Abstract
Existing literature on national governance models has focused on the analysis of long-term, stable public organizational processes between different levels of governmental and social organizations. In contemporary China, a considerable part of the organization and implementation process in public services relies on cooperation among different local governmental institutions and social groups. This type of process is characterized by short-termism and instability. This paper uses the perspective of the “control rights” theory to analyze the relationship between the three parties: the principal, management, and the implementation agency in the case of the phenomenon of grading criteria variation in the grading of history as a subject in the gaokao (college-entrance examination) of Province X. This paper shows that although the relationship shows a high degree of correlation, the three parties do not belong to the same bureaucratic organization and lack administrative oversight within the process, which increases the uncertainty in negotiation and maneuvering, resulting in two issues: First, the principal party and management party often have divergent views on targets. With the advantage of controlling incentive distribution, the principal party is able to involve itself in the inspection and evaluation of policy implementation, and therefore maintains the ability to arbitrarily intervene in the process. Second, implementation agency behavior is constantly influenced and modified by feedback from the principal party and the management party, and vice versa. In the process of continuous feedback and adjustment, the three parties gradually reach their own shared understanding of policy implementation that becomes the cause of local variation in grading standards. This paper suggests that unstable public organization process is an important area of study on contemporary Chinese governance. Control rights theory can be further explored as an analytic tool and strategies of various social forces in gaining organizational control should also be investigated in depth.
Introduction
Since 2000, the gaokao (college-entrance examination) system has been continuously reformed by the Chinese government with the aim of weakening the dominant role of the gaokao in high school education. On September 4, 2014, the State Council issued Implementation Opinions on Deepening the Reform of the Examination and Enrollment System, which called for reform to the gaokao system and assigned several provinces and cities to pilot the reform plans. “The total grade of the examinee is composed of the grade in Chinese, mathematics, and foreign language in the unified gaokao, and the grades of three subjects in the high school academic-level examination” (General Office of the State Council, 2014). In September 2014, as one of the pilot provinces, the government of Province X (2014) launched a regulation named Pilot Program for Deepening the Comprehensive Reform of the College Examination and Enrollment System in Province X. It stipulates that Chinese, mathematics, and foreign language are three required subjects of the gaokao. Additionally, examinees select three subjects from the following: ideology and politics, history, geography, physics, chemistry, biology, and technology (including general technology and information technology), as elective subjects in the high school academic-level examination, in order to combine the gaokao and high school academic-level examinations (gaozhong xueye shuiping kaoshi). On November 7, 2014, the department of education of Province X issued Measures for the Implementation of Academic-Level Examinations for Ordinary High Schools in Province X and Measures for the Implementation of the Examination of Selected Subjects in College Enrollment, which signaled the implementation of national policy in Province X.
According to the new gaokao and enrollment regulations in Province X, candidates can determine their text subjects independently according to their own interests and the requirements of their targeted majors in the colleges/universities to which they intend on applying. They will have two examination opportunities for each selected subject. From 2015 to present, Province X has formed a new gaokao pattern with selective examinations (xuankao) and academic examinations (xuekao) as the core. The amalgamation of academic-level examinations and the gaokao in Province X has changed the grading of history subject exam papers in Province X from once a year to twice a year, each time requiring about a week, in the spring and autumn respectively.
University A is a comprehensive university emphasizing normal education. Its educational target is training high school teachers. Since 2015, the history department of University A has been carrying out the task of grading the history subject test papers of academic and selective examinations in Province X. I have participated many times in the entire process of grading as leader of an assessment small group (pingjuan xuekezu xiao zuzhang) in the grading of history. In my experiences, the unified grading criteria, which should remain the same, have been changed during the whole process. At the beginning, the grading criteria were swinging between “grading precisely” (caidian geifen) and “grading roughly” (xiangjin geifen). “Grading precisely” means the criterion is closely linked to the knowledge points in the reference answer, while “grading roughly” means the criterion is broad and vague. In the middle and late stage of grading, the criterion would change to “grading roughly”.
