I wouldn’t draw any conclusions from this study. It is a study limited to NY state for one (strong government-union political relationships), and the authors th

I wouldn’t draw any conclusions from this study. It is a study limited to NY state for one (strong government-union political relationships), and the authors themselves mention that the various limitations (like confounding variables) preclude causal interpretations of the correlation they observed. Some relevant quotes: > Even with these adjustments, the available data and observational research strategy preclude strong causal interpretations due to bias from unmeasured confounders and selection into our study sample. > While our study design precludes causal interpretations, our results suggest that unions may have reduced COVID-19 deaths among nursing home residents by successfully demanding PPE for health care workers. The study also claims it adjusts for some of the following variables but in the analyses it wasn’t clear exactly how they adjusted for it: > Unionized facilities were also more likely to be for-profit, less likely to be associated with a chain, had lower LPN-to-resident ratios, and were located in more populous counties with higher per capita rates of confirmed COVID-19 cases (supplemental exhibit 1). The higher levels of PPE access, which the study thinks may be a trait of unionized facilities, could be due to many factors such as random luck, facility size, political connections, etc. I wonder how causally predictive those would end up being in studies that dive deeper into those variables.