Because the marginal cost, to the employer, of hours beyond 40 is zero – remember these employees are almost always paid a salary, i.e. a fixed amount regardless of hours worked (beyond 40). Thus, even if an employee's productivity plummets – but is still above zero – it's in the employer's interest to persuade (or prod) them to work more hours.

But of course even that's a gross simplification. There are lots of 'non-linear' considerations, e.g. burnout.

It's not that less is more; rather it's that more work has diminishing productivity returns. And, after a certain point, the returns dip negative due to fatigue-related risks (this is a unique example, but it's always major news when the maximum hours allowed for medical residents gets reformed/reduced on account of patient mortality related to doctor fatigue).

And that's just when we limit the scope of "returns" to productivity. As the article points out, there are many, many non-productivity benefits associated with a shorter work week.

I don't think many Americans will be on their death bed lamenting "oh, if only I'd spent more late nights/weekends at the office working on spreadsheets."