High cost of Scientific Journals is largely an artifact of how the journals of various academic societies were digitized. During the late 90's, the academic so

High cost of Scientific Journals is largely an artifact of how the journals of various academic societies were digitized. During the late 90's, the academic societies realized that making their archives internet accessible would be very high value. But 90's era internet economics meant that this would cost several hundred million dollars. But during the late 90's stock market bubble, it was also possible to raise several hundred million dollars. As a condition for digitizing journal archives, the publishers acquired the copyrights from the academic societies. This investment has turned out to be extremely profitable. Publishers have been able to charge immense institutional access fees for electronic access to the journals. When I was studying at Upenn, 20% of undergraduate tuition went to paying for institutional access. The notion of preprint archives have been a somewhat successful work around. This basically creates a free distributable version of the work before the copyright is transfered from the authors to the publisher. Unfortunately, physics and math are the only disciplines that have established the use of preprint archives as a norm.