For individuals, the gaokao is an important event and an inflection point in their life course. For society, the gaokao’s authoritativeness and fairness has made it the default system by which educators and the educated abide. Why, then, has the grading criteria always had similar variation in the history subject grading process in Province X since the new gaokao reform? Why is it difficult to carry out the “grading precisely” criterion, and why does this always give way to the “grading roughly” criterion? To answer these questions, it is necessary to clarify the governance mode and action strategy of the new gaokao grading in Province X.
In Province X, the education and examinations authority (jiaoyu kaoshiyuan) is a public institution (shiye danwei) with administrative functions which are directly affiliated to the department of education. Its responsibilities include “organizing and executing the ordinary gaokao in Province X (including unified examination subjects and selective examination subjects)”. The task of grading the new gaokao is the responsibility of the education and examinations authority, and is implemented by the grading institution (pingjuandian) (Education and Examinations Authority of Province X, 2015). A relationship of entrustment is formed between the education and examinations authority and the grading institution. Given this background, this paper introduces “control rights theory”, a concept which has attracted much attention in the study of national governance in recent years, to analyze this problem.
Literature review
With the subtle adjustment of the relationship between government and society, informal operations, strategic applications, and adaptive practices in the process of public organization have been widely studied by scholars (Ouyang, 2011; Zhou, 2013). Li-An Zhou (2014a: 8–9) adopts the subcontracting and employment systems in enterprise theory (Coase, 1937; Hart and Moore, 1990) in proposing the concept of the “administrative subcontracting system” and the three analytical dimensions of authority relations, economic incentives, and internal control.
By constructing an analytical framework with two dimensions of vertical contracting and horizontal competition (Zhou, 2014a: 6), the concept of the administrative subcontracting system has the potential to dominate our understanding of informal operations and flexible strategies in government governance. However, there is still much room for discussion on the connotations, extensions, and analytical dimensions of this concept. First of all, the basic logic for explaining the formation of the administrative subcontracting system is the economic hypothesis of cost–benefit measurement. This “overly general level of interpretation” “could almost explain the formation of any governance model” (Zhang, 2014: 87). Compared with a clear delineation of mandate and responsibilities between the subcontractor and the contractor, Zhang (2014: 85–96), noting several recessive characteristics of the administrative subcontracting system (diversification of the government’s role, group relations between multiple centers of control, and dependence on execution tools), puts forward the concept of the “administrative contracting system”, reflecting that the present management system’s use of existing organizations and social relations, within the scope of the fuzzy control rights, through acquiescence, exchange and invisible authorization, forms a mixed symbiosis model that accommodates organizations with different goals and interests. Second, the administrative subcontracting system focuses on the “contract awarding” actions of higher-level government bodies, emphasizes the power distribution between the upper and lower levels within the administrative organizations (Zhou, 2014b: 109), and pays insufficient attention to the “contract” mechanism and operational strategy of grassroots governments. Therefore, the administrative subcontracting system aims to investigate the relationship between government levels, while the administrative contracting system aims to investigate the internal operation process of grassroots governments at the city and county level (Yang and Yuan, 2017: 183–186; Guo, 2015). The former focuses on the “clear” boundary at the level of institutional norms, while the latter focuses on the “fuzzy” zone at the level of actual operation.
Although the administrative subcontracting system and the administrative contracting system complement each other in terms of problem focus, they both analyze the governance model under a static framework, and pay insufficient attention to the type of deviation and evolution process of this mechanism in the process of operation. Zhou (2016: 34, 45–46) redefines the difference as one between intra-administrative outsourcing and extra-administrative outsourcing. The key to distinguishing between extra-administrative outsourcing and intra-administrative outsourcing is whether the contractor and the principal are in the same bureaucratic organization’s system of power and promotion. Huang and Zhou (2017: 122–123) note the influence of the multi-hierarchical nature of the administrative contracting system on its operational modes: “administration” and “contracting”, the two kinds of mechanism, have different configuration requirements in the context of executive governance, internal party control, and incentivization; having failed to reach a proper equilibrium, the administrative contracting system engages in cyclical swings between the “administrative” and “contracting” poles, forming “administrative”- or “contracting”-dominant governance models, respectively. Xueguang Zhou (2017: 94) believes that the administrative subcontracting system focuses on matching and compatibility among the three elements of executive power, economic incentive, and internal party control, and does not involve the authority relationship and the distribution of control rights among the principal party, the management party, and the implementation agency. The differential distribution of control rights determines the relations of authority and governance patterns within a given organization. Xueguang Zhou (2017: 95–104) divided such control rights into three categories: target-setting rights, inspection and acceptance rights, and incentive-distribution rights. According to the different distribution modes between the employer, contractor, and the implementation agency, he proposed four corresponding governance modes: highly correlated, administrative subcontracting, loosely correlated, and federative.
