Rubinstein's Experienced Advice for “Lost” Graduate Students in Economics

One of my favorites: 

Q3. I have already written 30 pages. I have repeated myself several times, and my proofs are much longer than necessary. I have added uncertainty wherever I could and I have moved from a discrete case to Banach spaces. My adviser still says I hardly even have enough for a note. How long should my paper be? 

If you don’t have a good idea, then keep going. Don’t stop at less than 60 single-spaced pages. Nobody will read your paper in any case, so at least you have a chance to publish the paper in QJE or Econometrica.

If you have a really good idea, my advice is to limit yourself to 15 double-spaced pages. I have not seen any paper in economics which deserved more than that, and yours is no exception. It is true that papers in economics are long, but then almost all of them are deathly boring. Who can read a 50-page Econometica paper and remain sane? So make your contribution to the world by writing short papers—focus on new ideas, shorten proofs to the bare minimum (yes, that is possible!), avoid stupid extensions, and write elegantly!

And

Find comfort in my motto: “A paper that has not been rejected should not be published.” But beware of the faulty logic in assuming that “every paper that has been rejected should be published.”

See more here.