Staff at a Florida library created fake readers to foil computer systems that recommend throwing out books unread for months or years.
An investigation found bogus borrowers had “read” thousands of books.
Library staff have been suspended while officials work out if the fakes meant the libraries received more state cash.
Staff said their action had helped save cash as it kept perennially popular titles that would have had to be bought again if they had been thrown out.
Officials at the Lake County public resources department that oversees libraries in the Florida locale told local papers they had been alerted to the existence of the fictitious readers at the East Lake County library by an “unidentified person”.
An investigation by Lake County municipal staff found several bogus identities had been created by senior staff at the library – one dummy reader was named Chuck Finley after a retired major-league baseball player.
Reports seen by local newspapers revealed that Mr Finley had checked out more than 2,600 books in nine months.
Some titles had been checked out and returned within an hour.
In total, suggests the official report, the faked folks at the libraries boosted readership numbers at East Lake by almost 4%.
Readership numbers are important because they count towards the amount of official cash that libraries receive.
Last year, libraries in and around Lake County received about $1m (£820,000) from local government coffers.
In a statement seen by the Orlando Sentinel, the head librarian at East Lake said the false borrowers had been created only to retain titles staff knew were always popular.
Following the computer system’s recommendations to ditch unread books would mean libraries re-purchasing titles they had only just thrown out, he said.
And, he alleged, many other libraries used “dummy” reader cards to do the same.
Digital activist Cory Doctorow said the logging system had become a “straitjacket” that rode roughshod over the views of staff who knew what readers wanted.
The librarians had not fooled the software for personal gain, he said.
“They did it because they wanted to make the system better, to teach it how to weight the circulation data to reflect the on-the-ground intelligence.”