Control rights theory broadens the applicable radius of the administrative contracting system theory, which has great benefits to understanding the diversity and complexity of current national governance models. However, empirical studies using this theory are not abundant, and as such there remains great scope for employing this theory in understanding different types of governance models at present. First of all, both the administrative contracting system and control rights theories focus on the analysis of the long-term and stable public organization process between governments at different levels and social organizations. In contemporary China, a considerable part of the organization and implementation of public services depends on cooperation between public institutions belonging to different systems in a given administrative region. The delegation and management of education, health, environmental governance, poverty alleviation and relief, and so on are all grassroots administrative matters (Li-An Zhou, 2017: 61). Such processes are often characterized by short-term and unstable characteristics. China’s public institutions are not public administration authorities, and do not belong to the same bureaucratic organization system. They do not have the function of administrative management of other departments and institutions within the administrative system, nor does there exist a relationship of hierarchy between leaders and subordinates, nor is there the possibility of the personnel promotion, nor is it possible to use market mechanisms to produce contractual cooperative relations between them. Although the principal party has formal authority, this authority cannot be realized in the manner of a relationship of command, as between superior and subordinate, within the bureaucratic organization. Therefore, the principal party lacks administrative oversight over the management party. This makes it possible for both the principal and management parties to adopt action strategies based on their own interests, thus increasing the uncertainty of negotiation and maneuvering between them.
The business of public institutions is mainly the provision of specific public services developed from government functions. This means that public institutions have both professional and technical authority and a public function. The organizational process of public services involving different public institutions is usually led by the central and local administrative institutions, while the frontline business institutions are responsible for specific implementation. In terms of the right of formulation and interpretation of professional technology and industry criteria, the administrative institutions are generally responsible for the results of such transactions, and provide process supervision and business guidance to the frontline business institutions, in the form of contract awarding or entrustment. If the organizational system and market mechanisms are the key elements in determining the outsourcing system, professional and technical authority is the key element in determining the governance model of public institutions. According to Xueguang Zhou’s classification of governance models, it can be assumed that in the unstable public organization process involving the coordinated participation of public institutions, the principal party will give up the highly interrelated model of centralized operation and take the way of transferring part of the control right to the management party to arouse the latter’s enthusiasm. The ideal type of governance model should be between the contract type and the loosely related type that is closer to decentralized operation.
Second, in Xueguang Zhou’s analysis (2017: 103) of the allocation of control and corresponding management modes, the dimension of control right which is gained by the management party is the key to shaping the governance model, incentive allocation is a necessary condition for formation of relations of control, and the right of inspection and acceptance and the right of target setting are sufficient conditions which orientate different governance modes. The question is whether there is a possibility that management does not have the right to incentivize distribution, but does have the right to inspect and accept. This paper argues that the right of inspection and acceptance can be divided into the right of inspection and the right of acceptance. The former is exercised multiple times during governance, while the latter is exercised once at the end of governance. Different types of governance work assign different weights to processes and results. In results-driven work, the right of acceptance is substantive; in process-driven areas such as education and healthcare, right of inspection is much more meaningful and the right of acceptance is relatively minor and symbolic. The division of inspection and acceptance rights means it is possible for the management party to grasp the right of inspection or acceptance but not to grasp the right of incentive distribution.
