Report No: ACS26526 . World Measuring Rural Access: Update 2017/18 . February 2019 . GTR01 OTHER . . . -2- Standard Disclaimer: . This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Measuring Rural Access: Update 2017/18 -4- CONTENTS Acknowledgment ...................................................................................................................... 5 I. Introduction ....................................................................................................................... 6 II. Methodology ...................................................................................................................... 7 III. The Results: Update........................................................................................................... 8 IV. Country Notes .................................................................................................................. 12 A. Armenia .................................................................................................................... 12 B. Burundi ..................................................................................................................... 14 C. Lesotho ..................................................................................................................... 16 D. Liberia ...................................................................................................................... 18 E. Madagascar............................................................................................................... 20 F. Malawi ......................................................................................................................... 22 G. Mali .......................................................................................................................... 24 H. Nigeria ...................................................................................................................... 26 I. Peru .............................................................................................................................. 28 J. Rwanda ........................................................................................................................ 30 K. Sierra Leone ............................................................................................................. 32 L. Somalia ..................................................................................................................... 34 M. Iraq ........................................................................................................................... 36 N. Jordan ....................................................................................................................... 38 O. Lebanon .................................................................................................................... 40 References ............................................................................................................................... 42 Appendix. RAI estimated at sub-national level ...................................................................... 43 -5- Acknowledgment This update was prepared by Transport Global Practice under the supervision of Jose Luis Irigoyen, Senior Director, and Aurelio Menendez and Maria Marcela Silva, Practice Managers. The report was prepared by a World Bank team comprised of Atsushi Iimi (Task Team Leader), Adam Stone Diehl, Laban Maiyo, Muneeza Mehmood Alam, Fernanda Ruiz Nunez, Tatiana Peralta-Quirós and Farhad Ahmed, in collaboration with Steven Farji Weiss, Noroarisoa Rabefaniraka, Sevara Melibaeva, Rajesh Rohatgi, Kulwinder Singh Rao, Cheick Omar Tidiane Diallo, Aiga Stokenberga, Li Qu, Ben Gericke, Olatunji Ahmed, Ana Rodriguez Coteron, Stephen Muzira, Sofia Guerrero Gamez, Ana Silvia Aguilera, Muhammad Zulfiqar Ahmed, Emmanuel Taban, Heinrich Bofinger, James Markland, Haileyesus Mengesha, Mira Morad, Ibrahim Dajani and Ziad Nakat on country-specific issues. The work also benefited from support probided by Almud Weitz, Olivier Le Ber, Shomik Raj Mehndiratta, Benedictus Eijbergen, Nicolas Peltier-Thiberge, Marianne Fay, Neil James Fantom, Umar Serajuddin, Buyant Erdene Khaltarkhuu, Simon Ellis, Nancy Vandycke, Holly Krambeck and Wei Winnie Wang. The team would also like to acknowledge the generous cooperation of the Governments of Armenia, Burundi, Iraq, Jordan, Lebanon, Lesotho, Liberia, Madagascar, Malawi, Mali, Nigeria, Peru, Rwanda, Sierra Leone, and Somalia, as well as other organizations, such as the United Nations, which provided necessary data for the analysis. -6- I. Introduction 1. Transport connectivity is an essential part of the enabling environment for inclusive and sustained growth. In many developing countries, particularly in Africa, the vast majority of people are still not connected to local, regional, or global markets. Such rural accessibility is crucial to reduce poverty and promote inclusive economic growth. The Sustainable Development Goals (SDGs) aim to build resilient infrastructure, promote inclusive and sustainable industrialization, and foster innovation (Goal 9), for which Target 9.1 is to “develop quality, reliable, sustainable and resilient infrastructure…to support economic development and human well-being, with a focus on affordable and equitable access for all.