affordability). Data from 481 Primary Sampling Units (PSUs are clusters of households, grouped together for survey purposes) in the ENCOVI 2000 indicate that nearly 56 % of households for which PSU level information is available, had no access to surfaced roads while 13% had no access to motorable roads throughout the year. 48% of these households did not have any access to public transport.11 Household data also indicate that a majority of rural households have to walk distances of 14 kilometers to reach a market or a bank. Similarly, poor households in rural areas, have to walk nearly an hour to collect wood, their main source of energy for cooking.12 There is anecdotal data to show that poor road conditions have especially contributed to high agricultural production losses and low overall agricultural profitability in the country. It is estimated that there is a 100% wedge between farmgate prices and market prices in the department of San Marcos.13 Subsequent sections of this study explore the state of transport infrastructure and services in Guatemala and examine their correlation with access to essential services. In the next section, a brief description of the dataset used in the study is provided. This section is important because the ENCOVI 2000 dataset is a combination of two surveys taken at two different levels: the household level and the PSU level. In the ENCOVI 2000, PSUs were not selected randomly and are not representative of PSUs in Guatemala. Since most data on road and transport is contained in the PSU section of the ENCOVI 2000 survey, an analysis of access and poverty requires combining the data from two levels and examining the characteristics of households belonging to surveyed PSUs. This creates other analytical issues which are also discussed in the next Section (Section III). Section IV and V use data from the PSU portion of the ENCOVI 2000 dataset to examine access to road and transport infrastructure for households. Section VI uses data from the household portion of the ENCOVI 2000 dataset and examines access to public services such as water, health facilities, schools, wood. Section VII examines access to other opportunities. Section VIII concludes. III. DATA III.A. The ENCOVI 2000. The principal data source for this study is the Living Standards Measurement Survey (ENCOVI 2000) for Guatemala, collected by the National Statistical Institute (INE), Guatemala. The ENCOVI 2000 was specifically designed to include a unique level of data on transport due to the importance of this issue for poverty in Guatemala. Transport data was collected at two levels: at the household level and at the PSU level. The ENCOVI household survey covered 7,276 households and is statistically representative at the national level, at the urban/rural evel, for the eight regions and for the five largest ethnic groups. l However, as is usual with LSMS type surveys, the data is not representative at the departmental/municipal or community level. The `community' section of the ENCOVI questionnaire was administered in 481 PSUs. A description of the households which responded to the household part of the ENCOVI 2000 and belong to the PSUs that were surveyed, is presented in Table 3. Because a significant share of the information needed on transport 11These results must be read with caution. See discussion on data below. 12 World Bank calculations using the Living Standards Measurement Survey data (ENCOVI, 2000 and Instituto Nacional de Estadistica, Guatemala). These calculations use household level data, for the 7,276 households. 13The World Bank, 1997. 7 was gathered at the PSU level, several important points about the PSU questionnaire are worth mentioning at the outset. (a) The PSU. The PSU questionnaire was administered in clusters where the household survey was carried out. The PSU questionnaire thus gathers information for contiguous households and does not share any other common characteristics by design. (b) Administration of the PSU questionnaire. Each PSU questionnaire was administered to focus groups of individuals belonging to the PSU in question. These may or may not have included members of the actual households that were selected for the household survey. (c) Non-Response Rates. INE attempted to collect PSU information for all PSUs covered by the ENCOVI household survey. However this proved difficult, particularly in urban areas where boundaries between PSUs become blurred; there was a 40% non-response rate for ENCOVI PSUs. The rate of non-response was higher in urban areas. (d) Linking PSU information to ENCOVI households. Information collected in the PSU questionnaire can be linked to household-level data by virtue of geographic location codes entered in both datasets. PSU-level information is available for about two-thirds (4,873) of all households in the ENCOVI household survey. (e) Lack of Statistical Representativeness. Given the above definitions and sample methodology, the PSU dataset is not statistically representative of PSUs in Guatemala. Moreover, because of the non-response rate for the PSU questionnaire, households for which PSU-level information are available, are not representative for all households in Guatemala. (This is typical for all LSMS-type PSU "surveys"). For the study, since these PSUs are not representative but have data associated with them, they will be referred to as "ENCOVI PSUs," short for `PSUs that were surveyed in the ENCOVI 2000'. III.B. Other Data Sources. The study also draws on other sources of information such as data from the WDR (World Development Report) 2001 for international comparisons and a GIS dataset on roads constructed by the DGC. IV. TRANSPORT INFRASTRUCTURE This section uses PSU level data from the ENCOVI 2000 (ENCOVI PSUs) to examine characteristics of road infrastructure and transport services in Guatemala. The PSU level questionnaire of ENCOVI 2000, was administered to 481 PSUs of which 15 PSUs did not respond to transport questions. PSU level data can be linked to 4,844 households. Table 3 shows that PSU-level data is present for a higher percentage of rural and poor households than urban and non-poor households. Nonetheless, in urban areas, a higher percentage of non-poor households have PSU level data. Such information is important because PSU level data is not representative at the regional, ethnic or areal level. Knowledge of the distribution of households which can be linked to PSU-level data is important so that inferences made from PSU-level data may be appropriately weighted. IV.A Roads IV.A.1. A Description. Guatemala is a mountainous country with obvious physical obstacles for road construction. Map 2 overlays the elevation map for Guatemala on a map showing her road network. The map shows that the south-central portions of Guatemala are the most mountainous with elevations 8 going up to 4200 meters. This region also has the highest population density (Map 3). The considerable road presence in these regions of the country, hides two critical inadequacies: First, the poor and indigenous households suffer from lack of `access' to road infrastructure, and, road surface quality is a critical problem in Guatemala. (For the purposes of this study, `access' is defined as a function of two variables ­ affordability and availability of road infrastructure and services.). IV.A.2. Motorable Roads. In 2000, 13% of the ENCOVI PSU households had no access to a motorable road through the year (Table 4). However experts at the DGC (Direccion General de Caminos) and at BIOENERG, assess this percentage to be on the low side. A couple of reasons may account for this low figure. Firstly, these tables use PSU level data which is not representative for the whole country. Secondly, this figure does not account for seasonal non-access. As explained below, many areas in Guatemala suffer from road closure especially during the rainy season. This figure does not take this into consideration. Access to motorable roads is not significantly different between rural and urban areas, but poor and indigenous households belonging to ENCOVI PSUs have lesser access to motorable roads. Approximately 16% of these households did not have access to motorable roads. 23% of extremely poor ENCOVI PSU households had no access to motorable roads. IV.A.3. Other Road Types. The most prevalent road type in Guatemala are unsurfaced roads or earth roads. However as can be seen from Figures 1-4, the surface quality of these roads varies widely. Unsurfaced roads are frequently unpassable during the rainy season which lasts from May to October. On the other hand, surfaced roads (Figure 1) are by and large passable throughout the year. Surface quality is discussed in detail in the next section but it is important to see that, on average, households belonging to urban non-poor ENCOVI PSUs have relatively better access to surfaced roads: Households belonging to urban ENCOVI PSUs are more than two times as likely to have access to a surfaced road than households located in rural ENCOVI PSUs. Contrariwise, households belonging to rural ENCOVI PSUs have better access to unsurfaced roads. Regionally, households belonging to ENCOVI PSUs located in Metropolitan Guatemala were most favored with respect to access to surfaced roads: In 2000, more than two thirds of the households had access to surfaced roads. Compared to them, households belonging to ENCOVI PSUs located in the North-East were most disadvantaged: Less than one-fifths of the households had access to a surfaced roads. It is interesting to note that households belonging to ENCOVI PSUs in the North-East region, along with households belonging to ENCOVI PSUs in the North-West region also have least access to motorable roads. Amongst the ethnic groups, the Mam and the Qeqchi had least access to surfaced roads. A high correlation between access to good roads and, income and ethnicity suggests two things: First, that PSUs surveyed as part of the ENCOVI 2000 (ENCOVI PSUs) are relatively homogenous. Richer households tend to conglomerate together and poor households tend to conglomerate together (Table 3). Households with the same ethnic characteristics also tend to live near each other. Second: roads are not randomly placed: Richer and non-indigenous households seem to have better access to good quality roads than poorer and indigenous households. This in turn could indicate any of the following: The richer and non-indigenous households have a greater say in determining road construction and rehabilitation; and/or, that these households tend to locate near better quality existing roads; and/or, that the presence of roads has led to higher income and lower poverty. In this context, it is also interesting to see that households belonging to `communities' (surveyed in the ENCOVI 2000) that have existed for 20 years or more have better access to most road types than households that belong to communities that have existed for less than 20 years (Table 5). 