The joy of being in Florida is that I get a lot of US centric questions that I never get to hear in Canada or in the rest of the ROW!
While adding a senior co-author to a nearly finished manuscript has been around for quite a while (unfortunately in my mind) especially in the financial archival area, I am hearing that it is spreading to not only other archival realms but to other paradigms. This I think is troublesome on many grounds – first it leads to vita inflation for the senior authors; second it sends a very bad signal to more junior academics that merit of ideas can be trumped or at least supplemented by the name of the author (especially problematic one would think in non-blind review journals like JAR); third it suggests that all co-authors are not equal yet I rarely see the senior author who has been added late to the paper out of alphabetical order (indicating equal effort).
The solution is simple in my mind – have a clear statement, like much of the science literature has (and increasingly other areas of social science especially psychology), where the exact contribution of the author to the paper, including whether they were there at the beginning of the paper’s production or were added later, should be made clear.
Certainly there are valid reasons to add a co-author after the project has started. The need to access certain specialized data that was not part of the original plan, the need to add certain statistical expertise that was not envisioned as being needed by the original co-authors, the need to add additional experiments and hence access to new subjects that a co-author can provide, or the need to do some field research to understand a finding and the original authors are not comfortable with doing the field work that is needed even in a limited manner.
Personally, I do not recall ever joining a project after the empirical work was commenced. Nor have I ever added a co-author to one of my projects after the empirical work has commenced. I have added co-authors after the original idea was developed as I needed folks with special skills, special access etc Indeed, the only time I have joined projects after the analysis has been done is with former doctoral students who are having difficult times publishing their thesis papers. There, the added work beyond all the supervisory work at the beginning, makes it relatively easy for me to justify joining the project. After all I have shepherded it through many steps and provided numerous thoughts along the way. Getting deeper into writing the paper is just the last straw that in my mind breaks the camels back and leads to co-authorship. But this is a questions for another day – when do dissertation papers result in co-authorships with supervisors? While the US answer historically has been clear – not so much in the ROW. More later on this point.