Identifying the value of teamwork : application to professional tennis

Do workers vary in their ability to work with others? I compare a given worker's productivity in solitary production to their value-added to team production to identify team skills: a worker's contribution to team production above and beyond that given by general skills. The identifying assumption is that workers use general skills in both production functions, but team skills only in team production. Professional men's tennis provides a useful setting to compare solo work (singles) to teamwork (doubles). I find that around 50% of variation in team output is explained by team skills. This is robust to a variety of specifications, including nonlinearities in player inputs. Players sort positively-assortatively along both skill dimensions, yielding indirect returns to skills of about half the magnitude of the direct returns.