9 IV.B. Road Quality Neglecting to recognize road quality as a limitation to access is a serious omission, especially in a country like Guatemala. This is because very few households are completely isolated. However this masks the fact that road quality is extremely variable in the country and there are periods of time during the year, when communities are temporarily cut off from the rest of the country. There are several definitions of a road ­ and these range from a surfaced road which guarantees all year vehicular access, to a footpath which can only be used by beasts of burden or pedestrians. Measuring access to a `road' in such an unqualified manner would lead one to mistakenly believe that Guatemala has an extensive road network and that the road service index is skewed favorably towards the rural areas: Rural areas have access to a greater variety of roads than urban areas. But this masks the fact that realistically year round access is skewed in favor of urban areas whereas rural areas suffer from lack of access. IV.B.1. State of Roads (1995-2000). Roads have to be continually repaired, rehabilitated and re-surfaced to maintain their surface quality. In the absence of maintenance works, they deteriorate rapidly and become non-passable. In Guatemala improvement works are especially critical because of the long rainy season, which washes away road surfaces. Discussions with road fund managers and engineers at the World Bank and at the DGC reveal that most road works were only initiated in the past 6-7 years. These road works have concentrated on rehabilitating and repairing (restoring or upgrading) existing roads. Despite these efforts during 1995-2000, less than half of the households belonged to ENCOVI PSUs that reported an improvement in road quality, while the rest reported either a deterioration or status quo in the quality of roads (Table 6). As is expected, households in urban PSUs benefited most from these efforts. Although perceptions about improvement in road quality over the 1995-2000 period do not vary significantly between the rich and the poor, they vary amongst regions: Households in Metropolitan and South-west ENCOVI PSUs most frequently reported an improvement in road quality while those in the North and Central region most frequently reported a deterioration. The most important reasons documented for the reported deterioration in road quality were road closures caused by land slides, flooding and mudslides and the appearance of holes and ditches rendering roads intransitable. IV.B.2. Road closure (1999-2000). Road closure is serious problem in Guatemala. Almost one-thirds of the surveyed households belonged to ENCOVI PSUs that reported road closure during 1999-2000 (Table 7). One-fifth of these households belonged to ENCOVI PSUs that reported road closure for 5 days or more with serious adverse economic repercussions. Most households reported that prolonged road closures led to economic losses due to interruptions in paid employment and higher prices. The incidence of road closures does not differ significantly across areas, ethnic groups, and income levels. This is in tandem with the distribution of perceptions of road quality change over the past five years. However there is wide variation in incidence of road closures across regions (Table 7): Road closure seems to be a dire problem in the south-east. Almost half the households belonging to ENCOVI PSUs in the south-east region experienced road closures and 42% of these experienced road closures for five days or more. The other region most affected is the north ­ 34% of households belonging to ENCOVI PSUs in the north were adversely affected by road closures and one-third of these experienced them for five days or more. It is interesting that road closure was as likely to affect ENCOVI PSUs with surfaced roads, as those with unsurfaced roads. However unsurfaced roads are more likely to be closed for 5 or more days than surfaced roads. Further unsurfaced roads were also more likely to suffer from intransitability caused by 10 holes, ditches and overall poor maintenance.14 More than 30% of non-motorable roads were closed during 1999-2000 (Table 10). Although the frequency of road closure does not differ significantly across income groups, ethnic groups and areas, the causes for road closure differ a lot. Table 8 shows that in rural areas, floods, were the most commonly cited reason for road closures. In urban areas, road closures were more likely caused by `other' reasons such as political demonstrations and conflicts. Amongst the non-poor, 36% of the households belonged to ENCOVI PSUs that reported floods as the main reason for road closure. Only 18% of non-poor households affected by road closures belonged to ENCOVI PSUs that averred that landslides were the main reason for road closure. Amongst the regions, households in the north (where more than a third of the households were affected by road closures) were most afflicted by landslides (65%). In the north-east, the main culprit was winter floods (82% of the affected households were impacted by these). Amongst indigenous groups, the Kiche were most affected by floods while the Qeqchi and the Mam were most affected by landslides. Thus although the incidence of road closures does not vary much across ethnic groups, consumption quintiles and areas, causes of road closure vary a lot. Similarly the impact of road closures also varies depending on area, consumption quintile and ethnic group (Table 6). Households in urban areas are most likely to belong to ENCOVI PSUs where access to basic services was affected. On the other hand, households located in rural areas mostly belonged to ENCOVI PSUs that were denied access to schools (42% of households) and work/employment (34%) due to road closures. Most ENCOVI PSUs contended that access to markets, access to work and lower prices would be the three most important benefits of a reduction in road closure. This pattern of greater adverse impact on school and work access for the poor and much lesser for the rich, is also repeated when one looks at the five consumption quintiles, and at the ethnic groups. This is unfortunate and a fact that promises to widen the already large gap between the rich and the poor in Guatemala. IV.C. Road Works: Maintenance, Rehabilitation and Upgradation. Till 1996, there was very little road work in Guatemala. The population essentially depended on community efforts to restore roads which was very ad hoc. In 2000, less than half of the households in Guatemala belonged to ENCOVI PSUs that benefited from new road work on surfaced roads, while a little more than half reported that they had benefited from road works on an unsurfaced road (Table 7).15 Road works directed at surfaced roads usually guaranteed all-year passability for the stakeholder population, while road works directed at lower grades of roads unsurfaced roads or dirt roads did not guarantee all year pass-ability (Table 7). Since most rural areas have access to unsurfaced roads while urban areas have access to surfaced roads, the trend of directing significant efforts to improve and rehabilitate surfaced roads, is bound to expand the access gap between the rich and the non-indigenous sections of population on one side and the poorer, non-indigenous sections of the population. Most road works in Guatemala are undertaken by one of three agencies. These are: large companies (usually multinational corporations), small (local) companies or communities. It is surprising despite the civil strife and violent prolonged civil war that is widely known to have eroded social cohesion, communities in ENCOVI PSUs in Guatemala play a non-trivial role in road works (Tables not presented 14 Unsurfaced roads, if not surfaced with stones, become non-motorable in the rainy season. 15 ENCOVI PSUs were not asked about the construction and maintenance of footpaths ostensibly because these do not require any sort of `construction' effort, but just `come about' by frequent use 11 here). This underscores the importance of building social networks and capital. Most of the ENCOVI PSUs that responded that road works had been undertaken by communities in the area, rather than corporations, also opined that the quality of road works was better when undertaken by communities rather than when undertaken by corporations (Large or small). Box 3: Definition of Road Works Road works can be divided into three categories: Maintenance, Rehabilitation and Upgradation. 1. Road Maintenance: Requires that the existing road be in a good/fair condition. Road maintenance can be routine maintenance or re-graveling. a. Routine maintenance requires that the holes and ditches in roads be repaired and that roads be drained on a monthly or yearly basis (for paved roads). This type of road maintenance is preventative in nature. b. Re-graveling or Sealing: Is a higher degree of maintenance. This is usually done every 3-5 years. The procedure consists of re-graveling the road surface. The process lays a mix of asphalt, stone and sand on the road.16 2. Road Rehabilitation: implies retrieving the original condition of the road. The process is curative in nature. Rehabilitating a road is much more costly than road maintenance and is usually undertaken if the road segment did not benefit from regular maintenance. Depending on the material used, in Guatemala rehabilitation can cost anywhere from $10,000-$12,000 per kilometer. 3. Road Upgradation: Implies promoting the grade of the road e.g. move the road from being an earth road to a gravel road or paved road ("empedrados"). In Guatemala it can cost up to $ 20,000 to $22,000 per kilometer. V. PUBLIC TRANSPORT Despite the fact that public transport is popular in Guatemala, high demand for public transport has not translated itself into widespread availability. In Guatemala more than 90% of the households that have access to public transport use it almost everyday. But more than half the households in the ENCOVI PSU dataset belonged to PSUs that did not have access to public transport (Table 9). Unfortunately the absence of public transport effects the poor much more than it effects the non-poor who also have other, private, means of transport available to them. A non-poor household is more than twice as likely to be located in an ENCOVI PSU that has public transport than an extremely poor household. There are also large regional and ethnic inequities in public transport availability. Compared to other regions, households located in ENCOVI PSUs in the Metropolitan and the Central region, which have on average, higher levels of income than the other regions, are atleast one and a half times as likely to have access to public transport. Similarly in the poorest ethnic groups­the Qeqchi and the Mam­only one in five households belonged to ENCOVI PSUs that had access to public transport. The lack of public transport seems to be a supply side problem. In Guatemala public transport is provided by a small group of contractors who have a monopoly on the supply of public transport such as buses and Pullmans. The central government has also imposed a price ceiling on commuter charges for public transport, which makes it unprofitable for private entrepreneurs to enter into areas with low traffic volumes thus exacerbating the problem of access in remote areas. Public transport is assailed by insufficient infrastructure and bad service (Table 11).17 More than three- fourths of the households that had PSU level information available, reported that there was either a shortage of buses or that the buses that were available were in a very bad condition. The extremely poor sections of society were especially affected by the absence of buses. Regionally, the ENCOVI PSUs in the south-east followed by the north-west and the Peten suffered the most from the egregious quality of public transport. 16 Strictly speaking there is a small difference between `sealing' and surface treatment. The material used for sealing does not use stone. On the other hand, surface treatment uses stones in the mix as well, although the stone size in the mix is very small. 17Amongst the ENCOVI PSUs that have public transport available, 12 Despite the inequitable and regressive availability of public transport services, the five year period over 1995-2000 saw a marginal improvement in public transport services for the poorer, indigenous sections of society (Table 9). Also, almost the same percentage (~42%)of households in the eight different regions responded that they had witnessed an improvement in public transport services in the ENCOVI PSUs they belonged to. A Connection between Roads and Public Transport? A cross-tabulation between presence of motorable roads and availability of public transport indicates that the two are positively related (Table 10). In 2000, ENCOVI PSUs that had motorable roads also had a greater percentage of households without access to public transport. On the other hand in ENCOVI PSUs with no motorable roads show a much lower percentage of households with access to public transport. In Charts 1 and 2 below, it is easy to see that regions that have the best availability of public transport i.e. the metropolitan and the central regions, also have the highest percentage of ENCOVI PSU households that have access to motorable roads. On the other hand, regions that have low percentages of ENCOVI PSU households with access to public transport, namely the north, north east and north west, also have low access to motorable roads. Since households that are rich, urban and non-indigenous are more likely to belong to ENCOVI PSUs that have motorable roads than poorer, indigenous and rural households, the absence of public transport compounds problems of access for these sections of society in Guatemala. The high correlation between the presence of motorable roads and public transport also indicates that the absence of public transport may not be caused so much by the absence of demand but rather because of inadequate supply caused in turn by the absence of motor-worthy roads. Public Transport Availability (% HHs in ENCOVI PSUs) 80 70 60 50 40 30 20 10 0 Metropolitan North North-east South-east Central South-west North-west Peten Chart 1. 