The authors build their thesis around a dichotomy between "toxic employees" and "superstars" (not "rock stars" as claimed in the article summarizing the study). The authors rather woodenly make their way through various analyses that make productivity the dependent variable, while also delving into the question of whether toxicity is contagious within workplaces.
Overall I get the impression that the study authors have never actually spent time in toxic workplaces.
The most toxic employees are not those who "[engage] in behavior that is harmful to an organization, including either its property or people" -- as another commenter mentioned, this is a comically tautological definition. Rather, toxic employees are those who convince their boss that they are a superstar while making the company a worse place to work. A toxic employee isn't, as the study seems to imply, somebody who writes mean graffiti in the employee bathroom -- it's someone who berates junior staffers under the auspices of demanding excellent execution, and who finishes the year with sales growth in his division up 23%.