Authorship inflation is unethical.

I thank Dr. Kakuk for his response to my article. Unfortunately, he seems to have misunderstood my arguments. He claims that I “fail to differentiate the problem of authorship from the problem of IFs usage”. In fact, I never actually directly linked these two problems, except to say that both trap researchers in a prisoner’s dilemma. Kakuk also claims that “The inappropriate usage of IF is not the only factor that hampers researchers to follow authorship guidelines”. But I never claimed that IFs make researchers fail to respect guidelines; my point was simply that both authorship guidelines and impact factors trap researchers in a prisoner’s dilemma. Kakuk further claims that I “obscure how IF is calculated and fails to fully understand why authorship guidelines are not followed”. With regard to the first of these points, I do not obscure how IFs are calculated; I simply say that they are discredited and provided a reference to this effect [1]. With regard to the second point, I have explored in depth in a previous article why authorship guidelines are not followed [2], including the problem of incentives. Kakuk is right about these problems, but focuses too much on the number of authors, when author position is also highly important. Having many authors on a paper is not really a “win-win”, as high prestige tends to attach only to the first and last authors positions. It can easily be seen as negative to be middle author among several on many papers. Kakuk’s suggestion of dividing the “score” for a paper by the number of authors has some merit, but it would have to be modified to reflect the importance attributed to author position. (Furthermore, it might be necessary to include extra authors who are not traditionally thought of as such. I have suggested elsewhere that peer reviewers and ethics committees might sometimes meet authorship criteria [3,4].) Most universities now have research integrity guidelines that define authorship criteria and specifically forbid ghost and guest authorship. Therefore, Kakuk is quite wrong to state that for researchers who add superfluous authors “Other than uneasy feelings, the presence of excess baggage on the author list is of little detriment to the real authors”. This is simply incorrect. They should be worried about real detriment, both because they are acting unethically and because they could face major sanctions from their institution, including potentially losing their jobs for scientific misconduct. It is therefore misleading to claim that “At present, if researchers decide to involve many authors who did not make a real contribution to the research or its publication, there is no real down side for them: everyone gets credit”. [Correction added on 3 November 2014, after first online publication: David Shaw was responding to an earlier draft of Peter Kakuk’s correspondence. The quotation “Only some committees and journals in their authorship guidelines are attempting to place some restrictions on who can be presented as an author of a publication” did not appear in the published version, and the text has been changed to reflect this.] Ultimately, Kakuk’s proposal could make researchers more interested in reducing the number of authors on a paper, but it would not prevent senior academics from applying pressure to add more authors. Ethical researchers already have an incentive to respect authorship guidelines, and his proposal simply adds one more self-interested reason to act ethically, which is probably insufficient for most researchers to escape the prisoner’s dilemma. The problem of IFs can be easily resolved by reduction of reliance upon them, but the problem of authorship attribution is much more difficult to solve, even with strong enforcement of journal and institutional authorship and integrity guidelines.