Richard Grant at Naturally Selected recently posted his take on some recent discussions around scholarly publishing. In a comment, he mentioned that non-profit organizations may not be suited for scholarly publishing. I wouldn't know that so no contradiction there. But I thought it would be a nice thought exercise to make a rough, speculatory sketch of what the profit could buy that coprorate scholarly publishers make every year. The big three players are Thomson Reuters, Reed Elsevier and Wolters Kluwer. Together, they make about 5 billion US$ in adjusted operating profit (according to my  slide #26 that I researched a few years back using available financial records). Clearly, not all of that will come from library budgets, but my guess is that the largest chunk will come from some public budget or another.

So if these publishers would not make any profit at all, we'd have 5 billion every single year to spend on investments in scholarly communication. What would I do with that sum? I think I'd give it to libraries to be able to archive every single piece of literature and data their faculty produce. All that's needed for that on top of the infrastructure already present, is man-power and servers. So let's say we equip each and every university library on the whole planet (about 10,000 according to univ.cc) with a team of four experts (yearly salary of 100k) and a new $10,000 server each year. That means a total work force of specialists of 40,000 together with a distributed server park of 10,000 brand new servers. I'd think this should be overkill for any such task. Anyway, with just 4.1 billion per year you'd be able to equip every single university library on the planet with what should be more than sufficient resources to archive everything, make it accessible and provide efficient software to boot.

And then you'd still have 900M to burn on something interesting every year... wait, then the corporate scholarly publishing industry would be obsolete, so make that more than 10 billion every year.