> 2. Having a housewife leaves the man with more free time to work.

That's hypothesis 2 of the three that they looked at, that marriage "makes a man a more productive worker".

There is a need for a study. What if employers are simply assuming that married men are better workers (hypothesis 3), but married men aren't actually better workers? Than as an unmarried man, I'm getting discriminated against, and marital status is a protected class. This is good shit to look at.

No, marriage does not make a man more productive.

A productive man is more likely to get married (He's a good catch)

It's a selection bias on the part of the woman (or man, not that there's anything wrong with that)

I forgot 3, then:

3. Employers assume married men, especially if they have children, are hard workers and responsible

I'm glad you're being complete :)

But, can you see why it'd be good to study this? Which one of the three is it? Or if it a mix, and what is the mix?

Of course it would be good! My comment was facetious... in part ;-)

Those are the first and third of the the three hypotheses that they consider. They attempt to figure out how large the impact of each hypothesis is.