by Lars P. SyllCauses deduced in an experimental setting still have to show that they come with an export-warrant to the target population/system. The causal background assumptions made have to be justified, and without licenses to export, the value of ‘rigorous’ and ‘precise’ methods - and ‘on-average-knowledge’ - is despairingly small.