Data-driven tools can support decision-making and improve implementation – especially in crises like COVID-19
This page in:
Following major shocks due to conflict, natural disasters, or even pandemics such as COVID-19, this kind of data collection can be impossible—or too slow to be useful. Governments and development organizations can sometimes find themselves in a catch-22, with well-designed interventions being most needed in the very places where timely data are impossible to collect.
To improve our ability to respond to these challenges, the World Bank has developed an innovative toolbox for rapid data collection and evidence-based policymaking. The first suite of tools aims to collect poverty and inequality data in a timely and cost-effective way, while the second helps managers make informed decisions related to portfolio allocation, project targeting, and project monitoring. The tools help projects to quickly adapt to changing circumstances and reach the people most in need of assistance. In the context of COVID-19, we adjusted the tools to inform the World Bank’s emergency response.
Rapid monitoring tools
One of the most agile tools used to gather data about the impacts of a crisis are Rapid Response Phone Surveys (RRPS). These are quick surveys administered by telephone to households or businesses, with each interview typically lasting less than 20 minutes. RRPS allow for near real-time survey data collection and make it possible to cover an entire country and even reach respondents in remote areas. RRPS have proven to be invaluable in monitoring socioeconomic impacts during the COVID-19 pandemic, making data-gathering possible while minimizing the health risks for beneficiary populations and survey teams. The World Bank has already deployed more than 100 RRPS around the world to measure behavioral responses to COVID-19 as well as socioeconomic impacts on households, firms, and communities. An example with an interactive country dashboard is Kenya, where RRPS is continuously deployed to gather information from households and firms.
Another broadly applicable rapid monitoring tool is the Survey of Well-being via Instant and Frequent Tracking (SWIFT) that can generate poverty estimates using five- to 10-minute interviews. Instead of collecting comprehensive income or consumption data, SWIFT focuses on key indicators that are correlated with poverty, such as household size, ownership of assets, or level of education. These indicators are then converted into meaningful poverty statistics using estimation models. The SWIFT method applies machine learning techniques and survey-to-survey imputation based on the latest available household survey data to generate estimates comparable with official poverty and inequality statistics.
The World Bank has already applied this highly cost-effective methodology in over 50 countries. SWIFT has been adapted to monitor rapidly changing socioeconomic conditions due to COVID-19, including as part of some RRPS in some 20 countries. In Saint Lucia, using SWIFT highlighted significant differences in food insecurity and access to medical attention between richer and poorer Saint Lucians, showing the pandemic doesn’t affect everyone in the same way.
Portfolio and project management tools
These new and agile tools are not limited to data collection, but also include strong mechanisms for project and budget management. The Portfolio Footprint (PF) examines the distribution of World Bank projects spending within a country and its relationship with the needs of the population. This approach starts by estimating a World Bank portfolio’s past and committed expenditure in different regions of a country. These investments are then compared with different welfare and sector-specific indicators for those regions. This helps identify any potential misalignments between needs and World Bank investments, and informs discussions about targeting in the future. PFs are currently supporting portfolio mappings in Central America in an attempt to correlate project locations with COVID hotspots.
Toolkit Implementation by Country (darker color indicates more projects across all tools)
The Project Targeting Index (PTI) is an evidence-based targeting method that identifies high-priority geographical areas for intervention. PTI brings together relevant criteria such as poverty rates, the number of poor, and other socioeconomic data, and aggregates these numbers into a single indicator. The framework is customizable and can include project-specific criteria such as food insecurity, education, or access to services. After calculating the PTI, geographical areas such as districts or counties are then ranked in order of priority for project implementation. This kind of analysis can help guide project teams in selecting project sites and can introduce greater transparency and objectivity into the selection process. In Madagascar, the PTI approach was further refined to consider COVID-19 cases and inform the World Bank’s emergency response.
Beyond the planning and budgeting stages of programming, Iterative Beneficiary Monitoring (IBM) is tailored for tracking project implementation. This low-cost method gathers feedback directly from beneficiaries and provides managers with a real picture of what is happening on the ground, regardless of the stage of the project. IBM produces short reports and focuses on identifying specific stumbling blocks for project implementation. IBM is usually carried out through very quick phone interviews, collecting multiple rounds of small-scale data from a sample of project beneficiaries. Costs are kept to a minimum. IBM offers a problem-oriented feedback loop that enables managers to adjust project activities on-the-fly. The first IBM used in the context of COVID-19 was conducted in Morocco for the Municipal Performance Program, which aims to improve the delivery of services and infrastructure for city residents. IBM assessed both disruptions in service delivery and municipal responses to help understand the challenges of the project and improve crisis preparedness.
These agile and responsive tools for data collection and analysis allow the World Bank teams and country counterparts to make decisions based on statistical evidence even when access is difficult and time is short. In addition to reducing costs and allowing for more frequent monitoring, we can use these methods in countries with severe access restrictions, for example, due to insecurity or the risk of contagion during the COVID-19 pandemic.By investing in this kind of innovation, the World Bank will continue to increase its ability to provide effective and timely support to populations most in need during times of crisis.
This is very important, valuable and interesting work. Decision makers at the county level, especially the politicians, often plan and implement interventions with any supporting data. Neither do they wish to commit financial and other resources for the collection of strategic data for planning such as weather and hydrology. Is the data collected in Kenya accessible to the public or one has to go through government departments to access it?
Thanks for this interesting article - are these tools open source and would they be suitable for adoption by bilateral donors or large NGOs or UN agencies?