Covid-19 Correlations

Introduction

NPIs have become increasingly controversial as the collateral damage from ruined businesses, job and education losses, social isolation, and disrupted lives has accumulated.  Was this damage justified by NPI efficacy in reducing mortality rates?

Moreover, the severity of the collateral damage from NPIs raises a second question: is there evidence that non-viral effects were responsible for a significant portion of the pandemic deaths?

The fifty American states offer an opportunity to address these questions using correlational analysis.  In this paper we test the likely causal effect of eleven state characteristics that have been hypothesized to influence covid mortality:

•          Covid vaccine rate

•          NPI intensity.

•          Mask use.

•          Covid test rate.

•          Covid case rate.

•          Healthcare $ per capita.

•          ICU beds per capita.

•          Population over 65.

•          Obesity rate.

•          Population density.

•          Poverty rate.

We focus on the Coefficient of Determination (R2) results for these correlations to determine whether they support the causal hypotheses.

The discussion is organized in four sections:

•          Excess all-cause deaths are the best measure of pandemic mortality.

•          R2 ≤ 10% is rather strong evidence of no causal effect.

•          Were NPIs effective in reducing mortality rates?

•          Was pandemic mortality mostly caused by the sars2 virus?

Continue.