(1) Our researchers are paid (and their research projects are funded, often with tax-payers' money) to conduct research and to publish their findings ("publish or perish") so that other reseachers, at other universities and research institutions worldwide, can access, use, build-upon, cite and apply those findings. This is called "research impact". The productivity and progress of research depend on its impact.
(2) Research is published in peer-reviewed journals (24,000 worldwide, across all disciplines and languages, publishing about 2,500,000 articles per year).
(3) Unlike book-authors or journalists, research-article-authors do not seek royalties or fees for these writings: They write them only for the sake of research impact. (This is why these authors and their institutions were always willing, in the paper era, to undertake the effort and expense of mailing out hard-copies of their articles to any would-be users who requested a reprint, and were sometimes even willing to pay page-charges to the journal for publishing their article. Greater research impact means (i) career advancement, higher salary, more research income, prizes and prestige for the researchers and their institutions and, more important, (ii) greater research productivity and progress, hence greater benefits to the tax-payers who fund the research.)
(4) In the paper era, the only way for journals to cover the costs of peer-review and publication was to charge subscription tolls for access: Universities and research institutions paid the tolls so their own researchers could access and use the peer-reviewed research output of other universities and research institutions.
(5) No institution could ever afford toll-access to anywhere near all 24,000 journals; and most could only afford a small fraction of them -- a fraction that keeps shrinking with rising journal prices, even in the Web era.
(6) As a result, it was true in the paper era -- and is still true today, in the Web era -- that for each one of the 2,500,000 articles published yearly, most of its would-be users cannot access it. That means much of its potential research impact is being lost.
(7) In the paper era, this impact loss was unavoidable, but in the Web era it is no longer necessary. There are two complementary ways in which all access-denial -- and hence all impact-denial -- can now be remedied:
(8) (OAJ) New "open-access" journals can recover their costs other than by charging the user-institution, as toll-access journals do, for each journal or article they access. (For example, they can instead charge the author-institution for each outgoing article they publish). (But fewer than 1000 such open-access journals exist so far, publishing only about 5% out of the 2,500,000 articles that are published every year.)
(9) (OAA) For the remaining 95% -- the articles published yearly in the 23,000 toll-access journals -- the immediate solution to put an end to access denial and impact loss is for their authors to self-archive their full-texts online on their own institutional open-access websites for all would-be users worldwide.
(10) As soon as universities, research institutions and research funders extend their existing "publish or perish" policies from just publishing their research output to also providing open access to it -- via OAJ, by publishing it in an open-access journal whenever a suitable one exists, and otherwise via OAA, by self-archiving all their toll-access journal publications -- the open-access era will be upon us, and research progress and productivity will at last be maximised, instead of needlessly minimised, as it is now.
UNIFIED OPEN-ACCESS PROVISION POLICY:
(OAJ) Researchers publish their research in an open-access journal if a suitable one exists otherwise
(OAA) Researchers publish their research in a suitable toll-access journal and also self-archive it in their own research institution's open-access research archive.