Signalling effects. Imagine that you are one of five hundred postdocs applying for your dream professorship. Your papers are so highly technical that the hirin

Signalling effects. Imagine that you are one of five hundred postdocs applying for your dream professorship. Your papers are so highly technical that the hiring committee won't be able to judge them directly (certainly not before you make the short list). The most important component of your job application is your recommendation letters. But also very important is your CV, and hiring departments will want to see where you have published. Journals like Inventiones Mathematicae, Journal for die reine und angewandte Mathematik, Publications mathématiques de l'IHÉS, and so on have developed a reputation for publishing only outstanding papers, and so your CV will stand out if you have published them. Less so if you have published only in, say, Indagationes Mathematicae. Same when you are applying for tenure, promotion, grants, etc. Basically any time you need to demonstrate to anyone outside your immediate specialty that you are doing top-notch work. This could be done in other ways; indeed I would have better use for my brain cells than to keep track of a Byzantine hierarchy of journal quality -- but for now the system is quite entrenched.