� The Rural Access Index (RAI) was proposed and accepted as an indicator to measure this target. 2. The RAI is one of the most important global indicators in the transport sector. It measures the proportion of people who have access to an all-season road within an approximate walking distance of 2 kilometers (km). There is a common understanding that the 2 km threshold is a reasonable extent for people’s normal economic and social purposes. The definition is also simple enough to understand and use not only in the transport sector, but also in the broader development context, such as poverty alleviation. The initial RAI study in 2006 was based on household surveys and other simplified methods,1 estimating the global index at 68.3 percent, leaving a rural population of about one billion disconnected around the world (Figure 1). Figure 1. Rural Access Index, 2006 1 Roberts, Peter, K. C. Shyam, and Cordula Rastogi. 2006. “Rural Access Index: A Key Development Indicator.� Transport Papers TP-10. The World Bank Group, Washington, DC. -7- II. Methodology 3. The World Bank partnered with the Department for International Development (DFID) of the United Kingdom and the Research for Community Access Partnership to develop a new methodology to measure rural access, one which is sustainable, consistent, simple, and operationally relevant. Conceptually, the methodology is still focused on access to an all- weather road, but uses technically more objective, common parameters, such as roughness of the road and visual assessment. It measures the share of the population that lives within 2 km of an all-season road or an equivalent road as a proxy, for instance, a road in good condition, in rural areas. For more technical details, see World Bank (2016).2 4. This new methodology takes advantage of spatial techniques and data collected using innovative technologies. In recent years, various new technologies and data sets have been developed. For instance, high-resolution population distribution data have been developed by the international research community. The WorldPop data have the highest resolution (100 meters). Therefore, it is more or less known where people live. Digitized road alignment data, including road conditions, are also available at road agencies or in the public domain (such as OpenStreetMap). 5. Additional data required for the RAI calculation are minimal. Road condition data are needed at each road segment level. Many road agencies often possess their own road asset management systems, which provide detailed road condition data. There are a variety of ways of collecting such data. In traditional road inventory surveys, vehicle road profilers are used. In recent years, smartphone applications have been developed to assess road roughness while driving. Some other technologies, Figure 2. Spatial Technique for the New Rural such as high-resolution satellite Access Index Method imagery, also have the potential to evaluate road conditions in mass. Crowdsourced or open data may be particularly attractive from a data sustainability point of view. 6. All these potential data sources can be used to examine and monitor road conditions (also see World Bank (2016) for the discussion on alternative data sources). Available data may not always perfect: some data may only be partial with unofficial roads and paths excluded. 2 World Bank. (2016). Measuring Rural Access: Using New Technologies. -8- Others may be more comprehensive but lack information on road condition. It is essential to agree on how to measure road conditions given available data and country-specific circumstances. 7. By spatially combining (i) global population distribution data, (ii) geo-referenced road alignment data, and (iii) road condition data, the RAI is virtually computed by spatial software (Figure 2). The methodology ensures accuracy, consistency and sustainability. The new method also allows the index to be computed at any subnational level (such as districts or villages). The index is no longer one national number but often highlights significant inequality in rural accessibility across areas. It is intended to be used in actual road sector operations, for example, in the rural road prioritization context. III. The Results: Update 8. In the first phase report prepared in Figure 3. Comparison of Original and New RAI 2016, the new method was applied to Results eight pilot counties: Ethiopia, Kenya, Mozambique, Tanzania, Uganda, and Zambia in Africa, and Bangladesh and Nepal in South Asia. Rural access was found to vary significantly across these countries, from 17 percent in Zambia to 87 percent in Bangladesh (Figure 3). In total, it is estimated that about 34 percent of the rural population is connected, with roughly seven million people left disconnected.3 9. The results turned out to be comparable to the 2006 initial estimates in some countries, and significantly different others. This is because the methodologies are different between the two studies, and particularly because the 2006 study did not use actual data on the ground but relied on a model approach comparing similar countries. 