13 No Motorable Road (% HHs in ENCOVI PSUs) 25 20 15 10 5 0 Metropolitan North North-east South-east Central South-west North-west Peten Chart 2. VI. PHYSICAL ACCESS TO PUBLIC SERVICES From the previous two sections there are two main conclusions: Firstly, the North and the North-east are the most deprived in terms of motorable roads and public transport. This has impacted access to economic opportunities such as education and work in these regions. Secondly, the poor and indigenous groups have access to fewer and lower quality of roads and public transport. They are also most adversely affected by prolonged road closures and have witnessed very marginal improvements in public transport quality over the past 5 years. The following section examines if the absence of roads and public transport services has translated into longer travel times and distances for these sections of the population. A correlation will highlight the need for a remedial road policy in Guatemala. If on the other hand, the absence of roads and public transport has not translated into poor physical access to most services, it will indicate that poverty alleviation efforts and policy should be directed elsewhere. In this section, we examine physical access to public services such as water, health centers, employment, schools and other necessities such as wood. The hypothesis is that the absence of roads and transport services restricts physical access to many services and necessities. This part of the analysis uses the household portion of the ENCOVI 2000 dataset which is representative at the urban/rural level, for the eight regions and for the five largest ethnic groups. VI.A. Physical Access to Publicly available Drinking Water About one-thirds of the households of Guatemala have to go outside the premises of their home to reach a drinking water source. However on average, distances aren't very large. Of the many possible means of getting water, the public/private well seems to be the most popular and most households walk to reach it. A very small percentage of the households actually use a river/lake or rain water. Table 12 shows that for households that use a public water supply source such as a public water pipe or a well, distances range 14 from five-hundredths of a kilometer to three-fifths of a kilometer. On average, people have to walk anywhere between 5 minutes to less than 20 minutes to reach their water source. Although there is not much variation in travel distances and times, in the south-east region, the absence public transport is highlighted by the fact that households­which are predominantly poor in the region, have to travel much more (almost 20 minutes) compared to the national average (13 minutes) to reach publicly available water, even though travel distances are not very high. The most common means of carrying water seems to be head potage in Guatemala and this is more so in rural areas where 89% of the households walk to the water source and carry water on their heads or in their hands (Table 13). Only 1% of the households use a beast to carry water. Regionally however, except for the metropolitan region where about 65% of the population does not have water on their premises and has to walk to the water source, in all the other regions more than three-fourths of the population that gets water from outside head-carry or back-carry water. This trend does not seem to differ across consumption quintiles. If we combine data on motorable roads and examine the distance and time to reach drinking water, on average there is very little difference between travel times and distances to reach the public source of drinking water for households located in rural areas also (Table 24). On the other hand, there is a large distinction between time taken to reach water source between urban areas with motorable roads and those without motorable roads. In urban areas without motorable roads, households take more than half an hour to reach a public source of drinking water. In urban areas with motorable roads, households take less than fifteen minutes to reach a water source. The same trend is followed for households that belong to ENCOVI PSUs that are located in the central region. Otherwise across regions and ethnic classes, there is very little difference Summary Table of Main Results: Physical Access to Water. 1. One third households get water from outside sources predominantly a well (public or private) 2. Travel times or dis tances are not very long. 3. Most walk to get water ­ head carry or back carry water. 4. Rural and extremely poor households especially impacted by lack of transport facilities. Most walk. 5. The absence of motorable roads, especially for the non-poor in urban areas, more than doubles travel times to reach drinking water. VI.B. Physical Access to Wood In Guatemala, on average, almost three fourths of the households use wood for cooking and this is especially true for the poorer sections of the population (Table 15). However wood is not the only fuel used for cooking. Propane, kerosene and coal are also used. About 46% of the households that use wood for cooking, buy it from the market, while about 50% collect it, most probably from woodlots. As is expected, wood use is more significant in rural areas than in urban areas where less than half of the households use wood for cooking. Wood is used by almost all households in rural areas. More than 85% of the households in relatively poor regions like the North and the North-west, and in the not so poor regions of south-east, south-west and the Peten use wood for cooking. Keeping in mind the importance of wood in Guatemala, it is significant that average travel times to reach wood are more than an hour, each way. In the north-west, travel times can be almost one hour and twenty minutes, each way. People usually walk to the nearest wood lot Since most wood is carried on the head, this imposes a greater burden. Average travel distance are 1.4 kms. In the two poorest regions of the country, the north and the north-west, average travel distance is 1.8 kilometers. Otherwise, for households 15 that use wood for cooking, travel times and distances do not differ significantly across consumption quintiles or across ethnic groups. An interesting fact that emerges from combining household and ENCOVI PSU data is that on average at least, travel times to reach wood are slightly higher for households that belong to ENCOVI PSUs that have motorable roads compared to those that do not have motorable roads (Table 24 and Table 27). This may be because ENCOVI PSUs that do not have motorable roads tend to locate themselves closer to woodlots and wood markets ­ given their dependence on wood, than ENCOVI PSUs that have motorable roads ­ where preferences for other amenities may impact location decisions. It is easy to see that the absence of transport facilities impacts access to wood, a household necessity. There are important implications for policy. Firstly, in the long run, providing cheap alternative cooking fuel would significantly reduce the burden on household members, usually children who procure wood and carry wood on their backs. Secondly, community based sustainable wood lots and plantations should be encouraged. This also has other beneficial externalities. Finally, it indicates the need for more footpaths and carefully planned pedestrian ways which can provide an easy route to woodlots Summary Table of Main Results: Physical Access to Wood. 1. Three-fourths of households use wood for cooking. In rural areas almost all households use it for cooking. 2. Almost half the households buy wood from markets, while the other half collect it from woodlots. 3. Notwithstanding its importance, travel times and distances are large. On average, 1 hour and 1.4 kms respectively, each way. 4. Walking is the main `mode' to get to a market or a wood lot. Wood is usually carried by children. 5. Travel times and distances vary amongst regions but not across consumption quintiles or ethnic groups. 6. In the south-east and the north-west, travel times are an hour and twenty minutes, each way. 7. Implications for policy: Fuel substitution, encouraging community based plantations and building carefully planned, easily traversed pedestrian walkways (veredas). VI.C. Physical Access to Health services. On average people take three fourths of an hour to reach a health facility of any sort whether this is a public hospital, hospital of the IGSS, a private hospital, a polyclinic of the IGSS, a health center, a health post, a community center, a private clinic or a private consultation, a private pharmacy or a state pharmacy (Table 16). However in rural areas average times taken to reach a hospital are almost an hour. This is probably also caused by a combination of a relative lack of transport facilities and health services. The average time taken to reach a health facility in rural areas is much greater than that in urban areas, and almost the same percentage of population walks to the reach a health facility. On average, 63% of the population walk to a health facility, of any sort. Surprisingly, this percentage does not vary significantly between rural areas and urban areas. However as is expected, the poor and the extremely are more likely to walk to a health facility than the non-poor. In the north and the north-west, regions with the poorest transportation infrastructure, 77% of the population walks to get to the health facility of choice18. Across all categories, other than area, such as poverty groups, regions, consumption quintiles and ethnic groups, there is a correlation between travel time and mode of transport. Absence of transportation services has led to high cost of accessing health services especially in the north and the north-west. In these two regions, the average monthly payment for reaching a health facility is more than three times the national average (National average: 32 Quetzales per month; North: 103 18In these regions, it is believed that the health facility of `choice' is the nearest health facility 16 Quetzales per month; North west: 113 Quetzales per month.19 However, unless for some reason, people in the north and the north west suffer from ill-health more than the average, the high payments for transportation reflect the scarcity of cheap public transport in these regions.20 The data also show that in urban areas, on average, people have to pay about one and a half time to get to a health facility than in rural areas. This is also true for the non-poor versus the poor: the non-poor pay almost three times for transportation to a health facility than the extremely poor. This could be because urban areas­which are predominantly non-poor have better access to a greater variety of health services. Choice rather than distance may account for the higher transportation cost in this case. Amongst the indigenous groups, the Qeqchi, who are amongst the poorest indigenous groups, have to pay on average, the largest amount for transportation to a health facility. This could be because of the scarcity of health service facilities. If we compare people belonging to ENCOVI PSU with access to motorable roads with those with no access, the latter take much longer on average, to reach a health facility than the former (Table 24 and Table 27). Again this may be a combination of scarcity of health facilities and of transportation facilities. There is no difference in travel times for individuals belonging to ENCOVI PSU in urban areas, but individuals belonging to ENCOVI PSUs in rural areas take longer to reach a health facility. For the north and the north-east, this difference is especially very large. Amongst the indigenous groups, the Qeqchi who belonged to ENCOVI PSUs and had no access to a motorable road, have to travel more than an hour and a half to reach a health facility. For similarly defined Qeqchi, but who have access to a motorable road, travel times are about 45 minutes. Summary Table of Main Results: Physical Access to Health Facilities. 