This is one of the most unconvincing instruments I've ever seen in an economics paper. They don't even try to justify why variables like 'parents alive' (!), 'o

This is one of the most unconvincing instruments I've ever seen in an economics paper. They don't even try to justify why variables like 'parents alive' (!), 'other public benefits' (!!), 'owning your own house' (!!!), or 'number of children' or 'work experience' (!!!!) are all causally independent of intelligence. There's no way these form a valid instrument allowing causal inference. Hey, why not throw in education or grades while you're at it, after all, "The validity of the variables used to generate exclusion restrictions needs to be investigated in future research", since that would be as valid an instrumental variable as any of the others... (I also object to the media framing this as an argument for 3-day work weeks. If cognitive performance peaked at 3-days, employee performance could still be way worse due to the few hours split up, in addition to the economic overhead for businesses of having many part-timers.)