10. The same methodology has now been applied to 15 additional countries, in collaboration with the national road agencies. The estimated RAI varies from 11.4 percent in Madagascar to 68 percent in Armenia and 93 in Lebanon (Figure 4). The new RAI are broadly consistent with the original estimates from 2006, with the majority of countries finding a decrease in access. Again, it is noteworthy that it is difficult to compare them directly. In addition, the underlying data for the new estimates still may or may not be perfect. For Nigeria, for instance, the RAI is calculated based on about 107,000 km of roads of which road condition data are available. This represents only half of the total road network according to the 3 See World Bank. (2016). Measuring Rural Access: Using New Technologies. -9- national statistics. Malawi and Somalia also have the same issue. See the attached country notes for country-specific technical issues. Figure 4. Original and New RAI Estimates for Additional 15 Countries 11. At the cross-country level, the new estimates seem to be consistent with available macroeconomic data. Not surprisingly, for instance, poverty is high where rural accessibility is limited (Figure 5). The correlation coefficient is about -0.627. 12. There is significant inequality within each country (Figure 6). One of the advantages of the new spatial method is that accessibility can be computed at the subnational level. In Kenya, Uganda and Rwanda, rural connectivity is systematically high along the Northern Corridor. Similarly, rural access is relatively high along the North-South Corridor, connecting Dar es Salaam to Mbeya, Malawi and Lusaka, Zambia. Apart from the areas along the regional corridors, rural accessibility tends to be limited in Africa. 13. Strong correlation between poverty and rural accessibility can also be observed at the subnational level. When using data from Madagascar, Mozambique, Kenya, Rwanda, Uganda and Zambia where recent district-level poverty data are available, the correlation coefficient is -0.619 (Figure 7). The evidence supports proposed approaches by which the RAI can be used to plan and prioritize rural road investments at the subnational level. Figure 5. RAI and poverty headcount - 10 - Figure 6. Rural Access Index at the Subnational Level (Eastern and Southern Africa) (Western Africa) - 11 - Figure 7. RAI and poverty headcount at subnational levels - 12 - IV. Country Notes A. Armenia Overview of road network Basic Data Population 1 2.90 million (2015) 14. Armenia has a relatively well-developed Land area 1 29,740 km2 (2015) 1 Population density 97 per km2 (2015) road network, serving all areas of its Length of road 2 7,700 km (2015) economy. Most freight and passenger traffic is Paved road (%) 2 84% (2017) carried by road. The road network is 7,700 km Length of road (GIS) 3 8,280 km (2017) long with 1,400 km of interstate roads, 2,520 Of which, “Good quality 3,390 km (2017) road� km of regional roads and 3,780 km of local Of which, road quality data 30 km (2017) roads. The surface condition of these roads are missing varies from good to fair. According to a recent RAI 3 66 % (2017) 1 World Development Indicators road survey, close to 41 percent of roads are 2 UNOPS, EU & AfDB (2015) in good or very good condition, 19 percent are 3 World Bank estimates based on a road inventory survey in fair condition, and 40 percent are in poor or very poor condition. 15. Because of its difficult terrain, the road system of the country is of vital importance. Most of the road network was built in the 1960s and 1970s, and the majority of republican and local roads have deteriorated since independence. The roads linking villages to the main highways are often called "lifeline roads" in Armenia. They are vital for the communities, located dozens and hundreds of kilometers away from urban areas. With a significant part of them last rehabilitated in the Soviet era, the lifeline roads have deteriorated in the intervening years with many now in desperately poor condition, effectively cutting off rural communities from near-by towns and big cities. 16. To update the road network condition, a road inventory survey using a smartphone application was carried out in the entire country during May, 2016. In total, 8,280 km of roads were surveyed. While more than 90 percent of inter-state highways are in good or fair condition, nearly 60 percent of unpaved roads are in poor or very poor condition. 