orgtheory.net

Brayden

Tyler Cowen and Andrew Gelman offer some thoughts on coauthoring papers. As someone who has frequently coauthored, here are my thoughts and advice on the topic.

• Rates of coauthorship – It’s clear from these discussions that the disciplines have different coauthoring norms. In economics, coauthoring is probably less common than in political science or sociology, which are both data-driven disciplines. Management theory has a higher tolerance for coauthored papers than any of the social science disciplines. Like Tyler says, empirical papers are much more likely to be coauthored than theory-only papers.
• Productivity – Having a coauthor can be productivity-enhancing or it can be a big drag. On the productivity-enhancing side, coauthors can divide their labor to improve efficiency, provide continual checks on the other’s logic, help fill theoretical gaps held by either author, and collaborate to improve the quality of the final draft. On the productivity-destroying side, coathors who don’t have compatible schedules, work habits, or skill sets may actually impede the ability of either author to get anything done. My advice is figure out how compatible you’d likely be with a potential coauthor early on and then make a decision about how much time (if any) you’ll commit to doing a project with that person.
• Diversity – My sense is that having diverse coauthors is beneficial inasmuch as two or three coauthors with very different skill sets, knowledge, etc. will be able to offset the weaknesses of the other. But coauthor diversity probably has a curvilinear relationship with productivity. At a certain point, the level of diversity can push coauthors into having continual misunderstandings and either one author will take the project over and dominate or a final paper will never materialize.
• Number of authors – My experience and conversations with others tell me that the ideal number of coauthors is two and that any number greater than three is too many. Projects with more than three authors promote free-riding and the first author usually gets stuck doing most of the work. Part of the problem is that it’s much more difficult to figure out how to divide labor among four people than it is among 2 or 3. That’s not say that I wouldn’t sign on a project with more than three authors, but you have to realize that you’ll either be doing the majority of the work  or you’ll be a free-rider.
• Order of authorship – I’m not sure if there even is a norm about coauthorship order in sociology. My experience has been that whoever came up with the idea and is willing to take a first stab at writing a first draft of the paper ends up being first author. If more than two authors are in the mix, deciding the subsequent order is usually negotiated before a final draft is finished. This ordering may change if one person ends up committing more time than another, but the negotiation should be done openly. Whoever is the lead author should be sure to figure out author ordering early on so that problems don’t erupt later.
• Number of coauthored projects – This is a tricky one and I’m not the best person to offer advice on it (yet). Some would say that coauthoring early on in your career makes you look like a perpetual RA. Others would say that you can get away with it if there’s a clear “you” in all of the papers. My word of caution to grad students who are thinking about writing multiple coauthored papers with a senior scholar is to make sure that they find someone with whom there is a sense of equality in the coauthoring relationship. You want to develop your own professional identity and joining a team project that churns out incremental analyses will likely inhibit your ability to do that. Feel free to coauthor but make sure you’re doing things on your own too and try to find multiple coauthors with whom you can write papers that have similar themes or that contribute to the same intellectual conversation.