Basically every research institute, university, college, and industrial scientific business has subscriptions and deals with basically every journal out there with the exception of those with terribly low impact factors or terribly speciallized topics. The most important factors to consider when publishing research papers are impact factor and relevance. In other words, how many people will read it? and will the RIGHT people find it easily or even stumble upon it?

The price that these journals charge for access to their publications (subscriptions, never buy papers one at a time unless you absolutely have to) basically serves to keep them afloat.

Also, many major journals are moving toward free online access. Journal of Virology is an example. It is the highest journal specific to my field and is easily accessable free through pubmed or through their own website. The people (i.e. scientists) who need to see the research that is out there always have access to it through their emplyers. The papers in these journals are essentially useless to the public anyway, so truly, it doesn't need to be free. Of course, it is nice when it is though. The public gets scientific information via media sources. One of the jobs of scientists and journals is to get that information to those media sources in a format that anyone can understand. Books/textbooks are other sources for Joe Q.

The problem is that the journalists don't even read the papers they're writing about. I'll bet they plagiarize somebody's press release instead. And journalists almost never cite their sources in enough detail that you can actually trace it back to a specific scientific paper (again, probably because they read the press release instead).

The idea that the info would be useless to the public might be a decent average but they do have some use. I have. I don't trust the press to report accurately, they tend to report first indications of potential problems as some sort of proofs of things the researchers never claimed and such like that. I want to at least see the abstract, but I'd rather see the report itself. I've seen inaccuracies of the sort in everything from the marijuana debate to global warming, the press on average doesn't know science any better than I do and I'd think on average they are worse. Or dishonest, take your pick.

The problem is that when you want to get "published" you basically have to sign over the rights to reproduce the article, which is total bullcrap. I want to go GPL with any research I do too-