Third, the theories of subcontracting and control rights both focus on the subcontractor (the principal party) and the contractor (the management party). Xueguang Zhou (2017: 105) noted the phenomenon of “collusion” between the management party and the implementation agency, but believed that this phenomenon was more likely to occur under the administrative subcontracting system. Under the highly correlated governance model, collusive behavior is expensive and rarely occurs. Under the loosely connected and federalist governance models, there is no incentive for the management party to collude with the implementation agency. However, this analytical perspective, which focuses on incentive mechanisms and rational choice, does not pay enough attention to the implementation agency’s own subjectivity and behavioral preference. Although the implementation agency is at the bottom of the subcontracting system and is subject to the dual oversight of the management party and the principal party, it does not mean that the implementation agency’s own behavioral preferences will not have an adverse impact on management and the principal. In the process of task implementation, the principal party, the management party, and the implementation agency are often constantly coordinating and maneuvering according to the work and targets, which makes it possible for the implementation agency to have an adverse effect on the principal and management. The latter will constantly correct the behavioral preferences of the implementation agency and form a governance loop with both positive and negative feedback regulation mechanisms.
It can be seen that although the theory of control rights is an important perspective for understanding the public organization process, it is necessary to further clarify the possible dimensions into which control rights theory can be divided. Bringing unstable public organization processes into the investigation is not only helpful for expanding the applicable radius of control rights theory, but also helpful for enhancing the understanding of contemporary Chinese national governance issues.
Case introduction and research framework
Since 2015, the basic process of grading the history exam paper in Province X’s gaokao is as follows: First, the education and examinations authority determines the total number of high school teachers participating in the evaluation work according to the total amount of examination papers to be graded, distributes the numbers to the relevant high schools in Province X according to the indicators, and issues the mobilization documents of the grading task to the subordinate institutions (Education and Examinations Authority of Province X, 2017a). Second, the grading institution set up the grading leadership group, with subject’s grading group as its subordinate branch, the department’s administrative head and subject’s responsible dean act as the subject’s grading group’s leader (da zuzhang, “the big group leader” hereafter). According to the subject’s situation, the head of the big group should divide the grading teachers into several small groups, appoint small group leaders, and determine the grading quota (Education and Examinations Authority of Province X, 2015). Usually, one week before the official start of the grading task, the history department of University A will hold a grading task mobilization meeting to complete the preparation of the grading group and the division of tasks. The department head and the subject’s dean act as the head and deputy of the grading group. The former reads the grading rules, explains the matters needing attention, and is responsible for the overall coordination of the grading task. The latter unseals the sealed sample exam papers, reference answers, and other documents issued by the education and examinations authority, and carries out collective discussion, distributes questions to be graded to each small group, and determines the number of grading groups according to the question types in sample exam papers and the total number of the exam papers. Each group usually grades a test question on a 4–6-point scale. The dean responsible for the subject of the examination determines the total task of every group according to the total amount of the papers to be graded and allocates the grading teachers equally to every group determined by the education and examinations authority. The small group leaders are allotted equally among the teaching and research offices of the history department. According to the small group leaders’ own professional emphasis, after obtaining their consent, assign the relevant grading question, define the rights and responsibilities, and complete the subcontracting before the grading task starts.
As soon as all the grading teachers are in place, the grading task will officially start. All the grading teachers first receive a half-day training. After that, they enter the computer room of the Information Center of University A. Under the guidance of the staff, they will enter their respective workstations. The big group leader announces the grading rules and matters needing attention to everyone. The small group leaders check the members of their own group, implement the grading questions for them, and instruct the members on familiarizing themselves with the grading operating system. Grading is divided into a two-hour “trial” stage, followed by the “formal” grading stages. In the trial grading stage, the exam paper graded by the teacher is from a real examinee, but the grade is not official. The purpose is to make the grading teacher familiar with the answers, and to clarify the grading criteria and evaluation scale. After the trial stage, all the grading groups enter the formal grading stage. Because the grading criteria have been made clear in the former stage and cannot be changed during this stage, grading teachers should no longer question the grading criteria. During this stage, the grading teachers must accept the instructions from their own small group leader and external intervention from the big group leader of the grading institution and the inspectors of the education and examinations authority. When the formal grading stage finishes, the inspectors of the education and examinations authority, the grading teachers, and the members of the grading institution gather and convene the concluding meeting of the grading task. After the meeting, the grading teachers receive their remuneration uniformly and return to their work institution. The grading process is complete.