1. Average travel time is 46 minutes and 58% of the population walks to reach a health facility. - For people in rural areas, the extremely poor, people living in the north and in the Peten, travel times are nearly an hour. 2. 30% of people attending a health facility pay an average of 34 Quetzales to reach a health facility. - People in the north pay an average of 89 Quetzales for transportation to a health facility. 3. 9% of the population did not go to a health facility because it was too far or too expensive to get there. - 28% of people in the north and the north east and of the Qeqchi group did not go to a health facility because it was too far or too expensive. 4. People belonging to ENCOVI PSUs without motorable roads, take much longer than those with motorable roads, to reach a health facility. - The differences in travel times are much greater in the North, North east and the Peten., and for the indigenous groups, especially the Qeqchi. VI.D. Physical Access to Schools In Guatemala, children predominantly walk to school. On average, less than 15% of children attending school use means of transport to reach school irrespective of school type. The scarce use of transportation facilities along with relatively low school attendance indicates a possible correlation between the two. There could be two reasons for the correlation: First, there is relatively small incidence of schools, more pronounced in communities with poor quality of access. Second, long distances to available schools may be a factor in determining school attendance. VI.D.1. Pre-Primary School. In Guatemala, children below the age of 7 years, attend pre-primary school 19These figures do not account for frequency of trips to the health facility 20In this context, the previous section which presented data for households that belonged to ENCOVI PSUs showed that at least for PSUs that were included as part of the survey, the Peten had relatively well developed road and transport infrastructure. This indicates that the long travel times to reach a health center in the Peten may not be caused so much by the absence of adequate transport facilities but by the lack of health facilities. 17 (which includes nursery school, pre-kindergarten, kindergarten, tiny tots, or preparatory school21). Less than 10% of children, who attend pre-primary school, pay for transportation to reach the school (Table 17). What comes out vividly from Table 17 is that percentage of pre-primary school attending children who pay for transportation is relatively significant for the non-poor population and for those who reside in the metropolitan region. In other regions and amongst indigenous households, a very small percentage of pre-primary school going children pay for transportation. In fact, amongst the indigenous, there were no surveyed households that had children attending pre-primary school. Similarly in rural areas, the percentage of pre-primary school attending children who pay for transportation is very small (2%). This is probably because children probably end up doing household chores rather than attending pre-primary school. In urban areas, where the non-poor predominantly reside, 16% of the children who attend pre- primary school, pay for transportation. For households that have children attending pre-primary school and which pay for transportation, the average amount paid for school transportation is 1,080 Quetzales per year (US$138). In the metropolitan region, where most of the children who attend pre-primary school and who pay for transportation, are located, households have to pay a little more than the national average for pre-primary school transportation­Q1,400 (US$178) per school year. This is probably explained probably by the fact that school buses are provided by private contractors who have a monopsony on the supply of buses. VI.D.2. Primary Schools and Higher. Children take on average, 20 minutes to reach a school or higher (including secondary or graduate school). There is no significant difference in travel times to school amongst different areas and indigenous groups. However children in the metropolitan area take longer to reach a school than children in the rest of the country. This is despite the fact, that in the metropolitan region, a child is more likely to use a vehicle to reach a school than in the rest of the country. The longer travel time is probably explained by the fact that the metropolitan area has many more schools and a lot more traffic congestion. About 30% of children who are eligible to go to school, do so. Of these, only 14% use and pay for school transportation, and the rest walk. Most of these children belong to non-poor families and a majority reside in the metropolitan region. Almost all the children who reside in the north and in the north west, walk to school .Households pay an average of Q 890 per child per school year for school transportation. In the metropolitan and the central regions however, households pay an average of almost Q1,000 per school year per child for school transportation. Again the Qeqchi and the Mam seem most disadvantaged: Although all kids belonging to these indigenous groups walk, the ones who use school transportation have to pay more than Q1,000 per school year per child for transportation. VI.D.3. Comparison. It is immediately apparent from comparing the last columns in Table 17 and Table 18, that pre-primary school attendance in Guatemala is a sort of luxury good, and is more expensive that transportation to take children to primary schools or post-primary schools. Also, pre-primary schools are also afforded predominantly by the non-poor and the richest quintile. It is also true (tables not shown here) that less than 1% of the households in Guatemala received social assistance for transportation to reach schools and almost all this social Average travel times to reach schools are not very high in Guatemala and children take at an average 20 minutes to reach one (Table 14). Most of them walk to school, with a quarter of school going children using vehicles to reach their destination in urban areas. In rural areas, almost all children walk to school. What comes out very immediately in all this is that children in the south-east have to pay almost 10 times the national average for school transportation. Less than 1% received social assistance for reaching schools and almost all these households were located in urban areas. 21A small percentage of children aged less than 7 years also attend primary school, but they are ignored here. 18 Summary Table of Main Results: Physical Access to Schools. Pre-Primary School 1. Less than 10% of children less than 7 years of age, attend pre-primary school. Less than 10% of these pay for transportation to reach the school. - In urban areas, 16% of pre-primary school attending children pay for transportation to attend school. In rural areas, the figure is 2%. - In the metropolitan region, 22% of pre-primary school going kids pay for transportation. In al the other regions, the percentage is much lower than 10. 2. On average, households which pay for transportation for their kids, pay Q1,080 per school year for transportation. - In the metropolitan region this figure is Q1,400. - In the north and the north-west, households pay Q790 and Q890 respectively. 3. Payments paid per school year for transportation to a pre-primary school appear to increase with income. Primary School and Higher 1. 30% of eligible children attend primary school or higher. Of these, only 14% nationally use and pay for school transportation. 2. Average travel time to school is 20 minutes and this does not vary much across poverty groups, regions or across ethnic groups. 3. 83% of school going kids walk to school. Those who don't walk pay an average of Q 890 per year. - Children in the metropolitan region and the central region pay much more than the national average. - Children belonging to richer families and to the Qeqchi also pay much more than the national average for school transportation. VII. PHYSICAL ACCESS TO OTHER OPPORTUNITIES. VII.A. Transport And Agriculture In 2000, nearly half the households in Guatemala were agricultural households. Nearly half of these (45%) used motorized transport and paid for it to transport their agricultural goods to the market (Table 19). Since there is no rail transport system in Guatemala, all these households use either private transport or hired transport such as pickup trucks, trucks and trucks to transport their agricultural produce to the market. Expenditure on transport is a function of demand for transportation services as well as its supply. Thus, in the metropolitan region, 55% of agricultural households paid for transportation of their goods, while in the north, only 27% did. This indicates a scarcity of supply of transportation services in the north since households in the north are primarily agricultural. Otherwise there is not much variation in the percentage of agricultural households who paid for transportation services, amongst consumption quintiles, or regions. What does come out rather starkly is that amongst the Qeqchi, a very small percentage of agricultural households paid for transportation services, again indicating exclusion from transport services. Average expenditures on transportation services for transporting agricultural goods, including livestock, ranged from Q109 (US$13.9) in the metropolitan region to Q2,042 (US$260) in the north-east. VII.B. Private Transport Infrastructure It is not surprising that a very small percentage of the households in Guatemala own any motorized vehicle (a car/pickup truck/small truck/truck/motorcycle) or even a vehicle of any sort (motorized vehicle or a non-motorized one such as a bicycle/small cart) (Table 20). Only 9% of the population owns a car. Ownership of cars is more concentrated in urban areas, amongst 19 the non poor and in the metropolitan region. A miniscule percentage of the indigenous households own a car. Most households who possess a car, bought it second hand indicating that the secondary market in Guatemala is rather well-developed: Nearly 60% of the households which bought a car in 2000, bought a second hand car. Anecdotal evidence however suggests that a majority of these second cars are bought in the United States and driven into the country. On average, people change their cars almost every five years. Although the number of households who own 4 wheeled vehicles is not large, families that own vehicles seem to own a multiple number. Thus for every 5 families approximately, there is approximately one automobile/4 wheeled vehicle. For every three families, there is one 2 wheeled vehicle, whether it is a motorcycle or a bicycle. Urban families are almost four times as likely to own a four wheeled vehicle as are rural families. The discrepancy is not so great for 2 wheelers ­ 40% of the urban households own a 2 wheeler while 30 % of rural families own a 2 wheeler. One out of every three households in Guatemala owns a bicycle. Amongst the higher consumption quintiles, this percentage becomes higher belieing the often held belief that bicycles are inferior goods. Vehicular ownership also increases as one goes up the income ladder ­ almost 40% of the households in the fifth consumption quintile own a 4 wheeled vehicle ­ which could be either a car, a pickup or a small/large truck. The small ownership of privately owned vehicles further underscores the need to develop public transport infrastructure and networks. As is evident from previous sections, the absence of these services has meant either exclusion from merit services such as health and education. VII.C. Physical Access to Employment Travel times to work are usually a little above 40 minutes and almost 75% of the population walks to work indicating that residence locations are probably made after factoring in travel time to work. Almost 80% of the rural population walks to work and the percentage of population walking to work is larger as income levels go down. This is brought out vividly when one looks at the regional distribution of modes of transport used to reach place of employment. The metropolitan region, which is populated predominantly by the non-poor has the largest percentage of people using motorized vehicles to get to work. On the other hand, in the north and the north-west, which are the poorest regions, more than 90% of the population walk to get to work and only approximately 5% of the population use motorized means of transport. As is expected, the use of motorized means of transport increases with income. The high percentage of population using motorized means of transport in the metropolitan region is probably accounted for by the fact that more than 75% of the population in the metropolitan region work outside the municipio in which they live, thus necessitating the use of motorized means of transport. However, this does not seem to reduce their commuting time. Commuting times in the metropolitan region are the highest amongst all regions (52 minutes) explained also, probably, by large distances and considerable traffic congestion. Otherwise travel times do not differ significantly across income quintiles, regions, poverty levels and areas. The non-poor are more likely to experience two or more stages in their commute to work indicating their access to flexible means of transportation, while the poor usually only experience one stage while commuting to work. It is interesting to see that if we only examine the section of the population with ENCOVI PSU data and 20 compare commuting times for people who belong to PSUs with motorable roads with those that don't, on average, the former spend a longer period of time getting to work than the latter (Tables 24 and 27). This difference is especially pronounced for commuting population located in urban PSUs, those in the metropolitan region and in the Peten, and amongst the Kiche. Increase in commuting times in the presence of a motorable road can be explained by a variety of factors: First, the population that locates itself close to a motorable road is most probably behaviorally very different from the population that locates itself far from one. For the first, commuting may be a an important factor in their location decisions and they may be more likely to work in non-agricultural jobs, and jobs outside their municipio. On the other hand, the population that is present in ENCOVI PSUs without access to a motorable road is more likely to be agricultural and working in the vicinity of their residences. Second, these travel times do not account for modes of transport used or distances traveled. Third, travel times also do not account for congestion and other externalities. VII.C. Physical Access to Markets The presence and physical access to markets, where goods and services may be exchanged, is highly correlated with the presence of motorable roads (Table 22). Average times taken to reach a markets in ENCOVI PSUs without motorable roads, are approximately an hour and a half. ENCOVI PSUs without motorable roads are predominantly rural and time taken to reach markets exceed an hour and a half. However in rural areas, markets are located at an average distance of 12 kms and travel times are approximately an hour. The poor take much longer to reach a market than the non-poor indicating the lack of motorized means of transport. The absence of adequate transport means has also led to higher prices ­ more than two-fifths of the population belonged to ENCOVI PSUs that contended that the absence of an adequate road network had led to higher prices. VII.D. Provision and Access to Community Services There is preliminary evidence that the presence of motorable roads makes a significant difference in the availability of different types of services. Comparing Table 25 with Table 28, communities that have a motorable road are much more likely to have access to communication services such as phones and post offices. This is specially evident in the north east, south east and the north west.22 Rural areas also seem to benefit from the presence of motorable roads. The same is true for vulnerability-reducing services such as Banks, Cooperatives, Police stations and fire stations, which are nearer to communities with motorable roads than those without them. 22This is also very evident in the Peten region, but that difference is ignored here because of the low population density in the region. 21 VIII. CONCLUSIONS There are two main conclusions that come out vividly from this analysis: Firstly, road quality is a critical bottleneck for economic progress in Guatemala, impacting access to markets, employment and merit services. Lack of road works such as rehabilitation, regular and periodic repair and improvement works (filling up ditches and repairing holes) is integral to the road network of the country. This is especially important because the absence of good quality roads has exacerbated the isolation felt by poor indigenous communities, already deeply impacted by the long civil strife. Problems of road quality are most felt in the north and the north-west which are also amongst the poorest regions of the country. The absence of a `transitable' road network has also affected the provision and use of public transport ­ which is very important in Guatemala, especially because low per capita incomes make private transport infrastructure un-affordable. Thus most people walk to reach `merit services' such as hospitals and schools, as also to procure wood and water. Headloading is thus widespread and prevalent in Guatemala ­ which has important gender implications. Women spend 4 hours in some areas just to provide their households with wood. Again, in the north and the north-west, in the absence of motorized transport, policy should be directed towards emphasizing the use of non-motorized transport and building and repairing pedestrian walkways which would make traversing long distances over mountains, easier, for the rural poor. There is also preliminary evidence that the presence of motorable roads is correlated with greater availability of a variety of other services: communication services (phones and post offices) and vulnerability reducing services (such as Banks , Cooperatives, Police stations and Fire Stations). These communities are also less excluded ­ evident by the relatively lower travel distances to bus stops and other transportation infrastructure. Providing public means of easily available and affordable transportation will play an important role in integrating Guatemala and providing the poor and the indigenous sections of society with a perception of `progress' and inclusion. Secondly, it is important to provide for a decision making structure, within the administration, which is sensitive to the development priorities of the communities and considers their transport needs. In this context, encouraging the establishment of community institutions will not only empower communities and provide them with better access, as is evident from the little evident from communiities which have managed their own transportation needs, but community participation will have the double impact of securing social cohesion while assuring poverty alleviation by securing local employment and capacity building. Setting up institutions at the municipality or the community level that undertake road maintenance and rehabilitation will also ensure that these services are provided in a timely manner since decision makers will have a stake in the achieving results. This will also redress the problem of variable quality that afflicts the road system deeply. 22 REFERENCES Asian Development Bank "Assessing the Impact of Transport and Energy Infrastructure on Poverty Reduction (RETA 5947" Inception Report, May 2001. Burns, L.D. "Transportation, Temporal and Spatial Components of Accessibility" Lexington Books, DC Heath and Company, Massachusetts, USA. Dawson, J. and I. Barwell Roads are Not Enough: New Perspectives on Rural Transport Planning in Developing Countries. Intermediate Technology Publications, 1993. Gannon, Colin and Z.Liu " Poverty and Transport" TWU-30, September 1997. Gannon, Colin, K. Gwilliam, Z.Liu and C. Malmberg Calvo "Transport: Infrastructure and Services" A toolkit for Practitioners, April 2001. Jalan, Jyotsna and Martin Ravallion "Geographic Poverty Traps?" Institute for Economic Development, Discussion Paper, Boston University, 1998. Pankaj, Thampil " Designing Low-Cost Rural Transport Components to Reach the Poor" Infrastructure Note, Transportation, Water and Urban Development Department, The World Bank, March 1991. Puri, J "Guatemala: A Transportation Strategy to Alleviate Poverty. Approach and Concept Paper" The World Bank, 2001. The World Bank "Gender and Transport: A Rationale for Action" Note prepared by Michael Bamberger and Jerry Lebo, PREM Notes, Economic Policy, 1999. The World Bank, Main and Rural Roads Project, Project Appraisal Document, Loan 4260-GU, 2001. The World Bank "Quality Enhancement Review Report: Guatemala Poverty and Transport Study" Draft, September 2001. 23 MAP 1. 24 MAP 2. 25 MAP 3. 26 Table 1: International Comparisons of Paved Road Percentages and Goods Transported by Roads, Latin America and the Caribbean. Paved Roads (% Of National Road Network) 1990 1996 1997 1998 Bolivia 4 6 6 6 Costa Rica 15 17 17 21 Ecuador 3 13 19 17 El Salvador 14 20 20 20 Guatemala 25 28 28 28 Honduras 21 20 20 20 Mexico 35 37 37 30 Nicaragua 11 10 10 10 Panama 32 34 34 28 Latin America and the Carribbean 22 26 26 26 East Asia and the Pacific 17 12 12 17 Europe and Central Asia 77 82 83 87 Middle East and N. Africa 67 54 50 50 South Asia 38 42 41 57 Sub-Saharan Africa 17 17 16 15 Source: Various issues of the World Development Report published by the World Bank. Notes: 1. Percentage of paved roads is the percentage of roads that has been sealed with asphalt or similar road-building materials. However the WDR cautions that the data for most transport industries are not internationally comparable because unlike demographic statistics, national accounts etc. the collection of infrastructure data has not been standardized internationally. Data on roads are collected by the International Road Federation (IRF). National Road associations are the primary source of IRF road data; in countries where such an association does not exist or does not respond, other agencies are contacted, such as road directorates, ministries of transport or public works or central statistical offices. As a result compiled data are of uneven quality. 2. Goods transported by road is the volume of goods transported by road vehicles measured in millions of metric tons times kilometers traveled. 3. Goods transported by rail measures the tonnage of goods transported times kilometers traveled per million dollars of GDP measured in PPP terms. 4. "Na" means either "Not available" (in the case of country level data for "Goods transported by Road") or "Not applicable" (In the case of Continent-level information for "Goods transported by Road" and "Goods transported by Rail"). * Data taken from EIU, 2000 (pp. 30). 