The Proportionality of Common Sense Causal Claims

This paper defends strong proportionality against what I take to be its principal objection – that proportionality fails to preserve common sense causal intuitions – by articulating independently plausible constraints on how to represent causal situations. I first assume an interventionist formulation of proportionality, following Woodward. This views proportionality as a relational constraint on variable selection in causal modeling that requires that changes in the cause variable line up with those in the effect variable. I then argue that the principal objection derives from a failure to recognize two constraints on variable selection presupposed by interventionism: exhaustivity and exclusivity