(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Mapping urban living standards and economic activity in developing countries with energy data [1]

['Felix S. K. Agyemang', 'Department Of Planning', 'Environmental Management', 'University Of Manchester', 'Manchester', 'United Kingdom', 'Rashid Memon', 'Social', 'Economic Survey Research Institute', 'University Of Qatar']

Date: 2023-11

Urban data deficits in developing countries impede evidence-based planning and policy. Could energy data be used to overcome this challenge by serving as a local proxy for living standards or economic activity in large urban areas? To answer this question, we examine the potential of georeferenced residential electricity meter data and night-time lights (NTL) data in the megacity of Karachi, Pakistan. First, we use nationally representative survey data to establish a strong association between electricity consumption and household living standards. Second, we compare gridded radiance values from NTL data with a unique dataset containing georeferenced median monthly electricity consumption values for over 2 million individual households in the city. Finally, we develop a model to explain intra-urban variation in radiance values using proxy measures of economic activity from Open Street Map. Overall, we find that NTL data are a poor proxy for living standards but do capture spatial variation in population density and economic activity. By contrast, electricity data are an excellent proxy for living standards and could be used more widely to inform policy and support poverty research in cities in low- and middle-income countries.

Data Availability: The data underpinning the paper has been made available. Four datasets were used, including, household income and expenditure survey data (HIES), electricity consumption, nighttime lights (NTL), and OpenStreetMap (OSM). These can be accessed using the links below. HIES: https://www.pbs.gov.pk/content/microdata Electricity consumption: https://urldefense.com/v3/__ https://reshare.ukdataservice.ac.uk/856294/__;!!PDiH4ENfjr2_Jw!GYltBYJ3i-1Y489omSPD1XAdX8Q7M_PL_VG6LT9ZODmneA5sEYdmLdVUst6eMuGTEVEdavzygHQdjFYl6xdtosE4htz48ICIdsR5ow$ NTL: https://eogdata.mines.edu/products/vnl/ OSM: https://download.geofabrik.de/asia.html .

Copyright: © 2023 Agyemang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

I. Introduction

Rapid urban population growth in low and middle-income countries (LMICs) is contributing to the ‘urbanization of poverty’ [1]. Yet the true scale and nature of this challenge is unknown as we generally lack accurate and up-to-date data about household conditions in most cities in LMICs. Census data are infrequent and often undercount the urban poor [2–4], and traditional household surveys (e.g. Demographic and Health Surveys, or Living Standards Measurement Study data) generally do not have large enough samples to provide representative data for individual cities, let alone smaller areal units such as neighbourhoods. These traditional data sources are therefore of limited use in rapidly growing cities where policymakers need timely and spatially explicit information to support decision-making and efficient resource allocation. Given that virtually all projected population growth in coming decades will be absorbed by urban areas in low- and middle-income countries [5], urban data deficits are a significant obstacle to global poverty alleviation and development efforts [6].

Over the past decade, development economists have increasingly turned to non-traditional data sources, such as mobile phones, satellite imagery and volunteered geographic information. Of these, analysis of publicly available night-time light (NTL) emissions data captured by satellites has become the most popular means of estimating sub-national variation in economic activity and poverty in the absence of suitable administrative or survey data. It has long been recognised that energy consumption and economic development go hand-in-hand at the country level [7], and this simple insight inspired the use of NTL emissions as a proxy for economic activity in countries with limited data.

However, research to date has generally focused on the potential for NTL to capture production and income at regional scales. Yet there is also a well-established relationship between energy consumption and household welfare. Indeed, access to modern energy—i.e. energy from clean, affordable and renewable sources, such as electricity—is itself a key indicator of development [8–10], and was included as Goal 7 of the UN Sustainable Development Goals. Household energy consumption is closely associated with traditional measures of living standards, including income and consumption. While there is ongoing debate regarding the extent to which improved access to modern energy can drive improvements in household living standards [11–15], there is abundant evidence from experimental and observational studies that increases in income and assets are associated with increased total energy consumption (from all fuel types) at the household level. Importantly, rising living standards are associated with increased consumption of electricity in particular as households move up the ‘energy ladder’ (away from traditional biomass fuels) and diversify their ‘energy portfolios’ [16].

