Life expectancy loss by education level and sex: the impact of COVID-19 in the US (2020) and their forecasts

Abstract

Objective: Stratified life expectancy loss by education levels and sex helps measure particular mortality impacts during a catastrophic event. We propose a statistical approach to estimate them using the US case during the COVID-19 pandemic in 2020. Method: First, we estimate life expectancies according to available data, including those years when catastrophic events occur. Second, we use them to calculate a valid multivariate time series VAR(p) model, omitting the respective catastrophic(s) year(s). Through the model, we generate forecasts, which are compared with estimated data, and afterward, the life expectancy losses are quantified as their differences. Results: Less than four times the life expectancy losses with low education compared to the high education group. Our projections also indicate that life expectancies with almost all education falls outside the forecast intervals. Conclusion: The more educated the population is, the less life expectancy is lost. Women always outlive men within each education stratum. Long-term estimates continue to underscore gender disparities in life expectancy.

Keywords

• life expectancy loss
• VAR model
• multivariate forecasts
• education level