Tyler Cowen points us to a new journal, Rejecta Mathematica, that only allows submissions from articles rejected by other journals. Of course, that means that you have to submit somewhere else first. But it will be interesting to see how many submissions it gets from papers that could have been accepted somewhere else. After all, do you really want to signal your research as being in the definitely mistakenly rejected class as opposed to the possibly mistakenly rejected class? In economics, the experience of the BE Press Journals for which a lower tier journal explicitly implies rejection for an upper-tier one indicates a reluctance to play that game.