(WITHDRAWN) Economic Resilience and COVID-19 Social Control Policies across Sub-Saharan Africa: Spatial Patterns and Geographically Weighted Regression Analysis

Background and Purpose: The incidence and mortality of COVID-19 has increased dramatically around the world. In sub-Saharan Africa (SSA), due to resource, capacity, and infrastructure constraints, some countries have adopted strict non-pharmaceutical COVID-19 public health responses to curb the spread of the coronavirus. These measures include the deployment of total lockdown, curfews, isolations, and spatial distancing protocols. While these measures are imperative to abate the spread of COVID-19, scant evidence exists on how SSA countries consider the economic vulnerability of their populace while instituting initial social control policies. Yet different social control measures, including full lockdowns of the economy, may have a financial risk for already vulnerable communities. Therefore, this study aims to (1) assess geospatial patterns in economic resilience and social control intensity, and (2) use an economic resilience framework to explore the relations between economic resilience and countries’ social control policy choices.

Methods: We harmonized three county-level data sources. Financial resilience data (saving, emergency saving, employment, agricultural income, wages, and domestic remittances) were drawn from the World Bank’s 2017 Global Findex database. COVID-19 cases data were obtained from the Africa Centres for Disease Control and Prevention. Social control data (dates, length, and types of lockdowns) were obtained from official press releases of each country. We used multivariate cluster analysis, an unsupervised machine learning method, to identify natural clusters in the financial resilience profiles of SSA countries. We also conducted geographically weighted regression (GWR) analysis to assess how countries’ economic resilience predicts their social control policies.

Results: Across the 40 countries with complete data, 40% (n=16) reported initial full lockdown, 37% (n=15) reported initial partial lockdowns, while 22% (n=9) indicated no initial social control. We identified two clusters of financial resilience (Pseudo F-Statistic=13.42): 44.1% (n=15) countries were financially resilient, while 58.9% (n=19) were not. Different financial resilience indicators had varying contributions (i.e., r-squared): savings (.59), domestic remittance (.43), self-employment (.41), wages (.28), emergency funds (.03), and agricultural income (.02). GWR results showed a stronger positive relationship between emergency funding and initial social control response, especially in central and southern Africa. Similarly, there was a positive relationship between personal savings and initial social control response—stronger in central and west Africa. There was a negative relationship between the number of COVID-19 cases and the initial social control response. This relationship was significant for most countries and particularly strong in central and west Africa.

Conclusions and Implications: The study advances our understanding of the heterogeneity of class membership associated with financial resilience indicators among SSA countries and extends our knowledge of economic resilience in the area of pandemic response. Considering that emergency funding, personal savings, and the number of COVID-19 cases are predictive of initial social control policies, public health policies need to take into consideration the social realities of their populace. We found that when there are both savings and emergency funding, the initial social control relationship is more intense. Therefore, there is a need for social protection programs developed and implemented to help populaces cope during and after the pandemic.