The title is a not particularly well-evidenced suggestion that firms are paying litigators for access. Rather, they are paying these people for exactly what you'd expect: they're the absolute best law students in the country and they have work experience that gives them a leg up in the toughest types of litigation the firms handle.

The latter claim is much more interesting. Though obscured by the title, the purpose of the article is to cover an forthcoming PRQ study which argues that (with a credible but not iron-tight causal inference design -- basically achieving conditional ignorability through a matching estimation strategy and further adjusting for case-specific details in a pretty normal way) former clerks do much, much better at getting their former boss's vote than expected, even net of the confounders you'd expect (ideological agreement, relevant case characteristics).

The article, which is open access (https://journals.sagepub.com/doi/pdf/10.1177/106591292094813...), is short, written in quite plain English, and has a better discussion of theoretical arguments about the revolving-door.

Inside academic social science baseball: I am a tad surprised this ended up at PRQ and not, say, JOP or the APSR or even AER.