Overcoming Data Sparsity: A Machine Learning Approach to Track the Real-Time Impact of COVID-19 in Sub-Saharan Africa
IMF Working Papers, May 6, 2022
Source details
- Canonical URL
- Overcoming Data Sparsity: A Machine Learning Approach to Track the Real-Time Impact of COVID-19 in Sub-Saharan Africa
Other formats
Bibliographic details
- Authors: Karim Barhoumi, Seung Mo Choi, Tara Iyer, Jiakun Li, Franck Ouattara, Andrew J Tiffin, Jiaxiong Yao
- Published: May 6, 2022
- Series: IMF Working Papers
- DOI: https://doi.org/10.5089/9798400210136.001
Overview and Purpose
- Objective: Outline a machine-learning framework to track economic activity in real time for economies with delayed publication of official GDP statistics.
- Application: Framework applied as illustrative examples to selected sub-Saharan African economies.
- Main claim: The framework is able to provide timely information on economic activity more swiftly than official statistics.
Key Findings
- The COVID-19 crisis has had a tremendous economic impact for all countries.
- Delayed publication of official GDP statistics in several emerging market and developing economies hampers assessment of the full impact of the crisis.
- A machine-learning nowcasting framework can overcome data sparsity to track economic activity in real time.
- As illustrated for selected sub-Saharan African economies, the framework provides timelier information than official statistics.
Methodological Themes
- Approach: Machine learning applied to nowcasting economic activity.
- Focus areas and related topics: COVID-19, Economic forecasting, Foreign exchange, Health, Machine learning, Oil prices, Prices, Real effective exchange rates, Technology.
Policy-Relevant Implications and Recommendations
- Use of machine-learning nowcasts can supplement official statistics to inform timely policy responses during crises characterized by delayed data release.
- Policymakers in sub-Saharan Africa and other data-sparse environments can leverage real-time machine-learning estimates to better assess economic conditions amid rapid shocks such as the COVID-19 pandemic.
Publication and Metadata
- Authors: Karim Barhoumi, Seung Mo Choi, Tara Iyer, Jiakun Li, Franck Ouattara, Andrew J Tiffin, Jiaxiong Yao
- Date: May 6, 2022
- Series: Working Paper No. 2022/088
- Issue: 088
- Volume: 2022
- Pages: 23
- DOI: https://doi.org/10.5089/9798400210136.001
- Stock No: WPIEA2022088
- ISBN: 9798400210136
- ISSN: 1018-5941
- Subjects and Keywords: Africa; COVID-19; crisis in Sub-Saharan Africa; data sparsity; Economic Activity; GDP; GDP statistics; Global; learning framework; Machine Learning; machine learning approach; Nowcasting; Oil prices; Real effective exchange rates; Sub-Saharan Africa
IMF Working Paper — Overcoming Data Sparsity: A Machine Learning Approach to Track the Real-Time Impact of COVID-19 in Sub-Saharan Africa (Working Paper No. 2022/088, May 6, 2022).
Content in this bundle
- Working Paper