This one is kind of hilarious, although not unexpected.

You'll find the study here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2796267

What did they study?

They used two samples, amounting to 11,000 workers total, covering a wide range of demographics, employment types, industries and seniority. A measure of job satisfaction was constructed with the methodology you can read in the article, and correlations between job satisfaction and various parameters were calculated while controlling for confounds.

What did they find?

In their own words, this is the result from their first sample:

Column (2) shows that male worker job satisfaction is not significantly influenced by supervisor gender. Column (3) shows that the job satisfaction of women is significantly and negatively correlated with having a female supervisor.

and this for the second sample:

More relevant to our central question, we find that, even after including measures of supervisor attributes, women report lower levels of job satisfaction when their supervisor is also a woman.

In fact, they go so far as to say:

Indeed, our results tend to support those of Warning and Buchanan (2009) and Buchanan et al. (2012), who suggest that women have a “hidden bias” against female bosses; women tend to believe female bosses are no worse, and perhaps better, than males yet still do not want to work for one.

Let that sink in: women give their bosses performance evaluation that are no worse or slightly better than they give male bosses, but they still hate working under other women.

By the way, the result is not just an average; using data for workers who've changed from a male to a female boss or vice versa, the study finds:

For women, replacing a female supervisor with a man yields significantly improved job satisfaction compared to the alternative. When a woman replaces a male supervisor, job satisfaction significantly decreases, but only in the job fixed effects estimates (column (4) in Table 4). These effects are asymmetric in magnitude; the change from a woman to a man as supervisor has a larger effect in every specification than the change from a man to a woman. Finally, there remains no evidence among men that changes in supervisor gender correlate significantly with job satisfaction.

And in conclusion:

After controlling for worker fixed effects, as well as a host of demographic, workplace and supervisor characteristics, we find a persistent and negative relationship between women's job satisfaction and having female supervisors in two distinct data sets. The magnitude is non-trivial.

In RP terms...

Hierarchy, which is intimately connected to leadership, is a male a social structure. Men spent a few hundreds thousand years hunting fast and dangerous game: they had to collaborate, and they had to reward the best hunters and fighters, as a matter of survival. Women spent a few hundred thousand years gathering fruits and roots and jockeying for position with other women so they could access the best males, also as a matter of survival.

What that means is that women hate any woman and like any man who is ahead of them on the social ladder. Men might like or hate a boss, but they're no driven to wanting to bring him down the same way women are.