Technical note

1. Introduction

Access to infrastructure is one of the biggest constraints to development in emerging and developing countries. Low access to critical infrastructure such as water and sanitation, electricity, and roads not only affect the socioeconomic wellbeing of households, but also the productivity of firms. Inequality in the distribution of infrastructure services, uneven quality, and affordability of services remain as major issues. A significant share of urban population in MICs face severe constraints in accessing quality and affordable infrastructure and critical services. However, data on the extent of the access, quality and affordability gaps are lacking or only available at the country level. Granular data at the subnational level is required to inform analysis and decision making.

The main objective of this dashboard is to facilitate access to a unique data platform that contains subnational level data on infrastructure access, quality, and affordability gaps, with particular emphasis on MICs where IFC operates. This information was collated and analyzed applying a consistent methodology to estimate access, quality, and affordability of infrastructure at subnational level. The purpose of the dashboard is to: (i) inform business development, impact assessment and strategy formulation at IFC, (ii) facilitate benchmarking within IFC AIMM Frameworks, (ii) set foundations to build a global database useful for policy, investment, and analytics.

2. Data sources and country coverage

The information displayed in the dashboard comes three sources: (i) harmonized household survey and census data from the World Bank (WB) Global Monitoring Database (GMD); (ii) Spatial data on road networks from Global Roads Inventory Project; and (iii) NASA’s Gridded Population of the World.

• The WB GMD: is a platform that gathers and harmonizes household surveys and census data from around the world into single platform to allow comparability and monitoring of key socioeconomic trends around the world. The database essentially, “constructs globally comparable microdata across countries, regions and across years for global poverty monitoring and welfare measurement” (World Bank, 2019). The database is accessible to WBG staff via Datalibweb 1. Data on measures relating to water, sanitation, electricity, cooking and heating, housing, digital technologies, household assets (ownership of electric appliances, car, etc.) are drawn from the GMD into the constructed database. Table 1 in the annex, provides a list of the main indicators extracted from GMD 2

• Global Roads Inventory Project (GRIP): GRIP is a global database on road networks of 222 countries.3 it' was constructed by consolidating spatial data collected from sources including multigovernmental organizations, NGOs, research institutes, governments and crowdsource initiatives (Meijer et al., 2018). It classifies roads by road type and surface pavement attributes. Road type categories are highways, primary roads, secondary roads, tertiary roads, local roads and unspecified. Road surface categories are paved, gravel, dirt/sand, steel, wood, grass and unspecified. Road surface categorization helps to differentiate roads environmental impact and seasonality because, for instance, highways have higher impact on the environment but all-season usability than unpaved roads.

• Population data: The population data comes from Gridded Population of the World, version 4 (GPWv4) produced by NASA’s Socioeconomic Data and Applications Center (SEDAC). he extracted data in this project is population count for 2015. GPWv4 provides spatial explicit information on distribution of human population across the globe at a 30 arc second (approximately 1km at the equator) resolution. The georeferenced population data can be easily merged with georeferenced socioeconomic data and administrative boundaries. The essence of including the population data in the database is to allow users get a sense of the number of people living in the respective subnational regions and also assess the potential demand gap with respect to the various infrastructure availability.

3. Country coverage:

The dashboard contains information covering 1,410 subnational districts/regions from 107 countries.

4. Methodology

Household (micro) data on infrastructure access: For each infrastructure group (i.e., sanitation, water, electricity, road, housing, digital technologies,) the data is categorized in three types of indicators: access, reliability of service, and affordability (proxied by expenditure) by households. The primary information is obtained from GMD household surveys data around the world, and to ensure consistency, the dashboard presents latest available household survey in the database as of 2022. Table 1 presents information on the survey name and year in from which the subnational infrastructure data were computed. For 68 (64%) countries in the database, the indicators are calculated on surveys collected between 2015 and 2020.

To ensure that the estimates of infrastructure access at the subnational level is representative of the population, the approach of the World Bank’s Global Subnational Atlas of Poverty (GSAP)4 adopted, matching the households (survey respondents) to the subnational districts. GSAP has produced the list of subnational boundaries for each survey country for which the household surveys are representative. In some countries these boundaries matches either (i) first administrative boundary; (ii) second administrative boundary; or a combination of both. For each of these subnational units, the population weighted average access/uptake rate for the respective infrastructure services is calculated. In addition, infrastructure access across heterogenous groups (e.g., rural vs. urban, welfare deciles) within the subnational is also computed. These classifications allow to assess the extent of inequalities in access to the various infrastructure across groups.

Roads: For each subnational unit and road class, the total road length (km) and density (km/km2) for the various road classes is extracted and used 5.

Population: For each subnational unit, the total population and the population density (per square km) is used for calculations.

Indicators at national and subnational level: indicators on access, reliability of service, and affordability are calculated at the subnational level. The database also presents the same indicators at the national level, based on national averages for ease of comparison.

Annex

Table 1: Indicator list and definitions

Table 2: Country, Survey Year, and Number of Subnational Regions

1 https://worldbank.github.io/SARMD_guidelines/datalibweb-getting-access-to-sarmd.html

2 Note that subcomponents of these indicators, e.g. electricity access in urban areas in the subnational, are also included in the final database.

3 https://www.globio.info/download-grip-dataset

4 https://thedocs.worldbank.org/en/doc/122521616013871062-0090022021/original/GMGD2019documentationnotev02.pdf

5 The dplyr and sf packages in R stat were used in the extraction.