See also SPARC's comments on the benefits of institutional repositories.
One metric that is frequently used to understand the degree to which articles of a journal are noticed is the “impact factor” or IF. Journal Citation Reports (JCR) calculates the IF for many science and social science journals giving authors, readers, and publishers a rough estimate of the visibility of a journal. The JCR database includes many OA journals some that have surprisingly high IF’s given their very short publication history. For example, PLoS Biology and PLoS Medicine, both no more than 10 years old, ranked number 1 (out of 86 journals) and number 5(out of 153) in their respective JCR categories of Biology, and General/Internal Medicine. Several studies have also shown that free online articles increase citation rates (see for example, Lawrence, 2001 or Eysenbach 2006).
Another class of research metrics known as "altmetrics" (alternative metrics) takes into account the rate at which articles are downloaded, mentioned or archived in various social media. The effect of information dissemination via websites such as Mendeley, CiteULike, and Twitter as well as many blogs is almost immediate providing seemingly real-time measures of the influence of an article.
In a study published in Science in February 2010, authors Evans and Reimer determined that people tended to cite OA journals more than non-OA journals by about 8%. But more striking was the difference in citation rate among poor nations (defined by gross national income) where the researchers in these countries cited OA journals as much as 20-25% more frequently.
Currently there are several proposed or active federal mandates in the U.S. There are also various state proposed policies. Listed below are the federal laws and bills.
Since information in self-archived repositories is not necessarily peer-reviewed, isn't the content potentially of lower quality than what is found in traditional journals? Each repository develops its own policies on who may submit and what can be submitted. Some have strict guidelines and membership requirements others are more open. For example arXiv, the mathematics and physics central repository, is very open, accepting primarily pre- and post-prints. To ensure a minimal level of quality however, it utilizes a system of moderators who can reject items based on inappropriate subject matter, inappropriate format of the submission, duplicated content, or submission of copyright protected material. However, experience from other high energy physics repositories is that most content in fact has been subjected to internal review prior to submission to arXiv or an alternative repository. Many deposited articles in fact go on to be formally submitted for publication.