Capping credit card late fees?
This seems like an empirical question and her entire piece has no actual empirical data. She only cites an industry trade group statement that is also devoid of evidence. It sure is easy to be a libertarian economist.
It could be right on the empirical question, but I'm not sure how theory could answer it a priori, moralizing stories aside. Issuers have other means besides late fees available to them if late fees are restricted. Those other means might be less abused. Or have better incidence.
It'd be like arguing credit access for poor people is restricted if issuers can't punch poor people that default. If an issuer behavior is abusive enough, on balance it could be beneficial to regulate it.