Can we utilize the "long tail" of our collections?

The Anderson book initially started losing me this week as I felt it was getting deeper and deeper into economics. I, personally, hated taking economics in college (and hoped I'd find an industry to work in where I never had to deal with it.) The more of Anderson I read as the week progressed, however, the more I realized how economics really plays a huge role in libraries (as I know now, alas, it does in almost all fields. I guess this is even more evident in harder economic times.)

Anderson's discussion about all hits eventually losing their "hit status" and becoming a part of the tail made me think of the carts and carts of library materials we weed from our collections on a regular basis. Because most libraries are still brick and mortar institutions, this is inevitable (physical limitations prevent them from being able to "carry everything.") Individually, I don't think libraries can truly be effective Long Tail organizations because of these limitations. If they work together with other libraries, though, both to ensure that they can collectively cover all niche markets, and to build their digital collections (promoting the "libraries without walls" concept), they can, as a whole, be long tail information providers. I think we can most effectively make this happening by focusing our collection development efforts on the "tails within tails" that Anderson discusses (pp. 139-141.) I don't think all libraries can really be everything for everyone. But if they combine their efforts and each works to fill a certain niche (library A, for example, builds a solid business collection; another library in the area, library B, focuses on medicine, etc.) they will be better prepared to meet everyone's needs and wants.