[throwaway81523]

> Holland's 'Statistics and Causal Inference' paper (1986)

In case anyone else wants to take a look:

https://www.jstor.org/stable/2289064

https://doi.org/10.2307/2289064