This study addresses the two-way process in which a subordinate and a superior engage in influence activities (bottom-up) and favoritism (top-down) in subjective Performance Evaluation. The research context is the Chinese government's evaluation of Chinese state-owned enterprises (SOEs) by the State-Owned Assets Supervision and Administration Commission of China (SASAC). We analyze archival records of the government's evaluation scores, score adjustments, and evaluation ratings given to 63 SOEs between 2005 and 2007. These analyses are also interpreted based on insights gained from in-depth field interviews with SASAC officials and chief financial officers (CFOs) of SOEs. Results indicate that the political connection of SOE CFOs, the geographic proximity of SOE headquarters to the SASAC central office, and political rank of the firm affect the SASAC's evaluations. Data Availability: Data used in this study cannot be made public due to a confidentiality agreement.

We examine the structure of communications in accounting research by analyzing patterns of citations among authors who have published in five major journals between 1984 and 2008. Understanding communication structures is important because they shape academic knowledge creation, which prominent scholars have claimed has become narrowly focused and self-perpetuating in accounting due to a specific type of communication structure - 'tribalism.' We use a mathematical algorithm and other analyses to distinguish among five types of communication structures. We find that the field contains multiple clusters, with some clusters being centered on research topics alone, a finding consistent with a 'normal academic field.' Remaining clusters are more narrowly based – on combinations of topics, methods and theory bases – and all but one of them represent a “small world” structure because they are close together and exhibit frequent communication. Both normal academic fields and small worlds have been shown to contribute positively to innovation in research. The economics-based archival financial accounting cluster exhibits some properties of a tribal structure because, while researchers in other clusters communicate toward this cluster, the cluster sends most of its outbound communication to itself. A contribution of our study is that it shows that tribalism is not as rampant as previously suggested. Also, our findings suggest the field has become less tribal over time. Further, we identify 'hub' researchers who attract communications from multiple clusters and whose articles build on, and cite, work from multiple clusters. These individuals are instrumental in moving fields away from tribalism. Finally, we discuss possible determinants and consequences of the existing communication structure.