27 Table 2: Poverty by Area, Region and Ethnicity, Guatemala 2000. Extremely Poor All Poor Non-Poor ( % of Total Households) ( % of Total Households) ( % of Total Households) All 15.7 56.1 43.9 Urban 2.8 27.1 72.9 Rural 23.8 74.5 25.5 Regions Metropolitan 0.6 18.2 81.9 North 39.1 84.0 16.0 North-East 8.9 51.9 48.1 South-East 20.1 68.5 31.5 Central 8.8 51.7 48.3 South-West 17.0 64.0 36.0 North-West 31.5 82.1 17.9 Peten 12.9 67.9 32.1 Quintiles First 78.3 100 0 Second 0 100 0 Third 0 80.6 19.4 Fourth 0 0 100 Fifth 0 0 100 Ethnic Groups Non-Indigenous 7.7 41.5 58.5 Indigenous 26.3 75.8 24.2 Kiche 19.1 64.4 35.6 Qeqchi 37.6 82.9 17.1 Kaqchikel 13.6 62.6 37.4 Mam 34.3 89.9 10.1 Other Indigenous 30.7 82.2 17.8 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 28 Table 3: Distribution of Households in ENCOVI PSUs by Region, Area and Poverty Group, Guatemala 2000. (Households for which PSU level information is Available) No. of HHs for HHs with HHs in HHs in All Extremely Non-Poor which ENCOVI ENCOVI PSU Urban Rural Poor poor HHs HHs PSU level data is level data/ PSUs/ PSUs/Tot HHs available. Total ENCOVI Total al HHs. (%) ENCO ENCOVI VI HHs HHs. (%) (%) (%) (%) (%) All 4844 66 33 67 57 12 43 Regions Metropolitan 430 46 55 45 21 2 79 North 673 84 43 57 66 17 34 North-East 426 71 52 48 35 0 65 South-East 542 67 25 75 62 13 48 Central 836 66 27 73 53 12 47 South-West 768 69 30 70 60 10 40 North-West 863 72 19 81 80 29 20 Peten 335 57 28 72 44 2 56 Areas Urban 1578 46 100 0 24 3 76 Rural 3266 85 0 100 67 17 33 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 29 Table 4: Access to Road Type, by Area, Region , Ethnicity and Poverty Group, Guatemala 2000. (Households for which PSU level information is Available) HHs with Access to Road Type/ Total HHs (%) Among those with access to Road Type: Share by Characteristic (%) Surfaced Unsurfaced Dirt Footpaths No Motorable Roads* Surfaced Unsurfaced Dirt Footpaths No Motorable Roads* All 39.8 69.9 37.9 62.5 13.3 100 100 100 100 100 Urban 65.3 58.7 22.3 34.9 10.1 51.1 26.1 18.3 17.4 23.9 Rural 28.3 75 44.9 75 14.7 48.9 73.9 81.7 82.6 76 Poverty Levels Non-Poor 52 67.4 32 58.8 8.9 60.4 44.7 39.1 81.1 31 All Poor 29.4 72.1 43 86.1 17 39.6 55.3 60.9 18.9 69 Extremely Poor 20.5 69.4 42.4 86.1 23.4 7.1 13.6 15.3 18.9 24.3 Regions Metropolitan 65.2 57.3 22.3 34.3 3.2 33 16.5 11.8 11 4.7 North 25.8 66.3 15 75.1 17.8 6.5 9.5 4 12.1 13 North-East 18.7 77.2 51.2 34.9 18.3 4.3 10.2 12.5 5.2 13.7 South-East 32.6 73.4 69.9 77.8 14.4 7.9 10.2 17.8 12 10.1 Central 47.4 83.6 30 61.5 9.5 12.4 12.5 8.3 10.3 7.4 South-West 37 76.2 38.7 69.3 16.1 22.7 26.7 25 27.1 31 North-West 29.3 61.7 48.2 91.7 20.5 9.6 11.5 16.6 19.1 19.5 Peten 47.7 70 51.1 67 2.3 3.6 3 4.1 3.2 0.5 Quintiles First 20.7 70.5 43.5 86.3 22.3 9.2 17.8 20.3 24.4 29.7 Second 29.6 71.7 44.1 77.9 16.4 15.1 20.9 23.7 25.4 25.1 Third 37.1 74.4 40.4 61.3 11.3 18.7 21.4 21.4 19.7 17.3 Fourth 48.3 70.8 33.7 52.5 8.9 26.4 22.1 19.4 18.3 14.4 Fifth 60.7 62.2 28.7 38 8.9 30.6 17.8 15.2 12.2 13.5 Ethnic Groups Non-Indigenous 44.5 69.5 41.8 49.7 11.2 62.5 55.7 61.8 44.6 47.2 Indigenous 33.9 70.4 32.9 78.9 15.9 37.5 44.3 38.2 55.4 52.8 Kiche 43.6 71.7 27.7 86.9 4.4 9.8 9.1 6.5 12.4 2.9 Qeqchi 21.3 59.3 13.2 67.6 23 4.5 7.1 2.9 9 13.9 Kaqchikel 43.1 86.6 30.2 69.2 14 9.5 10.8 7 9.7 9.9 Mam 28.9 72.6 41.4 89.7 17.9 6.3 9 9.5 12.5 11.7 Other Indigenous 32.1 61.9 50.2 80 20.5 7.5 8.2 12.3 11.9 14.4 * Type A road is Carretera pavimentada o balastrada; Type B:Carretera de tierra o terraceria ;,Type C: Carretera de herradura sin balastre; Type D: Veredas. HHs can have access to more than one type of road. A motorable road is a surfaced or an unsurfaced road. Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala 30 Table 5: Years of Existence Of `Communities' and Road Access, Guatemala 2000. [HHs for which PSU level information is available] No. of years for HHs with access to HHs with access to HHs with access to HHs with access to which community Type A road/Total Type B road/Total Type C road/Total Type D road/Total to which HH HHs HHs HHs HHs belongs, has existed. (%) (%) (%) (%) > 20 years 79 81 82 83 10-20 years 12 9 12 12 <10 years 9 9 6 5 Total 100 100 100 100 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala 31 Table 6: Perceptions on change in Road condition, Guatemala (1995-2000) (Households for which PSU level information is Available) HHs which reported HHs which reported HHs which reported no Total improvement/Total worsening/Total change/Total HHs HHs HHs (%) (%) (%) All 43.8 16.4 39.8 100 Area Urban 52.6 12.2 35.2 100 Rural 39.7 18.3 41.9 100 Poverty Levels Non-Poor 48.1 15.6 36.3 100 All Poor 40.1 17.1 42.8 100 Extremely Poor 34.5 21.7 43.8 100 Regions Metropolitan 53.4 6.4 40.2 100 North 27.9 24.9 47.2 100 North-East 41.2 13.1 45.8 100 South-East 44.5 9.8 45.8 100 Central 33.3 25.0 41.7 100 South-West 57.3 17.8 24.9 100 North-West 22.0 21.9 56.1 100 Peten 53.9 20.8 25.2 100 Quintiles First 34.4 21.2 44.4 100 Second 41.5 16.3 42.2 100 Third 43.6 13.7 42.7 100 Fourth 43.6 14.9 41.5 100 Fifth 54.6 16.9 28.5 100 Ethnic Groups Non-Indigenous 45.9 17.9 36.2 100 Indigenous 41.0 14.5 44.5 100 Kiche 41.5 16.1 42.4 100 Qeqchi 24.5 15.2 60.3 100 Kaqchikel 46.5 8.4 45.1 100 Mam 57.1 13.8 29.1 100 Other Indigenous 34.5 19.3 46.2 100 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala 32 Table 8: Households perceptions of New Road Works by Road Type, Guatemala (1995-2000). (Households for which PSU level information is Available) Reported New Road works for All year passability of newly Reported road works for All year passability of newly worked surfaced roads. worked surfaced roads. unsurfaced roads. unsurfaced roads. HHs with new Among those who HHs with Among those who have HHs with Among those who HHs with new Among those who have roads/ Total reported new road new roads/ all year passability: new roads/ reported new road roads/ Total all year passability: HHs works: Share of HHs Total HHs Share of HHs by Total HHs works: HHs Share of HHs by by characteristic characteristic Share of HHs by characteristic (%) (%) (%) (%) (%) characteristic (%) (%) (%) All 45.7 100 90.5 100 56.6 100 77.3 100 Urban 74.3 62.6 98.2 67.9 45.5 30.9 97.7 39.1 Rural 27.8 37.4 77.6 32.1 63.6 69.1 68.2 60.9 Poverty Levels Non-Poor 63.7 71.2 93.1 73.3 42.6 38.4 84.9 42.2 All Poor 26.9 28.8 83.9 26.7 71.3 61.6 72.6 57.8 Extremely Poor 16.5 4.9 94.3 5.1 77.9 18.5 66.5 15.9 Regions Metropolitan 81.9 43.3 100.0 47.9 30.3 13.0 100.0 16.7 North 26.0 3.5 37.3 1.5 71.4 7.8 69.5 7.0 North-East 37.8 8.3 43.7 4.0 39.3 6.9 38.3 3.4 South-East 34.9 10.0 100.0 11.1 70.9 16.5 78.2 16.7 Central 51.6 6.8 100.0 7.5 8.1 0.9 100.0 1.1 South-West 42.3 22.5 88.1 21.9 82.8 35.5 88.4 40.6 North-West 12.0 2.9 100.0 3.2 68.6 13.3 51.1 8.8 Peten 24.3 2.7 100.0 3.0 67.3 6.1 71.0 5.6 Quintiles First 21.5 8.2 93.2 8.4 76.3 23.4 71.0 21.5 Second 26.5 10.4 78.6 9.0 69.8 22.1 75.0 21.4 Third 36.2 13.2 80.9 11.8 64.4 19.0 70.5 17.4 Fourth 59.9 30.6 91.7 31.0 46.8 19.3 81.5 20.4 Fifth 70.0 37.7 95.5 39.8 37.2 16.2 92.7 19.4 Ethnic Groups Non-Indigenous 53.7 67.4 89.4 66.6 44.2 44.7 82.6 47.8 Indigenous 34.9 32.6 92.7 33.4 73.4 55.3 73.0 52.2 Kiche 56.9 8.6 99.0 9.4 68.2 8.3 86.4 9.3 Qechi 6.3 1.2 100.0 1.3 80.9 12.4 41.9 6.7 Kaqchikel 32.3 4.9 100.0 5.4 47.4 5.7 73.8 5.5 Mam 49.4 9.9 100.0 10.9 95.7 15.5 93.6 18.7 Other Indigenous 32.9 7.9 71.0 6.2 67.9 13.2 69.0 11.8 * In the ENCOVI 2000 a surfaced road is Carretera pavimentada o balastrada; Unsurfaced road:Carretera de tierra o terraceria ;,a dirt road: Carretera de herradura sin balastre; anda footpath: Veredas. HHs can have access to more than one type of road. Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala 33 Table 7: Perceptions on Road Closure, Guatemala, 1999-2000. [Households for which PSU level information is Available] Road Closure Road closure for 5 or Main reason for Road Closure Main Activity affected due to Road closure. more days. [HHs perceiving any of the following as main reason [HHs whose activity is affected/Total HHs which for road closure/Total HHs which reported road reported Road closure] (%) (%) closure] (%) (%) HHs Share of HHs Share of Landslide Mudslide Flood/winter Rd. Other Work Market School Health Other reporting road HHs reporting HHs sank access center closure/Total reporting roads reporting access HHs road closure, closure for road by 5 or more closure for characteristic days /Total 5 or more HHs which days, by reported characterist road ic closure All 27.6 100 19.9 100 27.9 10.6 35.5 4.7 21.3 27.9 10.6 35.5 4.7 21.3 Area Urban 28.5 32.5 19.8 31.2 15.1 12.7 21.4 9.0 41.7 15.1 12.7 21.4 9.0 41.7 Rural 27.1 67.5 20.0 68.9 34.0 9.5 42.3 2.7 11.5 34.0 9.5 42.3 2.7 11.5 Poverty Levels Non-Poor 26.6 44.5 18.5 42.7 17.7 10.6 35.5 7.3 28.9 17.7 10.6 35.5 7.3 28.9 All Poor 28.4 55.5 21.3 57.3 36.1 10.5 35.6 2.7 15.2 36.1 10.5 35.6 2.7 15.2 Extremely Poor 31.9 15.9 27.2 18.8 42.8 11.9 33.8 5.9 5.7 42.8 11.9 33.8 5.9 5.7 Regions Metropolitan 25.1 17.8 21.1 20.6 10.6 15.3 10.5 16.3 47.3 10.6 15.3 10.5 16.3 47.3 North 34.4 12.1 30.0 14.6 64.8 0.0 31.3 0.0 4.0 64.8 0.0 31.3 0.0 4.0 North-East 20.7 7.5 16.3 8.1 13.7 4.2 82.1 0.0 0.0 13.7 4.2 82.1 0.0 0.0 South-East 47.8 16.1 41.5 19.3 36.4 8.7 34.6 11.2 9.0 36.4 8.7 34.6 11.2 9.0 Central 31.9 12.0 22.2 11.6 22.0 23.8 39.6 0.0 14.7 22.0 23.8 39.6 0.0 14.7 South-West 20.3 18.8 10.5 13.4 15.0 0.0 44.3 0.0 40.8 15.0 0.0 44.3 0.0 40.8 North-West 28.4 13.1 16.1 10.2 44.1 20.1 25.4 0.0 10.3 44.1 20.1 25.4 0.0 10.3 Peten 24.6 2.6 14.5 2.1 0.0 23.2 73.1 0.0 3.7 0.0 23.2 73.1 0.0 3.7 Quintiles First 31.0 19.9 25.5 22.6 45.0 11.8 32.0 5.2 6.0 45.0 11.8 32.0 5.2 6.0 Second 28.9 21.3 20.0 20.4 32.0 9.2 40.4 1.2 17.1 32.0 9.2 40.4 1.2 17.1 Third 25.4 18.7 18.6 18.9 27.0 10.0 35.5 1.3 26.2 27.0 10.0 35.5 1.3 26.2 Fourth 23.2 18.2 16.1 17.4 24.4 8.5 35.8 3.9 27.5 24.4 8.5 35.8 3.9 27.5 Fifth 30.1 21.9 20.7 20.8 12.0 12.9 33.7 11.4 30.0 12.0 12.9 33.7 11.4 30.0 Ethnic Groups Non-indigenous 28.9 58.6 21.6 60.5 18.7 10.8 38.1 8.1 24.4 18.7 10.8 38.1 8.1 24.4 Indigenous 25.9 41.4 17.9 39.5 40.9 10.2 32.0 0.0 16.9 40.9 10.2 32.0 0.0 16.9 Kiche 32.9 10.3 11.4 4.9 23.3 12.6 43.3 0.0 20.8 23.3 12.6 43.3 0.0 20.8 Qeqchi 31.7 9.2 26.4 10.6 51.5 0.0 40.5 0.0 8.0 51.5 0.0 40.5 0.0 8.0 Kaqchikel 26.2 8.9 18.0 8.5 24.8 18.6 28.4 0.0 28.2 24.8 18.6 28.4 0.0 28.2 Mam 14.1 4.4 11.2 4.9 67.5 0.0 18.3 0.0 14.2 67.5 0.0 18.3 0.0 14.2 Other Indigenous 25.0 8.4 22.8 10.6 53.9 15.0 19.7 0.0 11.4 53.9 15.0 19.7 0.0 11.4 Source: The World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 34 Table 9: State and Physical Access to Public Transport, Guatemala 2000. [Households for which PSU level information is Available] Access to Public Transport Change in State of Public transport Frequency of Use of Public Transport HHs with (1995-2000) No access HHs with Share of HHs HHs which HHs HHs HHs which use HHs which HHs which Other to pub. access/Total with access to reported which which Public transport use public use public frequency Tpt./Total HHs Pub. Ttp.,by Improved reported reported everyday/Total transport transport /Total HHs characteristic. Pub. worse Pub. same Pub. HHs with access couple of once a HHs with Tpt./Total Tpt./Total Tpt./Total to public tpt. times a week/Total access to HHs w/ HHs HHs week/Total HHs with public tpt. Pub. Tpt. w/ Pub. w/ Pub. HHs with access to Tpt. Tpt. access to public tpt. public tpt. (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) All 47.3 100 41.6 19.4 39.0 92.2 6.8 0.3 0.6 52.7 Area Urban 65.0 43.1 32.0 26.0 42.1 98.3 1.7 0.0 0.0 35.0 Rural 39.3 56.9 48.8 14.5 36.7 87.6 10.7 0.6 1.1 60.7 Poverty Levels Non-Poor 58.7 57.2 41.3 22.3 36.4 96.5 3.1 0.0 0.4 40.9 All Poor 37.6 42.8 42.0 15.5 42.5 86.5 11.8 0.7 1.0 62.4 Extremely Poor 27.4 7.9 39.0 21.2 39.8 81.8 15.7 1.4 1.1 72.6 Regions Metropolitan 74.4 30.7 32.4 29.3 38.2 100.0 0.0 0.0 0.0 25.6 North 35.0 7.2 48.3 16.6 35.1 87.5 12.5 0.0 0.0 65.0 North-East 35.0 7.3 35.7 23.5 40.8 96.4 3.7 0.0 0.0 65.0 South-East 45.4 8.9 49.9 13.3 36.8 90.9 6.1 0.0 3.1 54.6 Central 67.6 14.8 42.5 19.5 38.0 91.4 6.3 0.0 2.3 32.4 South-West 33.6 18.1 49.7 8.9 41.4 90.6 9.4 0.0 0.0 66.4 North-West 38.3 10.3 45.3 16.0 38.7 74.0 22.7 3.3 0.0 61.7 Peten 42.9 2.7 40.2 7.6 52.3 94.5 5.5 0.0 0.0 57.1 Quintiles First 28.9 10.8 36.8 20.0 43.2 81.1 15.5 1.8 1.6 71.1 Second 36.9 15.9 41.2 14.4 44.4 86.7 12.2 0.4 0.8 63.6 Third 47.2 20.2 45.1 14.7 40.2 91.3 7.7 0.3 0.7 52.8 Fourth 54.6 24.9 46.2 20.1 