No, it's the opposite: when subscriptions were for printed copies, it's not like Elsevier would send an agent to set fire on your archives as soon as you discon

No, it's the opposite: when subscriptions were for printed copies, it's not like Elsevier would send an agent to set fire on your archives as soon as you discontinued a journal. In the 1990s, electronic access was seen as a cost reduction technique: some publishers would even provide it for free to their subscribers, others for a small fee; libraries would buy less copies of the same journal. Later, publishers offered to terminate the paper copies and replace them with electronic-only access, but libraries sensed a trap: what happens if we terminate access in the future? By the early 2000s, most universities had learnt to include such a perpetual access clause whenever they switched to electronic-only access. (See EBLIDA advice in 2001: <http://www.eblida.org/Activities/Publication/Licensing_digital_resources.pdf>. http://www.eblida.org/Activities/Publication/Licensing_digit...) All this work was not completed quickly: around 2010 I remember we (the board of directors) still had to approve some such conversions in my university. (Germany status in 2010: <https://www.mpg.de/230683/access__hosting_studie_e.pdf>. https://www.mpg.de/230683/access__hosting_studie_e.pdf>.) However, most of these clauses have never been used so far, and for some universities it's a lot of work to invoke such perpetual rights, track the back copies etc. https://oadoi.org/10.1080/0361526X.2017.1383964 https://oadoi.org/10.1080/0361526X.2017.1383964 Not all universities have a self-hosted archival copy, either. https://www.lockss.org/use-lockss/post-cancellation-and-perpetual-access https://www.lockss.org/use-lockss/post-cancellation-and-perp...