An economics paper would usually try to exploit a natural experiment changing work hours for people in a way independent of their intelligence and jobs. For example, new labor legislation banning 40-hour work weeks, causing a drop in worked hours: you should be able to see cognitive scores of the same individual increase after the legislation takes effect.
Measure productivity of the same people working different numbers of hours per week? (And the study needs to run longer than one week at each number of hours. Working overtime boosts your productivity - the first week.)
How do you measure the productivity of cognitive work over an entire week? It's kind of subjective, and having the subjects self-report their level of satisfaction would probably be unreliable.
EDIT: I guess the trick might be to not use a survey, but to give the participants all the same project to work on (very expensive), then adjust the results based on their previous performance. I think an ideal group for this would be students because they're all already working on the same projects. Tell one (random) group that they can have some money if they promise to work on their class project for no more than one hour a day. Could this work?