33.7 94.5 4.8 0.1 0.6 45.4 Fifth 66.4 28.2 37.0 24.8 38.3 98.1 1.7 0.0 0.2 33.6 Ethnic Groups Non-Indigenous 55.0 65.0 39.5 21.9 38.7 96.4 3.1 0.0 0.5 45.0 Indigenous 37.9 35.0 45.5 14.8 39.7 84.3 13.7 1.0 1.0 62.1 Kiche 42.4 7.7 60.1 12.8 27.1 74.2 21.4 4.4 0.0 57.6 Qeqchi 19.5 3.3 48.6 2.8 48.6 89.8 10.2 0.0 0.0 80.5 Kaqchikel 61.1 12.1 57.6 10.6 31.9 85.7 11.5 0.0 2.8 38.9 Mam 23.8 4.2 19.4 2.0 78.7 80.9 19.2 0.0 0.0 76.2 Other Indigenous 39.1 7.7 24.8 35.8 39.4 92.0 8.0 0.0 0.0 60.9 Source: The World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 35 Table 10: Availability of Public Transport and Motorable Roads, Guatemala 2000 [[Households for which PSU level information is Available] If Motorable Road is located in PSU: If Motorable Road is not located in PSU: HHs which respond that HHs which reported HHs which HHs which reported public transport is road closures/Total respond that road closures/Total available/Total HHs. HHs public transport is HHs available/Total HHs. (%) (%) (%) (%) All 50.3 (2.9) 27 (2.6) 28.3 (6.2) 31.5 (6.3) Area Urban 65.7 (5.4) 34.3 (5.4) 58.3 (9.7) 41.7 (9.8) Rural 42.8 (3.4) 57.2 (3.4) 18.8 (7.1) 81.1 (7.1) Poverty Levels Non-Poor 60.1 (3.6) 25.9 (3.2) 44.2 (7.9) 32.9 (7.8) All Poor 41 (3.2) 27.9 (3) 21.2 (6.9) 30.9 (7.2) Extremely Poor 31.3 (4.1) 30.4 (4.4) 14.6 (6.8) 36.7 (9.8) Regions Metropolitan 73.5 (8.5) 26.5 (8.5) 100 (0) 0 (0) North 41.7 (7.8) 58.3 (7.8) 4.3 (2.7) 95.7 (2.7) North-East 42.1 (10.2) 57.9 (10.2) 3.1 (3.3) 96.9 (3.3) South-East 52.0 (8.3) 48.0 (8.3) 6.2 (5.1) 93.8 (5.1) Central 67.3 (5.5) 32.7 (5.5) 70.2 (17.5) 29.8 (17.5) South-West 31.1 (6) 68.9 (5.9) 46.6 (13.9) 53.4 (13.9) North-West 45.1 (6.1) 54.9 (6.1) 11.9 (8.4) 88.1 (8.4) Peten 43.9 (9.1) 100 (0) - - Quintiles First 33.2 (3.9) 30.3 (4) 14.1 (6.0) 33.3 (9.2) Second 40.1 (3.8) 28.3 (3.7) 20.7 (6.3) 32 (7.9) Third 48.6 (4.0) 25.6 (3.5) 35.9 (11.6) 24 (8) Fourth 56.4 (4.6) 23.1 (3.2) 36.2 (8.4) 24.1 (7.3) Fifth 67.5 (3.6) 28.7 (4.1) 55.6 (10.3) 44.3 (11.2) Ethnic Groups Non-Indigenous 57.6 (3.7) 27.9 (3.4) 33.9 (7.9) 36.1 (8.9) Indigenous 40.4 (3.6) 25.6 (3.3) 23.2 (9.2) 27.4 (7.6) Kiche 43.0 (7.3) - - - Qeqchi 25.3 (6.8) - - - Kaqchikel 58.0 (7.1) - - - Mam 28.0 (8.9) 12.1 (6.2) 0 23 (15.3) Other Indigenous 42.8 (6.8) 27.9 (3.4) 24.6 (11.5) 36.1 (8.9) Note: Numbers in brackets ( ) are standard errors. " - " shows not enough sample points. Source: The World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 36 Table 11: Main Reported Problem with Public Transport, Household Perceptions, Guatemala, 2000. [Households for which PSU level information is Available] HHs reporting main problem with public transport/Total HHs with access to public transport (%) Insufficient Bad Bad Service Faraway Unreliable No Other Total Buses condition of Stops frequency and night Buses schedule service All 50.6 16.3 7.8 2.6 4.7 4.6 13.4 100 Area Urban 43.4 24.1 10.1 0.0 5.5 4.0 12.9 100 Rural 56.0 10.5 6.1 4.5 4.0 5.1 13.8 100 Poverty Levels Non-Poor 44.5 19.5 10.7 1.6 5.8 3.5 14.3 100 All Poor 58.7 12.1 3.9 3.8 3.2 6.2 12.2 100 Extremely Poor 69.1 4.9 2.7 3.3 3.3 5.4 11.4 100 Regions Metropolitan 34.5 24.2 17.6 0.7 5.2 3.2 14.7 100 North 57.1 12.0 0.0 0.0 8.3 6.8 15.9 100 North-East 37.8 13.5 2.4 12.3 2.5 4.6 27.0 100 South-East 75.1 9.1 0.0 0.0 0.0 2.2 13.6 100 Central 61.2 24.6 3.5 0.0 3.7 2.2 4.8 100 South-West 45.1 8.7 6.4 7.2 10.0 9.1 13.5 100 North-West 69.6 10.1 4.2 0.0 0.0 6.4 9.7 100 Peten 69.6 2.1 7.4 6.2 0.0 0.0 14.7 100 Quintiles First 67.9 5.0 4.0 2.6 2.4 6.7 11.4 100 Second 59.3 12.5 2.9 3.9 3.4 8.6 9.3 100 Third 51.8 17.9 4.6 5.3 3.4 3.0 14.1 100 Fourth 48.2 20.6 7.9 1.5 4.3 3.3 14.2 100 Fifth 40.4 17.9 14.2 0.8 7.6 4.0 15.2 100 Ethnic Groups Non-Indigenous 47.1 19.1 8.2 1.9 6.1 3.9 13.7 100 Indigenous 57.1 11.1 7.1 3.8 2.1 6.0 12.7 100 Kiche 58.8 4.5 1.7 0.2 3.8 5.6 25.6 100 Qeqchi 56.0 17.6 0.0 0.0 8.5 11.4 6.4 100 Kaqchikel 50.1 12.5 11.4 11.0 0.9 3.3 10.8 100 Mam 42.7 0.0 14.9 0.0 0.0 21.5 21.0 100 Other Indigenous 74.7 19.0 4.7 0.0 0.6 0.0 1.0 100 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 37 Table 12: Travel time and Travel distance to Water Source, by Households, Guatemala 2000. (Time and Distance to source measured in Minutes and Kilometers respectively). Travel time (Minutes) Distance (Kilometers) Public Well River/ Water Public Well River/ Water water (public/pr lake truck Rain Other water (public/pr lake truck Rain Other pipe ivate) pipe ivate) All 13 10 18 5 5 12 0.21 0.14 0.65 0.07 0.05 0.12 Area Urban 13 15 12 9 4 14 0.13 0.13 1.56 0.13 0.00 0.08 Rural 13 10 19 2 6 9 0.25 0.14 0.62 0.02 0.05 0.15 Poverty Levels Non-Poor 13 9 15 5 3 12 0.18 0.08 0.70 0.07 0.02 0.10 All Poor 13 11 20 7 7 10 0.22 0.18 0.64 0.11 0.06 0.13 Extremely Poor 9 12 18 30 4 12 0.22 0.28 0.65 1.50 0.01 0.19 Regions Metropolitan 13 19 20 4 17 - 0.09 0.34 0.40 0.05 0.09 - North 8 10 19 5 8 - 0.38 0.37 0.63 0.05 0.19 - North-East 12 6 27 0 15 5 0.15 0.07 0.66 0 0.15 0.12 South-East 19 18 17 6 10 8 0.32 0.24 0.43 0.02 0.02 0.08 Central 13 8 23 0 10 - 0.13 0.05 0.51 0.30 0.10 - South-West 12 8 19 0 4 9 0.12 0.10 1.33 0 0 0.04 North-West 14 12 15 14 5 - 0.28 0.19 0.42 0.25 0.10 - Peten 9 12 10 43 4 19 0.18 0.20 0.18 0.66 0.04 0.46 Quintiles First 11 11 19 30 3 11 0.22 0.26 0.66 1.50 0.01 0.20 Second 17 12 22 - 15 8 0.23 0.15 0.63 0.19 0.15 - Third 11 10 16 7 3 12 0.22 0.13 0.53 0.07 0.02 0.12 Fourth 14 8 14 3 2 13 0.21 0.08 0.79 0.05 0 0.11 Fifth 11 11 31 7 3 5 0.07 0.06 0.96 0.07 0.04 0.02 Ethnic Groups Non-Indigenous 13 10 22 5 8 12 0.21 0.10 0.57 0.05 0.06 0.12 Indigenous 13 10 22 5 8 12 0.20 0.20 0.68 0.42 0.05 0.12 Kiche 11 9 17 5 - 6 0.16 0.19 1.64 0.01 - 0.05 Qeqchi 7 6 15 30 5 11 0.32 0.17 0.55 1.50 0.05 0.18 Kaqchikel 13 10 25 - - 6 0.10 0.18 0.50 - - 0.19 Mam 11 10 16 30 5 11 0.19 0.17 0.34 1.00 0 0.06 Other Indigenous 21 15 21 0 5 8 0.24 0.28 0.59 - 0.09 0.12 Source:World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. Note: `-' shows that there are no data points in that category. 38 Table 13: Mode of Transport to Water Source, By Households, Guatemala 2000. HHs walking HHs using a HHs using a HHs using a HHs using a HHs using to get beast to boat to carry bicycle to truck to other means water/Total carry water/Total carry carry to carry HHs water/Total HHs water/Total water/Total water/Total HHs HHs HHs HHs. (%) (%) (%) (%) (%) (%) All 84.6 1.2 0.2 0 0.5 13.5 Area Urban 63.8 0.8 0.4 0.1 0.8 34.2 Rural 88.7 1.3 0.1 0 0.4 9.4 Poverty Levels Non-Poor 80.1 1.2 0.2 0 1.0 17.5 All Poor 87.1 1.3 0.2 0.1 0.2 11.3 Extremely Poor 93.4 0.8 0 0 0.2 5.6 Regions Metropolitan 65.6 0.1 0.6 0 0 33.7 North 88.3 0.7 0 0 0 11.0 North-East 79.2 1.0 0 0.2 0.8 18.9 South-East 87.4 3.4 0 0 1.5 7.7 Central 84.6 0.6 0 0.2 0.3 14.3 South-West 90.9 0.6 0.2 0 0.4 7.9 North-West 83.8 1.3 0.4 0 0.7 13.9 Peten 89.7 6.1 0 0.4 0.3 3.4 Quintiles First 94.0 1.1 0.2 0 0.0 4.7 Second 88.4 1.9 0.5 0 0 9.24 Third 75.3 1.2 0 0.2 0.4 23 Fourth 80.0 1.4 0.4 0 1.7 16.5 Fifth 70.8 0.9 1.3 0 0 27 Ethnic Groups Non-Indigenous 82.6 1.7 0.2 0 0.7 14.8 Indigenous - - - - - - Kiche - - - - - - Qeqchi - - - - - - Kaqchikel - - - - - - Mam 91.6 0.3 0.7 0 0.4 7.0 Other Indigenous 86.8 0.8 0.2 0.1 0.2 11.9 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica Guatemala. Note: `-' shows that there are no data points in that category. 39 Table 14: Mode of Transport by Water source, for Households, Guatemala 2000. Source of water HHs with Travel time MODE OF TRANSPORT water to water HHs HHs HHs using a HHs HHs HHs Total source/Total source. walking to using a boat to reach using a using a using HHs water beast to water/Total bicycle truck other source/Tot reach HHs to reach to means al HHs water/T water/To reach to otal tal HHs water/ reach HHs Total water/ HHs Total HHs (Mins) (%) (%) (%) (%) (%) (%) (%) Water pipe inside house 53.4 - - - - - - - Water pipe inside property 13.3 Public water pipe 4.3 13 83.1 0.63 0.45 0.11 0.17 16 100 Well (public/private) 15.3 10 90.1 1.45 0 0 0.17 8.2 100 River/lake 8.4 18 - - - - - - - Water truck 0.5 5 - - - - - - - Ran 0.6 5 - - - - - - - Other 4.2 12 71.2 2.14 26.4 0 0 0 100 Total 85 1.34 0.24 0.34 13.2 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadica, Guatemala. Note: `-' shows that there are no data points in that category. 40 Table 15: Time, Distance and Mode for Transporting Wood by Households, Guatemala 2000. Wood Use One way Travel One way Travel Most favored Mode for (% of HHs) Time (Mins.) Distance carrying wood. (kms.) Walking Beast/Truck All 74 64 1.4 77 16 Urban 46 51 1 81 13 Rural 96 67 1.4 77 17 Poverty level Non poor 56 58 1.1 71 19 All poor 96 66 1.5 81 14 Extremely poor 99 66 1.7 88 10 Regions Metropolitan 34 43 0.7 88 8 North 95 65 1.8 95 4 North East 79 68 1.1 78 10 South East 90 72 1.6 62 32 Central 82 61 1.6 67 20 South West 88 60 1.1 77 14 North West 95 77 1.7 83 16 Peten 92 69 1.8 52 38 Consumption Quintiles First 100 67 1.7 87 10 Second 99 71 1.5 80 16 Third 91 63 1.4 75 17 Fourth 77 57 1.1 74 17 Fifth 35 52 0.9 62 26 Ethnic Groups Non-Indigenous 63 63 1.2 71 20 Indigenous 92 64 1.5 84 87 Kiche 90 45 1.2 89 7 Qeqchi 97 56 1.7 93 3 Kaqchikel 86 60 1.5 74 18 Mam 98 67 1.5 82 16 Other Indigenous 91 89 1.6 84 15 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. Travel times and distances calculated only for households that use wood. 41 Table 16: Time, Cost and Mode of travel taken to reach a Health Facility by Individuals, Guatemala 2000. Average Most favored mode of Percentage of sick Average monthly Percentage of sick who travel Time transportation to reach people paying for Payment for did not visit a health One way. the Health Facility transportation to transportation to center because ** (% of population using the Health facility *** Health Facility*** i) the doctor was toofar mode if using a health ii) tptn. was too facility).** expensive iii) No means of tptn existed. (Minutes) Walking Vehicles* (%) (Quetzales) (%) All 46 58 39 29 34 9 Urban 36 61 37 27 44 2 Rural 57 61 34 31 25 13 Poverty level Non poor 42 53 45 31 35 5 All poor 54 68 28 26 31 13 Extremely poor 68 78 18 15 21 17 Regions Metropolitan 41 57 37 33 43 2 North 64 70 22 9 89 29 North East 52 46 43 29 39 22 South East 55 61 32 19 31 8 Central 38 57 38 27 31 4 South West 45 56 39 36 22 4 North West 43 75 22 23 21 5 Peten 63 54 35 27 53 11 Consumption Quintiles First 56 82 18 14 23 17 Second 49 69 31 27 27 11 Third 45 65 34 31 33 9 Fourth 43 58 341 38 38 6 Fifth 37 50 49 27 35 3 Ethnic Group Non-Indigenous 46 53 44 30 36 6 Indigenous 46 67 31 30 31 13 Kiche 37 63 34 36 30 7 Qeqchi 65 69 27 18 80 28 Kaqchike 43 67 29 27 53 3 Mam 58 68 32 30 13 7 Other -Indigenous 47 75 27 17 27 25 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. * These include all types of vehicles which use engines and which need a `minimum' road presence. These types of vehicles include small trucks, taxis, cars,motorcycles, trucks and ambulances. **Is calculated for those people who visit (i.e. are not already in one) some health care centers include a public hospital, hospital of the IGSS, a private hospital, a polyclinic of the IGSS, a health center, a health post, a community center, a private clinic or a private consultation, a private pharmacy or a state pharmacy. It excludes time taken by persons who were taken care of at their own home or at other homes. *** Payments made by those who went to any health facility even if the person was not ill and if the person went somewhere other than to another home to get treatment. 42 Table 17: Mode of transport and Time to reach a Pre-Primary School, for Children, Guatemala 2000 Percentage of Population Transportation children paying for school Cost of Reaching a attending pre- transportation/ Pre-Primary primary Population School school/ All attending Pre- children of Primary school. age 6 years or less. (%) % (Quetzales)* All 9.4 9.2 1,080 Urban 16 16 1,090 Rural 7 2 980 Poverty level Non poor 16 16 1,140 All poor 6 1 290 Extremely poor 5 1 250** Regions Metropolitan 17 22 1,400 North 6 3 790 North East 7 3 890 South East 9 2 480 Central 10 7 540 South West 9 4 770 North West 5 1 380 Peten 9 3 750** Consumption Quintiles First 5 0.7 250** Second 7 1.2 350 Third 8 0.6 150** Fourth 13 8.4 770 Fifth 24 23.4 1,220 Ethnic Group Non-Indigenous - 9.2 1,080 Indigenous - - - Kiche - - - Qeqchi - - - Kaqchike - - - Mam - - - Other May - - - Other Indigenous - - - Source: World Bank calculations using ENCOVI 2000, InstitutoNacional de Estadistica, Guatemala. * Uses unweighted means.Only 67 observations. ** Only 1 observation. 