David & Brachet (2011) find labor turnover contributes twice the organizational forgetting that skill decay does

organizational-forgettingeconomicsquantitativelearning-curve

Applying the continuous/discrete channel split from claim-david-brachet-2011-splits-forgetting-into-continuous-and-discrete-channels to a 1991-2005 dataset of Mississippi ambulance companies and their paramedic workforce, Brachet and David found both channels present but of unequal size: "We find evidence of organizational forgetting, which results from sizable skill decay and turnover effects, with the latter having twice the magnitude of the former." This is the paper's own headline quantitative result, stated in its abstract — labor turnover (the discrete channel) accounts for roughly twice the forgetting that individual skill decay (the continuous channel) does in this setting.

The figure is specific to the paramedic/ambulance-company population studied and is not asserted as a general ratio across industries — compare Argote's ship-production estimate and Benkard's aircraft-production estimate, which put the organizational-forgetting rate itself at very different levels industry to industry.