17. The road network density of Armenia is 279 km per 1,000 square km, which is low compared to other countries in the region, reflecting in part the difficulties to provide basic access to the rural population. Despite of the Government’s recent efforts, e.g., the Lifeline Road Improvement Project in 2009 and the Lifeline Roads Network Improvement Project approved in 2012, road quality remains a matter of concern. The lack of maintenance during the long civil war has damaged the road network significantly. While the paved road network is relatively well maintained, the condition of unpaved roads is mostly poor. Classification and standards - 13 - 18. While 83 percent of interstate roads are in good or very good condition, evidence of the relative importance that the government places on the maintenance of the long-haul and inter-city networks, only 36 percent of secondary and tertiary roads are in such condition. This compares favorably with other countries in the region such as Albania or Georgia, where less than 20 percent of rural roads are in good condition, but still lags behind Macedonia and Serbia. Data issues and assumptions 19. A country-wide road survey was carried out to obtain geo-referenced data of the entire road network of Armenia from March to May 2017. The survey used a smartphone application, RoadLabPro, collecting road surface type, category and road condition data, as well as geo-tagged information of service and market location. The survey result is assessed based on the observed IRI and type of pavement: For paved roads, a road is considered as very good when 1 7.0 Unpaved < 4.0 4.0 - 5.0 5.0 - 9.0 9.0 - 16.0 > 16.0 Source: Ministry of Public Works. 42. For population data, the WorldPop data (2010 edition) is used, and urban areas are defined based on the 1995 University of Columbia (CIESIN) urban area imagery. These data give a rural population estimate of about 3.9 million in Liberia. Estimated Rural Access Index 43. Given the current road condition, the RAI is estimated at 41.9 percent. While about 1.6 million rural people have access to the road network in good condition, about 2.3 million are left unconnected. The estimated RAI is lower than the previous estimate in 2006, which was 66 percent, although the two results cannot be compared directly due to methodological differences. 44. Rural accessibility differs among counties and districts. RAI is high around Monrovia and along the Monrovia-Ganta Corridor. Although urban areas are excluded from the RAI calculation, the district of Greater Monrovia has a RAI of 94.6 percent. For many districts in Montserrado, Bong and Nimba Counties, RAI is estimated at 30 percent or higher. In many other districts, accessibility is less than 10 percent. For some districts, the RAI is zero, meaning that there are no roads in good condition in the district. - 20 - E. Madagascar Overview of road network Basic Data Population 1 24.9 million (2016) Land area 1 581,800 km2 (2016) 45. In Madagascar, limited transport Population density 1 42.8 per km2 (2016) connectivity is a common constraint across all Length of road 2 31,640 km (2015) sectors. The country possesses important Paved road (%) 2 17.8 % (2015) transport infrastructure, including roads, Length of road (GIS) 2 21,640 km (2017) Of which, “Good quality 7,042 km (2017) railways and ports. However, their quality is road� generally poor due to past underinvestment Of which, road quality data 0 km (2017) and under-maintenance. In the road sector, are missing RAI 3 11.4 % (2017) official statistics account for about 32,000 km 1 World Development Indicators of classified roads, which translates into a 2 Autorite de Routier Madagascar road density of 5.4 km per 100 km2, low 3 World Bank estimates based on a road inventory survey compared to other countries in the region. 46. The vast majority of non-primary roads (i.e., provincial and community roads) need to be repaired and rehabilitated. Paved national roads (about 5,600 km or 18 percent of the total network) are relatively well maintained. However, about two-thirds of secondary and tertiary roads are estimated to be in poor condition. Classification and standards 47. The classified roads comprise about 11,890 km of national roads, 12,250 km of provincial roads and 7,500 km of community roads. The Ministry of Public Works and the Madagascar Road Authority are jointly responsible for the national road network, which is composed of 2,563 km of primary national roads (RNP), 4,784 of secondary national roads (RNS), and 4,543 of temporary national roads (RNT). 