Zhang Qiong ｜ Associate Professor of School of Public Administration and Policy of Renmin University of China
Recently, Associate Professor Zhang Qiong’s research paper, with partners from University of California (Berkeley), University of Hong Kong and University of Chicago, was published online by the American Economic Review, one of the top five journals in economics.
Titled “Subjective Performance Evaluation, Influence Activities, and Bureaucratic Work Behavior: Evidence from China”, this article looks at how subjective performance evaluation affects the behavior of Chinese civil servants at grassroot level and provides empirical evidence, which has been largely absent from the literature, on the existence and consequence of influence activities among these government employees.
The research was supported by the National Natural Science Foundation of China (project approval number: 72003188). Co-authors of this article include Professor Alain de Janvry and Professor Elisabeth Sadoulet of the University of California, Berkeley, Associate Professor He Guojun of the University of Hong Kong, and Assistant Professor Wang Shaoda of the University of Chicago.
Point of view| Influence activities and job performance—Want your subordinates to work well? Don’t let them know who evaluates their performance!
Due to the lack of consistent and comparable objective indicators reflecting individual work performance, subjective performance evaluation is very common in both the private and public sectors. However, subjective performance evaluation may open the door to “influence activities”, encouraging employees to spend too much of their time and energy on catering to the preferences of the evaluator rather than improving their actual work performance.
Influence activities refer to that, in addition to the widely mentioned ways of direct bribery or personal favors, subordinates spend more time and energy on influencing their superiors in charge of the evaluation in order to improve their performance scores. This kind of behavior may deviate from the goal of the organization, but the behavior itself has some practical significance. For example, the subordinates may allocate more efforts to tasks assigned by the evaluator, try to get more involved in projects initiated by the evaluator, or work harder on job dimensions that are more observable or valuable to the evaluator. But sometimes, this kind of behavior can be meaningless. For example, subordinates may allocate a lot of energy to butter up the evaluator or deal with many other matters besides work for the evaluator.
Although a sizable body of theoretical literature has discussed the formation and consequences of influence activities, there is very little empirical evidence to back it up. This is mainly because these behaviors are not easy to observe, and even if they are observed, it is difficult to infer that they are driven by intentions of improving evaluation outcomes (rather than simply being hardworking or friendly), making it difficult to classify them exclusively as influence activities.
By analyzing different performance evaluation schemes, the paper studies the existence and impact of influence activities. It finds that influence activities are prevalent in China’s local bureaucratic system: when subordinates know the identity of their evaluating supervisor, they will spend more energy to influence the evaluating supervisor in their favor, which leads the evaluating supervisor to give them a higher work performance score than the non-evaluating supervisor. When subordinates are not sure who evaluates them, they will work harder, which ultimately leads to a higher evaluation score of their work performance by all parties.
Specifically, the research team analyzed two different evaluation schemes. In the “revealed” scheme, subordinates were informed about the evaluating supervisor’s identity at the very beginning. Under this scheme, the supervisor chosen to be responsible for the evaluation gave significantly more positive assessments of the subordinates than the non-evaluating supervisor. At the same time, subordinates devoted more efforts to tasks assigned by the evaluator, deemed the assignments from the evaluator as more important, and their work performance improved more in areas that were valued more highly by the evaluator.
In the “masked” scheme, to introduce uncertainty about the evaluator’s identity, subordinates were not informed about the identity of the supervisor responsible for their evaluation, and the evaluator was randomly picked at the end of the evaluation cycle. Under this scheme, no significant difference was found in the evaluation outcomes between the two supervisors. This shows that once subordinates do not know who the evaluator is, their influence activities will be significantly reduced. More importantly, subordinates under the “masked” scheme chose to work harder and devote more time and energy to job dimensions deemed important by the organization (both supervisors), which corresponded to better performance outcomes under a series of indicators (such as supervisor assessment, co-worker assessment, performance pay, overtime, etc.).
The study benefited from the opinions and suggestions of many scholars and experts and was strongly supported by local governments in China. More than 800 villages and towns were involved in the two evaluation schemes. The findings of the research not only provide an important decision-making basis for the design and management of performance evaluation system of various organizations, but also highlight the institutional advantages of China’s dual-leadership governance model. Under the dual leadership of the “party leader” and the “administrative leader”, civil servants devote more energy to aspects beneficial to the overall organization, rather than catering to the preference of a single leader. In addition, to give full play to the checks and balances embedded in the dual arrangement, a random mechanism can be introduced in the evaluation for grassroot civil servants to improve the operating efficiency of the entire system and better achieve organizational goals.