Competing societal objectives in epidemic mitigation: A modelling study of COVID-19 in the Philippines

Rob Johnson^1*Rolly Czar Joseph T. Castillo²Cymon Lubangco²Timothy Robin Teng²Mark Tolentino²Christian Morgenstern¹Patrick Doohan¹David Haw³Giovanni Forchini⁴Elvira Pederes de Lara-Tuprio²Joselito T Sescon²Katharina Hauck¹

• ¹Imperial College, London, United Kingdom
• ²Ateneo de Manila University, Quezon City, Philippines
• ³University of Liverpool, Liverpool, United Kingdom
• ⁴Umea Universitet, Umeå, Sweden

School closures and suspension of non-essential economic activities are highly effective respiratory-pandemic mitigation strategies because they effectively interrupt disease transmission. However, they come with high societal costs. Most of these costs are borne by workers who lose their income, especially those who are not supported by welfare benefits, and students whose future income depends on their education. In countries where many households live close to the poverty line, closures should be designed to minimise impacts on the most vulnerable. The objective of this study is to learn and compare policy responses that minimise the number of householdspeople that fall below the poverty line, maximise GDP, or maximise societal welfare in a model of the COVID-19 outbreak in the Philippines. Towards this objective, we construct social welfare functions that take into account lives, education, poverty, and GDPquantify societal welfare in terms of lives, education, GDP, and we introduce poverty as a novel fourth dimension. We then use a population microsimulation model together with an economy-structured, an epidemiological model, and GDP and education projections to determine intervention strategies involving the partial closure of schools and economic sectors with the objective of mitigating the epidemic while minimising societal losses. We find the cost of reducing poverty is substantial in terms of the other outcomes, making a case for poverty reduction as an important tool for increasing societal resilience and preparedness for crises such as pandemics. From a modelling perspective, we identify the need for timely data collection in order to create tools to assist in future epidemics.