9.2.1 Why self-citations are not usually problematic
Self-citations are not necessarily bad or a reflection of manipulative behavior. If an academic does programmatic research, with individual studies building upon each other, one would expect him or her to refer back to their earlier work. To not do so would be violating academic integrity as it would make the later research appear more novel than it actually is.
In order to accumulate a significant number of self-citations, an academic also needs to have a significant number of publications in the first place, as one cannot cite oneself unless one publishes. Therefore, having a large number of self-citations usually goes hand-in-hand with having a high level of research productivity in general. However, most highly-cited academics have a very modest proportion (typically 5% or less) of self-citations.
Only academics with relatively weak citation records and/or junior academics tend to have a high proportion of self-citation. This is especially true in the Social Sciences and Humanities, where, due to publication delays, several years can pass between citing an article and having that citation appear in print. Hence only the authors of the article will be able to provide early citations to their paper as they are aware of it before it appears in print.
However, for both junior academics and academics with weak citation records, the low number of overall citations would already have provided a strong signal. Excluding self-citations wouldnt provide a lot of additional information. Only in very unusual circumstances would a more senior academic be able to game his or her citation record by consistently citing his or her own work.