It can be seen that during the grading process of history subject in Province X’s gaokao, the education and examinations authority is the principal party, who is responsible for the grading task. The grading institution is the management party, who has accepted the task from the principal party, in this case to carry out the specific grading task. Grading teachers are the implementation agency, who accept the supervision of the grading institution. The education and examinations authority, the grading institution, and the high schools which the grading teachers belong to are all public institutions. Their main functions are that of educational public service. They do not belong to the same bureaucratic organizational system, so the grading progress is an example of an “unstable public organization”, in that it requires participation and coordination between different public institutions. But in the grading progress, these three parties—the education and examinations authority, the grading institution, and the grading teachers—gather in one place and are externally isolated, as such they complete the grading task in a closely related, highly interactive mode. This aspect contributes to the formation of a “highly correlated” governance mode, rather than “administrative subcontracting” or “loosely correlated” modes.
The author has participated in grading selective examinations and academic examinations in history in Province X since 2015. Based on this experience, I can not only grasp all kinds of problems in grading teachers’ work, but can also get information and suggestions from regulators such as the education and examinations authority and the grading institution, so as to get deeply involved in the whole process of grading. I used participant observation; made an ethnographic written record of previous grading processes; interviewed relevant stakeholders including grading teachers, the education and examinations authority, and the grading institution; and collected relevant policy documents on the new gaokao reform. Based on these materials, this paper intends to use control rights theory to analyze the organization and implementation process of the grading process of history in Province X’s gaokao, so as to understand the internal logic of the variation of grading criteria. In terms of framework design, the dimension of control rights and its allocation in the grading process are first analyzed. Second, the mechanism of variation in grading criteria is analyzed by examining how the behavioral preferences of the implementation agency have reverse effects on the principal and management parties. Finally, I summarize the model characteristics and influences of this kind of public organization process.
Distribution of control rights in the grading process
The difference of target setting between the education and examinations authority and the grading institution
Under the administrative subcontracting system, the contractor has discretionary power: it executes the specific tasks assigned by the subcontractor and also controls the upward transmission of relevant information, which leads to the superior subcontractor’s assessment and control relying on the final result (Zhou, 2014a: 9). The contractor can pass the examination only when its work objective is completely consistent with the subcontractor’s result objective. However, in the grading process, there is a difference of target setting between the principal party and the management party. After grading the examination papers, the education and examinations authority should report the grading results to the department of education, which is a functional department of the government, and answer questions raised by the examinees, parents, and even other social organizations within the prescribed time. The average score and the “third grading rate” (sanpinglü, see below) are the two indexes that the education and examinations authority are most concerned about.
Under the restrictions of the unified examination system, most of China’s provinces no longer distinguish the high school academic-level examination from the gaokao (Dong and Feng, 2013: 63). The education and examinations authority of Province X adopts the unified organization of examination papers and grading for the academic-level examination and the selective examination in the gaokao. In this way, the grading criteria of the academic-level examination, which aim to emphasize the passing rate, conflict with that of the gaokao, which aim to emphasize selectivity. This poses constraints to the education and examinations authority and grading teachers. In the view of one of the inspectors of the education and examinations authority of Province X, if the average score is moderate and distribution normal, this clearly indicates that the examination paper’s topic is successful, because the examination itself has succeeded in its two functions of maintaining the passing rate and creating the conditions for selection, which is good for students, parents, and society. If the average score is too low, it will affect the passing rate of the academic-level examination. However, if the average score is too high, it is not conducive to the selection function of the examination (interview subject S5, inspector of the education and examinations authority of Province X, April 22, 2016).