For example, a randomized controlled trial in rural India found that cash and asset transfers increased total household fuel consumption and specifically electricity for lighting [17]. Similarly, [18] used a Regression Discontinuity Design to measure the causal impact of a cash transfer program on fuel choice and expenditure in Pakistan and found a significant impact on monthly per capita fuel expenditure. These findings are consistent with observational studies based on survey data in wide range of urban contexts, including Brazil [19], China [20, 21], Ghana [22], India [23], Mexico [24] and South Africa [25]. The link between income and electricity consumption (as opposed to other fuels) is particularly strong in urban areas where access is near universal: globally, an estimated 97% of urban residents have access to electricity [10].

The strength of the income-energy consumption association varies by country and context (weather and socio-cultural factors are also significant) but is highly consistent within each context. Electricity consumption data may therefore be a useful proxy for household living standards in urban areas in the absence of detailed data on household income, assets or expenditure. It has the added advantage of being measurable in near real-time, which may be particularly useful for policymakers seeking information about the household-level impacts of economic shocks in LMICs.

There are two primary channels linking income to residential electricity consumption, both mediated by household appliances—particularly those associated with lighting, cooling, heating, washing, and refrigeration, which constitute a substantial proportion of total household electricity consumption. Rising income increases electricity consumption (a) directly due to greater use of existing appliances in the home (intensive margin), and (b) indirectly through the adoption and use of new appliances (extensive margin). Across countries there is roughly an S-shaped relationship between income and ownership of energy-consuming appliances, with low levels of ownership among at the bottom end of the income distribution, a sharp increase at context specific thresholds and a plateau in the upper deciles [26]. In low- and middle-income countries, the adoption and use of new appliances is a strong indicator of improvements in household living standards and can be observed indirectly with energy consumption data.

Could nightlights serve as a proxy for living standards as well as economic activity in cities in LMICs? Previous research has confirmed a strong correlation between NTL emissions and GDP at various spatial scales [27–30] and has used NTL emissions to produce subnational estimates of the incidence of poverty [31–33]. However, the accuracy of NTL predictions is highly dependent on both the context and scale of application [28, 34]. The extent to which such data could be used to identify variation in living standards or economic activity within cities in LMICs remains unclear [35]. We tackle the question in the context of Karachi, Pakistan using high resolution data from residential electricity meters. First, we demonstrate empirically that household electricity consumption is a strong proxy for living standards in Pakistani cities.

We find a modest association between NTL emissions and median monthly electricity consumption, but a visual inspection shows that even high-resolution NTL data disguise substantial spatial heterogeneity in living standards across the city and can produce highly misleading results. In contrast, proxies for population density, ‘establishment density’, social infrastructure, and road density explain over half of the observed spatial variance in NTL emissions. These results are consistent with those of Mellander et al (2015), which used highly detailed demographic and administrative data to test these associations in the Swedish context.

Our analysis makes two key contributions to the literature. First, we demonstrate that electricity data can be used to generate high-resolution areal estimates of living standards in urban areas in the absence of traditional sources such as census or administrative data. These small area estimates could be used to significantly improve the efficiency of social policy targeting in cities [36]. Second, we show that even relatively high-resolution night lights data (VIIRS) on their own are not suitable for estimating intraurban variation in living standards, even in very large cities. However, they do capture information on spatial variations in population density, infrastructure density and economic activity.

Georeferenced electricity data are not as readily available as NTL data and must be treated with care due to privacy concerns, but they do exist in all cities and should be available to the governments that either own or regulate supply networks. Given high rates of electricity access in urban centres, such data could serve as a valuable tool for policy makers and researchers concerned with understanding and raising urban living standards. We also show that volunteered geographic information, such as OSM data, may have useful modelling applications in cities with scarce data.

In the next section we present the methodology underpinning the paper. This is followed by results and discussion, structured into three subsections. The first subsection presents original analysis of nationally representative household survey data to empirically establish a close association between household electricity consumption and living standards in Pakistan and demonstrates the potential use of electricity data for mapping urban living standards. The second subsection compares NTL emissions and electricity consumption in Karachi using gridded data. The last subsection models NTL emissions at the cell level as a function of proxy measures for economic activity. Section IV concludes.

[END]
---
[1] Url: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0291824

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.