48. While provincial governments manage the country’s provincial roads (12,250 km), the communes are responsible for provisionally 7,500 km of communal roads. Most of the provincial and communal roads are in poor condition. In addition, there are presumably a number of unclassified feeder roads in rural areas, which are also in very poor condition. - 21 - Data issues and assumptions 49. To update the road network condition, a road inventory survey using a smartphone application was carried out covering the entire country in May 2017. In the survey, the road conditions are classified based on measured roughness, i.e., international roughness index (IRI). The thresholds for condition classification were decided based on an actual field experiment in Madagascar. Excellent Good Fair Poor Paved <1 1-2 2-4 >4 Unpaved <3 3-5 >5 Source: Madagascar Road Authority. 50. For population data, the WorldPop data (2010 edition) is used, and urban areas are defined based on the 1995 University of Columbia (CIESIN) urban area imagery. These data give a rural population estimate of about 19.1 million in 2015. Estimated Rural Access Index 51. The RAI is estimated at 11.4 percent, leaving 16.9 million people unconnected in rural Madagascar. This is lower than the previous estimate in 2006, which was 25 percent. The current estimate is considered to be more accurate because it is based on actual road and population data. It seems to be consistent with the fact that road density is extremely low in Madagascar. Population density is also low, and the road condition is generally poor. 52. Rural accessibility differs significantly between semi-urban areas and the rest of the country. RAI is estimated at greater than 80 percent in Toamasina, Mahajanga and Antsiranana Districts. In most rural districts, rural accessibility is less than 5 percent. - 22 - F. Malawi Overview of road network Basic Data Population 1 18.6 million (2017) 53. According to the Government statistics, Land area 1 94,280 km2 (2017) Malawi has a well-established road network Population density 1 197.6 per km2 (2014) Length of road 2 15,451 km (2014) comprising 15,451 km of classified roads Paved road (%) 2 26.4 % (2014) (main, secondary, tertiary, district, and urban Length of road (GIS) 3 12,859 km (2016) designated roads) and some 9,000 km of Of which, “Good quality 3,441 km (2016) road� unclassified roads. This translates into a Of which, road quality data 1,244 km (2016) relatively high road density of 14 km per 100 are missing km2, favorably compared with its neighboring RAI 3 23.1 % (2016) 1 World Development Indicators countries (e.g., 5.5 km in Tanzania and 3.9 km 2 Road Authority, Malawi. in Mozambique). The current network is 3 World Bank estimates based on a road inventory survey considered to be sufficient to serve the entire country, estimated to cover about 80 percent of the country’s total population. However, the quality of the road network, especially secondary and tertiary roads, remains largely poor. According to the latest official road survey conducted in 2011, paved roads were mostly in good or fair condition, but about 33 percent of unpaved roads were in poor condition. Based on a more recent condition survey in 2016, of which the results remain to be validated, the rural road condition seems to have been deteriorated: 64 percent of the unpaved road network surveyed is in poor condition. Classification and standards 54. In Malawi, about 4,000 km or 26 percent of the total road network is paved, and the rest remain unpaved and in largely poor condition. The Malawi Roads Authority (RA) is a quasi- government body, established in 2006, which is responsible for constructing, rehabilitating and maintaining public roads under the Ministry of Works and Public Infrastructure. While the RA’s primary responsibility is for Main (3,357 km), Secondary (3,125 km) and Tertiary Roads (4,121 km), the responsibility for other types of roads are delegated to local governments. In the country there are 3,500 km of District Roads and some 9,000 km of Community Roads, which are currently unclassified. 55. In the 2014/15 revised budget, the Roads Fund Administration (RFA) allocates MWK3.3 billion or 0.3 percent of GDP to road - 23 - maintenance by the RA and local governments. A fuel levy accounts for about 90 percent of the RFA’s total revenue. While most of the Main Roads managed by the RA are paved and well maintained, other minor roads, such as District and Community Roads, tend to be underfunded. They are largely in poor condition and need to be rehabilitated. Data issues and assumptions 56. The road condition data in Malawi is fragmented. The latest comprehensive official road condition survey was carried out in 2011, though some assessments focused on primary roads are carried out every year. A quick road condition survey using a smartphone application, which measures road roughness while driving with it, was also carried out in 2016. The coverage is not complete but about 90 percent of the classified road network excluding urban roads. The survey was only focused on non-urban areas. 57. Paved roads are classified into 4 categories based on roughness measured: Excellent (IRI<2), Good (26), out of which the first three are considered for the index calculation. For unpaved roads, average speed is used as a proxy: Excellent (speed>80km/h), Good (70