I checked several math books out of the library over the summer, and they were sitting on my desk when the library emailed me saying they were 'missing', and I was being charged to replace them. I took them back to the library and they aren't missing anymore. Using the methods of this paper, in the interim period while they were declared missing but before I took them back, they were considered 'stolen', even though my account was charged a standard book replacement fee which was well above the cost to replace each book.
So 'missing' is quite often the same as 'extra overdue'. (in most university libraries you end up paying to replace any book you don't return anyway, so I doubt any ethicist would consider it stealing, but that isn't relevant to my point)
If ethicists just check out a lot of books, and they need them for longer than other people use other books, or else ethics books are in short supply so people often request them (so that they can't be renewed and end up being overdue and thus missing more quickly than other books), you would recreate this same data while all of the ethics books could still make their way back to the library in the end.
Here is another possibility: Students check out books from the library so they can avoid buying that books for a class. This would seem to be more common in philosophy than in other subjects because philosophy (or english literature say) tends to require original texts while other subjects have dedicated textbooks, and dedicated textbooks aren't generally held by libraries. If a student takes out a book but it is then put on hold by a professor, it generally is either recalled to the library or can't be renewed. The student just holds the book until the end of the semester, but it is marked missing by the time it is returned at the end of the semester since the hold accelerated the date it was due. Since the data was collected between "October 17 and December 13, 2006,", it is unclear what the data would look like if you did the sampling between say June and August.
One of my thousand gigs is adversarial critiquing of statistical sampling procedures (say, for lawsuits) and unaccounted-for seasonality is a big winner. It's a great way to skew samples in a deliberate matter.
The books mentioned in your anecdote would not have been categorized as missing using the paper's classification (which doesn't match your library's), only as overdue.
The paper's definition being: "missing (including books more than one year overdue)."
So 'missing' is quite often the same as 'extra overdue'. (in most university libraries you end up paying to replace any book you don't return anyway, so I doubt any ethicist would consider it stealing, but that isn't relevant to my point)
If ethicists just check out a lot of books, and they need them for longer than other people use other books, or else ethics books are in short supply so people often request them (so that they can't be renewed and end up being overdue and thus missing more quickly than other books), you would recreate this same data while all of the ethics books could still make their way back to the library in the end.
Here is another possibility: Students check out books from the library so they can avoid buying that books for a class. This would seem to be more common in philosophy than in other subjects because philosophy (or english literature say) tends to require original texts while other subjects have dedicated textbooks, and dedicated textbooks aren't generally held by libraries. If a student takes out a book but it is then put on hold by a professor, it generally is either recalled to the library or can't be renewed. The student just holds the book until the end of the semester, but it is marked missing by the time it is returned at the end of the semester since the hold accelerated the date it was due. Since the data was collected between "October 17 and December 13, 2006,", it is unclear what the data would look like if you did the sampling between say June and August.