43 Table 18: Mode of transport and Time to reach a Primary School or higher, for Children, Guatemala 2000 Percentage of Percentage of Average one Population using the Transportation Cost of students who students who way travel time following means to reach a Reaching a Primary (or attended paid for school to reach a secondary school/ Higher) School over the Primary school transportation/ Primary (or Population going to School year. or higher/All Students who Higher) School Primary (or Higher) children (>7 attended School. years of age) primary or (%) higher school * (%) (%) (Minutes) Walking Vehicles** (Quetzales) All 30 14 20 83 14 890 Urban 34 23 23 74 25 960 Rural 28 7 18 91 8 720 Poverty Levels Non poor 32 24 23 73 26 940 All poor 29 4 17 95 4 610 Extremely poor 25 1 16 97 1 470 Regions Metropolitan 32 32 26 73 27 990 North 28 6 17 96 4 660 North East 28 6 16 80 19 830 South East 29 4 16 91 8 770 Central 30 12 20 84 15 990 South West 33 12 19 87 11 710 North West 27 4 19 95 4 820 Peten 34 7 18 87 11 820 Consumption Quintiles First 26 0.7 17 97 1 670 Second 31 3 17 96 4 550 Third 30 7 17 89 10 630 Fourth 30 12 19 85 15 620 Fifth 35 37 27 60 37 1,030 Ethnic Group Non-Indigenous 32 17 22 80 19 920 Indigenous 28 9 18 95 5 730 Kiche 26 13 17 90 10 720 Qeqchi 25 4 17 98 2 1,020 Kaqchike 27 19 18 91 9 710 Mam 31 3 21 98 1 1,070 Other Indigenous 28 4 18 97 2 890 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. This table pertains to people above the age of 7, who attended either adult education classes, preparatory classes, primary school, middle school, university or graduate school. ** Includes all modes of transport other than boat and beast. This could be bicycle, small truck, truck, taxi, car, motorcycle or other means of transport. 44 Table 19: Expenditure on Transportation by Households, Guatemala 2000 (Quetzales). Percentage of hhs who paid Amount paid for for transporting transporting agricultural agricultural goods/Total products and livestock* agricultural households.* (Over a year) (Quetzales) (%) All 45 309 Urban 52 274 Rural 44 316 Poverty Levels Non Poor 48 650 Poor 44 145 Extremely Poor 36 101 Regions Metropolitan 55 109 North 27 230 North East 41 2,042 South East 46 260 Central 62 176 South West 42 160 North West 50 135 Peten 37 519 Consumption Quintiles First 37 113 Second 46 153 Third 51 175 Fourth 51 220 Fifth 40 1,855 Ethnic Groups Non-Indigenous 42 543 Indigenous 47 172 Kiche 49 102 Qeqchi 28 126 Kaqchike 70 175 Mam 41 283 Other Indigenous 44 154 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. * For agricultural households only. 45 Table 20 : Vehicle Ownership for Households, Guatemala 2000 (Percentage of all households) Ownership Car/Automobile Pick- Small Truck Motorcycle Bicycle Small cart Boat Other Any Any up truck pulled by ox vehicles Motorized vehicle vehicle All 9.0 6.0 2.0 0.5 3.1 28.0 0.1 0.2 0.8 2.3 5.5 Urban 20.0 5.8 5.0 0.8 4.1 33.1 0.2 0.0 0.9 3.8 7.7 Rural 2.0 6.1 0.2 0.2 2.3 23.7 0.0 0.3 0.7 1.2 3.8 Poverty Group Non Poor 17.0 10.0 4.0 0.9 5.3 33.1 0.0 0.4 0.5 4.2 7.7 Poor 0.0 0.9 0.0 0.0 0.4 22.7 0.1 0.0 1.1 0.1 2.9 Extremely Poor 0.0 0.0 0.0 0.0 0.0 11.5 0.0 0.0 0.0 0 1.4 Regions Metropolitan 30.2 9.1 6.2 0.0 3.6 24.4 0.0 0.0 0.5 5.6 8.3 North 0.5 0.0 0.6 0.0 0.8 21.6 0.0 0.0 0.0 0.2 2.7 North East 0.5 4.1 0.4 0.7 2.3 34.5 0.0 1.9 4.5 1.0 4.3 South East 1.6 3.3 1.5 1.2 1.4 30.1 0.0 0.0 1.2 1 4.7 Central 4.9 3.1 1.4 1.2 2.2 40.8 0.0 0.0 0.0 1.3 6.3 South West 1.0 8.0 1.6 0.8 5.7 30.1 0.0 0.0 0.6 1.9 5.3 North West 1.5 4.3 0.0 0.0 1.3 11.3 0.0 0.0 0.0 0.7 2 Peten 3.9 3.8 0.0 0.0 0.9 46.8 1.8 0.0 0.0 1 6.2 Consumption quintile First 0.0 0.0 0.0 0.0 0.0 12.2 0.0 0.0 0.0 0 1.4 Second 0.0 2.0 0.0 0.0 1.0 24.1 0.4 0.0 0.0 0.3 3.2 Third 0.0 0.2 0.0 0.0 0.0 30.7 0.0 0.0 3.0 0 3.8 Fourth 1.1 7.9 0.2 1.4 3.6 22.6 0.0 0.9 0.0 1.7 4.1 Fifth 30.2 13.3 7.7 0.6 7.8 41.3 0.0 0.0 0.9 6.8 11.3 Ethnic Group Non-Indigenous 14.1 5.9 3.1 0.3 4.0 31.3 0.1 0.3 1.2 3.1 6.6 Indigenous 1.6 6.1 0.7 1.0 1.5 23.0 0.0 0.0 0.0 1.2 3.8 Kiche 1.9 7.6 2.1 2.3 0.0 26.8 0.0 0.0 0.0 1.6 4.8 Qeqchi 0.3 0.0 0.0 0.0 0.0 23.6 0.0 0.0 0.0 0.0 2.7 Kaqchike 0.0 8.1 1.1 1.5 0.7 31.0 0.0 0.0 0.0 1.3 5.1 Mam 0.0 11.9 0.0 0.0 4.4 18.1 0.0 0.0 0.0 2.0 4.1 Other Indigenous 4.5 2.0 0.0 0.0 1.2 11.6 0.0 0.0 0.0 0.9 1.9 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 46 Table 21: Commuting Time and Distance for Individuals (first job), Guatemala 2000. Population Population using the Population reporting the Population Average Working outside following as the main following number of reporting the commuting the municipio (to mode of transport to Stages to get to following Time (last which they work/Working work/Working frequency of week) belong) /Total population. population. commuting/ population Working population Walking Motorized 1 2 More 1 2 transport than 3 (%) (%) (%) (%) (%) (%) (%) (%) (Minutes.) All 28 75 18 72 15 13 84 15 42 Urban 49 67 32 59 21 20 83 15 42 Rural 14 79 12 83 11 6 84 14 43 Poverty Level Non Poor 44 61 30 60 22 18 82 16 41 All Poor 15 84 9 83 10 7 85 14 43 Extremely Poor 13 91 6 91 5 4 90 9 42 Regions Metropolitan 78 63 34 42 25 33 86 13 52 North 9 90 6 92 6 2 89 10 39 North East 8 69 17 78 17 5 78 19 39 South East 18 79 12 84 12 4 86 11 46 Central 18 68 21 76 14 10 81 17 40 South West 14 74 18 78 14 8 85 13 36 North West 11 92 5 90 6 4 77 21 43 Peten 6 66 10 93 5 2 87 12 44 Consumption Quintile First 13 91 5 91 5 4 89 10 41 Second 11 83 7 85 9 6 83 16 44 Third 24 72 18 70 18 12 82 17 43 Fourth 36 67 25 60 23 17 84 14 42 Fifth 57 53 42 57 20 23 80 17 39 Ethnic Groups Non-Indigenous 36 67 27 66 18 16 84 14 40 Indigenous 16 85 14 82 10 7 84 15 45 Kiche 20 77 22 74 19 7 84 15 35 Qeqchi 8 92 7 93 6 1 91 9 40 Kaqchikel 22 74 26 68 15 17 80 19 56 Mam 8 95 5 92 4 4 85 14 40 Other Indigenous 17 88 11 86 7 7 83 16 49 Note: These are figures reported for the first job, and the commuting figures are average over a week. Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. * Means of transport include public and private means. 47 Table 22: Physical Access to Markets, with and without Motorable Roads, Guatemala 2000. (Households for which PSU level information is available). With a Motorable road Without a Motorable Road Distance (kilometers) Time (minutes) Distance (kilometers) Time (minutes) All 8 49 12 79 Urban areas 2 24 1 31 Rural areas 11 59 15 94 Regions Metropolitan 4 38 2 19 North 9 83 6 87 North-east 10 40 13 46 South-east 10 42 26 142 Central 9 41 7 70 South-west 5 43 9 71 North-west 9 67 10 86 Peten 30 64 159 360 Source: World Bank calculationsusing ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 48 Table 23: Correlates of the Presence of Motorable Roads, Guatemala 2000 Water Wood Health Work Facility Distance Time Distance Time Time Time (Kmeters) (Mins.) (Kmeters) (Mins.) (Mins.) (Mins.) All 0.3 13 1.4 64 46 44 Urban areas 0.1 14 1 47 35 46 Rural areas 0.3 13 1.5 69 53 43 Poverty Levels Non-Poor 0.2 11 1.2 59 43 42 All Poor 0.3 14 1.6 67 51 45 Extremely poor 0.4 13 1.7 66 58 45 Regions Metropolitan 0.2 16 0.7 39 42 62 North 0.4 13 1.9 63 51 38 North-east 0.4 17 1.2 74 51 36 South-east 0.4 21 1.8 79 54 50 Central 0.1 8 1.6 56 40 37 South-west 0.3 10 1.2 64 46 34 North-west 0.3 13 1.7 76 43 44 Peten 0.2 13 1.8 67 59 45 Consumption Quintiles First 0.4 14 1.7 67 56 43 Second 0.3 17 1.6 74 49 45 Third 0.2 12 1.5 61 46 42 Fourth 0.2 12 1.1 59 50 42 Fifth 0.2 10 1.0 54 38 46 First 0.4 14 1.7 67 56 43 Ethnic groups Non-Indigenous 0.2 14 1.3 66 47 42 Indigenous 0.3 13 1.5 63 46 45 Kiche 0.6 12 1.3 50 38 32 Qeqchi 0.4 10 1.6 51 45 40 Kaqchikel 0.2 13 1.6 54 45 60 Mam 0.2 11 1.5 70 59 41 Other Indigenous 0.3 18 1.6 89 50 51 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. A health facility is described as a public hospital, hospital of the IGSS, a private hospital, a polyclinic of the IGSS, a health center, a health post, a community center, a private clinic or a private consultation, a private pharmacy or a state pharmacy. 49 Table 24: Time taken to Common Community Services if the Community has a Motorable Road (Minutes), by households, Guatemala, 2000 Phone Post Inter- Urban Bus Bank Cooperative Police Fire S tation Office Urban Bus Stop station stop All 24 40 26 19 41 41 36 38 Urban areas 12 23 17 14 21 17 14 17 Rural areas 29 47 31 24 52 49 46 49 Regions Metropolitan 12 38 20 12 31 39 24 26 North 35 47 47 11 58 39 51 54 North-east 22 28 27 17 32 28 24 44 South-east 24 36 20 18 44 38 37 48 Central 26 40 26 47 40 38 33 36 South-west 19 37 21 15 40 41 38 36 North-west 43 56 38 58 56 44 50 37 Peten 37 48 27 18 56 95 47 91 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. Table 25: Distance to Common Community Services if the Communityhas a Motorable Road (Kilometers), for households, Guatemala, 2000 Phone Post Inter- Urban Bus Bank Cooperative Police Fire Station Office Urban Bus Stop station stop All 4 9 3 4 11 10 8 13 Urban areas 2 3 2 1 5 4 3 8 Rural areas 5 11 4 6 14 12 11 16 Regions Metropolitan 3 7 3 3 6 7 4 9 North 7 13 5 1 14 11 12 16 North-east 2 8 3 2 10 9 7 15 South-east 2 7 2 4 11 9 8 15 Central 3 8 3 3 10 9 8 10 South-west 3 5 4 2 10 9 6 11 North-west 6 11 5 22 14 11 11 25 Peten 9 30 5 16 34 30 29 41 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 50 Table 26: Correlates of the Absence of Motorable Roads, Guatemala 2000 Water Wood Health Work Facility Distance Time Distance Time Time Time (Kmeters) (Mins.) (Kmeters) (Mins.) (Mins.) (Mins.) All 0.3 15 1.32 61 59 43 Urban areas 0.04 35 0.99 44 35 36 Rural areas 0.3 13 1.39 64 69 45 Poverty Levels Non-Poor 0.07 21 0.94 51 48 45 All Poor 0.4 13 1.44 64 65 42 Extremely poor 0.5 16 1.75 67 68 33 Regions Metropolitan 0.02 2 0.73 27 28 53 North 0.5 13 1.96 73 97 41 North-east 0.1 7 1.10 31 72 34 South-east 0.07 6 1.2 67 71 44 Central 0.03 47 1.05 46 40 45 South-west 0.4 13 1.13 59 53 49 North-west 0.3 19 1.53 78 41 38 Peten 0.04 13 0.45 47 75 17 Consumption Quintiles First 0.4 16 1.74 65 67 35 Second 0.3 11 1.34 62 56 46 Third 0.2 9 1.08 70 69 51 Fourth 0.1 9 0.96 43 54 48 Fifth 18 46 541 44 44 44 Ethnic groups Non-Indigenous 0.2 15 1.04 49 54 41 Indigenous 0.4 14 1.53 70 66 44 Kiche 0.2 15 2.23 67 46 50 Qeqchi 0.5 13 1.97 70 100 42 Kaqchikel 0.2 13 1.04 65 65 73 Mam 0.3 16 1.19 60 66 32 Other Indigenous 0.2 14 1.60 79 37 33 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 51 Table 27: Time taken to reach Common Community Services if the Community does not have a Motorable Road (Minutes), for households, Guatemala, 2000 Phone Post Office Inter- Urban Bus Bank Cooperative Police Fire Station Urban Bus Stop station stop All 47 71 61 60 58 65 71 50 Urban areas 17 18 16 31 42 25 27 34 Rural areas 57 87 75 107 65 76 85 56 Regions Metropolitan 12 21 13 71 20 18 25 22 North 62 82 88 23 91 75 94 1 North-east 16 19 37 27 21 20 22 24 South-east 80 134 60 156 118 127 126 127 Central 36 36 22 80 94 40 59 43 South-west 46 73 68 5 41 80 73 24 North-west 53 78 71 34 42 44 69 37 Peten 201 290 230 - 350 350 350 420 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. Table 28: Distance to Common Community Services if the Community does not have a Motorable Road (Kilometers), for households, Guatemala, 2000 Phone Post Office Inter- Urban Bus Bank Cooperative Police Fire Station Urban Bus Stop station stop All 7 12 6 7 15 13 13 18 Urban areas 0 1 1 0 2 3 1 3 Rural areas 9 15 8 18 20 15 17 24 Regions Metropolitan 0 2 1 0 4 1 4 3 North 7 9 7 1 13 6 13 1 North-east 6 13 9 12 15 9 10 10 South-east 10 24 3 30 34 27 32 31 Central 3 7 2 1 8 10 7 17 South-west 5 8 6 0 7 9 9 9 North-west 10 13 6 3 17 12 13 22 Peten 85 135 85 - 185 185 185 219 Source: World Bank calculations using ENCOVI 2000, Instituto Nacional de Estadistica, Guatemala. 52 C:\DOCUME~1\wb17071\LOCALS~1\Temp\paper2wopics 20 May 2003.doc November 13, 2003 2:25 PM 53