Quantifying Cities for Sustainable Development: Transforming urban data collection to support research and policy making in developing countries

Lead Research Organisation: University of Bristol
Department Name: Geographical Sciences

Abstract

Our project addresses a critical gap in social research methodology that has important implications for combating urban poverty and promoting sustainable development in low and middle-income countries. Simply put, we're creating a low-cost tool for gathering critical information about urban population dynamics in cities experiencing rapid spatial-demographic and socioeconomic change. Such information is vital to the success of urban planning and development initiatives, as well as disaster relief efforts. By improving the information base of the actors involved in such activities we aim to improve the lives of urban dwellers across the developing world, particularly the poorest and most vulnerable. The key output for the project will be a freely available 'City Sampling Toolkit' that provides detailed instructions and opensource software tools for replicating the approach at various spatial scales.

Our research is motivated by the growing recognition that cities are critical arenas for action in global efforts to tackle poverty and transition towards more environmentally sustainable economic growth. Between now and 2050 the global urban population is projected to grow by over 2 billion, with the overwhelming majority of this growth taking place in low and middle-income countries in Africa and Asia. Developing evidence-based policies for managing this growth is an urgent task. As UN Secretary General Ban Ki Moon has observed: "Cities are increasingly the home of humanity. They are central to climate action, global prosperity, peace and human rights...To transform our world, we must transform its cities."

Unfortunately, even basic data about urban populations are lacking in many of the fastest growing cities of the world. Existing methods for gathering vital information, including censuses and sample surveys, have critical limitations in urban areas experiencing rapid change. And 'big data' approaches are not an adequate substitute for representative population data when it comes to urban planning and policymaking. We will overcome these limitations through a combination of conceptual innovation and creative integration of novel tools and techniques that have been developed for sampling, surveying and estimating the characteristics of populations that are difficult to enumerate. This, in turn, will help us capture the large (and sometimes uniquely vulnerable) 'hidden populations' in cities missed by traditional approaches.

By using freely available satellite imagery, we can get an idea of the current shape of a rapidly changing city and create a 'sampling frame' from which we then identify respondents for our survey. Importantly, and in contrast with previous approaches, we aren't simply going to count official city residents. We are interested in understanding the characteristics of the actually present population, including recent migrants, temporary residents, and those living in informal or illegal settlements, who are often not considered formal residents in official enumeration exercises. In other words, our 'inclusion criterion' for the survey exercise is presence not residence. By adopting this approach, we hope to capture a more accurate picture of city populations. We will also limit the length of our survey questionnaire to maximise responses and then use novel statistical techniques to reconstruct a rich statistical portrait that reflects a wide range of demographic and socioeconomic information.

We will pilot our methodology in a city in Pakistan, which recently completed a national census exercise that has generated some controversy with regard to the accuracy of urban population counts. To our knowledge this would be the first project ever to pilot and validate a new sampling and survey methodology at the city scale in a developing country.

Planned Impact

Who will benefit from this research?

Beyond academic beneficiaries (see above) we expect two further groups to benefit from the research:
1. Policy makers and development practitioners will benefit from the use of the City Sampling Toolkit, which improve the knowledge base that informs urban plans and interventions designed to enhance welfare, productivity and resilience;
2. Urban residents in low and middle-income countries will ultimately benefit from better informed urban planning, development and resilience initiatives.

How will they benefit from this research?

By 'policy makers' we mean urban planners, government officials, statisticians and bureaucrats involved in designing, implementing and evaluating initiatives at local and national levels. By 'development practitioners' we mean those working in non-governmental organisations or multilateral agencies (e.g. UN, World Bank, etc). These groups depend upon accurate and timely data to make important decisions about how scarce financial resources are allocated. Incorrect information can have a wide range of negative consequences for urban residents-our ultimate intended beneficiaries.

For example, inaccurate estimates of city populations can lead to underinvestment in critical infrastructure and services, including water, waste and transport. This, in turn, can result in higher disease burdens, congestion and elevated risks from natural disasters such as floods and fires. Similarly, underestimates of youth populations may lead to under provision of educational facilities, which can reduce labour force productivity. By providing a cost-effective tool for gathering accurate information on population characteristics and dynamics we hope to improve the quality of the work done by policy makers and development practitioners.

Within Pakistan, where we will conduct our pilot, our research will feed directly into ongoing debates that will shape urban policy in the years ahead.

What will be done to ensure that they have the opportunity to benefit from this activity?

While our ultimate intended beneficiaries are urban residents, our impact activities will focus on reaching policy makers and development practitioners. We will make our City Sampling Toolkit freely available online. To ensure that it actually gets used, we'll undertake a stakeholder mapping exercise during the inception phase of the project to identify specific organisations and networks to target through our communications channels throughout the research process, from inception to dissemination. This will include relevant multilateral organisations and affiliated agencies (e.g. UN-Habitat, UNDP, World Bank, Cities Alliance), bilateral agencies (e.g. DfID), international networks (e.g. Commonwealth Local Government Forum, Council for the Development of Social Science Research in Africa), research institutions and think tanks (e.g. International Institute for Environment and Development, Indian Institute for Human Settlements), and organisations that specialise in sharing knowledge and building capacity (e.g. Bond, INASP).

Within Pakistan we are partnering with the independent Collective for Social Science Research in Karachi. We will expand our engagements within Pakistan as the project progresses.

Publications

10 25 50
 
Description We have developed a technique for mapping household welfare at the neighbourhood scale using electricity consumption data. We have compared our results to a similar approach that relies on night lights data from satellites and demonstrated that night lights data cannot be relied upon to produce small area estimates of household energy consumption (which is a proxy for household welfare).
Exploitation Route We will be publishing our findings and publicising them in policy circles, as we believe the technique we've developed could be used for poverty mapping in cities in many low- and middle-income countries.
Sectors Government, Democracy and Justice

 
Description CEGA / CSSR - Combining satellite imagery and machine learning to target social protection in Pakistan 
Organisation Center for Effective Global Action
Country United States 
Sector Public 
PI Contribution We are developing machine learning techniques to estimate rural poverty from satellite imagery.
Collaborator Contribution Our partners at the Collective for Social Science Research in Pakistan are providing training data derived from a large household survey conducted by the Sindh Province. They will also be conducting a ground-truthing survey to evaluate the quality of our poverty predictions.
Impact None yet.
Start Year 2020
 
Description IGC Partnership - Using spatial data for targeting the poor: Reaching the most vulnerable 
Organisation International Growth Centre (IGC)
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution We are collaborating on the survey design, sampling and data analysis drawing directly on data and tools generated from the ESRC project
Collaborator Contribution Our partners have co-designed and executed the survey in the field. They will also be analysing the data to inform social policy in Karachi.
Impact None yet. Survey is ongoing.
Start Year 2020