The current grading system in Province X adopts the method of multiple people grading the same test question to reduce the grading deviation. According to the Detailed Rules for the Grading Task of Selective and Academic Level Examination Papers in Province X, “the principle of grading scores shall be determined by the subject’s grading group, and the ‘two-grading system’ shall be implemented. The threshold range is controlled between one-sixth and one-tenth of the total score, and the average of the two scores is taken within the threshold range. If the scores of the first and second assessments exceed the threshold range, the group members with the final arbitration authority will make a third assessment and take the average of the two scores that are close to each other within the two threshold ranges” (Education and Examinations Authority of Province X, 2015). The computer grading system sets the score within an integer range, eliminating the presence of decimals. A threshold deviation of 1 point is allowed only if the score of a test question is above 6 points. When the test question’s full score is 5, if Teacher A gives 5, Teacher B must also give 5 to pass. Since it is impossible to give a score between 4 and 5, any score given by Teacher B below 5 will exceed the threshold range of the test question and must be assessed by another grading teacher for the third time. This “third grading” situation is more common when grading teachers have different grading criteria. Therefore, the “third grading rate” of a given test question is an important basis for reflecting whether the grading group members have unified the grading criteria of this question. A low “third grading rate” indicates that the grading criteria is unified throughout the grading process, which is an important embodiment of the fairness of the examination.
Unlike the education and examinations authority, University A, as the management party of the grading task, is not responsible for the grading results and does not need to “explain” the results to the higher education authorities, students, parents, and other social organizations. With a fixed total amount of grading to be undertaken, the history department of University A tends to shorten the working period and reduce the continuous consumption of human and material resources. The time quota set by the education and examinations authority for the grading task is two weeks. However, in previous mobilization meetings, the dean of the history department and the responsible dean of the subject reduced time for the grading task from two weeks to 5–7 days, and stressed that it was better to finish the grading task as soon as possible. This means that the actual grading time is usually half of the period stipulated by the system, or even shorter.
As the grading implementation agency, the grading teachers are also not responsible for the grading results. When the grading task is completed, the contracting relationship terminates. Although the salary during the grading period is paid by the education and examinations authority, “the grading teachers’ accommodation expenses, transportation expenses, travel expenses, and other expenses are reimbursed when they return to the original work unit” (Education and Examinations Authority of Province X, 2017a). The units in which the grading teachers work not only bear the above costs, but also deal with the impact of teachers’ absence on the routine teaching order. The salary of high school teachers is directly linked to the academic performance of their students, which makes them want to complete the grading process as soon as possible and return to teaching work. Therefore, “being in a hurry” (gan shijian) becomes the same target orientation for both the management party and the implementation agency in the grading process. They all hope to end this “drudgery” as soon as possible. During every stage of high school teaching, every lesson has a strict schedule. If you are delayed by something, it means that the rest of your teaching order has to be completely disrupted. Our teaching is performance-based, so not only do schools hate it, but you don’t want it to happen either. High school teachers usually have a heavy teaching load, so they are usually reluctant to ask for colleagues to cover for them. Moreover, each teacher’s teaching style is different, and the grasp of knowledge points is not completely consistent, which is not conducive to students’ acceptance. Even if your colleagues cover for you, you have to supplement their work when you return. (Interview subject S15, grading teacher, November 10, 2016) A high “third grading rate” means that a significant portion of the task is done for nothing. If the test question’s grading criteria has been grasped well, the “third grading rate” can usually be kept below 5%, and no higher than 10% or so. The total number of examinations to be graded in history is in the tens of thousands every time. For a test question, if the “third grading rate” is around 40%, it means that your group members have to grade 40% more. It’s a lot of work, and it’s futile and thankless. (Interview subject S12, the subject dean of the history department of University A, April 22, 2016)
The division of the right of inspection and acceptance
Although there is a possibility that the management party will be absorbed in the target-setting process, the will of the principal is dominant in all the governance modes, including the highly correlated, administrative subcontracting, and loosely correlated modes. Therefore, when there is a conflict between the management party and the principal party in the target-setting process, the principal party needs to deal with this conflict and make the policy-implementation process in line with the expected target setting. In the administrative subcontracting system, the right of inspection and acceptance is inseparable and controlled by the subcontractor. For the education and examinations authority, although the grading task can be entrusted to the grading institution for specific implementation, only when the grading work is finished and the grading results are officially released will it mean that the government and the public officially start to check and accept the reliability and validity of the educational examination and accountability. The problem is that the grading results are automatically frozen after the grading task has finished and cannot be changed. In order to keep the grading result from “deviating” and reach the expected target, the education and examinations authority divides the right of inspection and acceptance into the right of inspection and right of acceptance by maintaining real-time monitoring and correction of the grading process and controlling the average score and “third grading rate” within the ideal thresholds.
In the process of inspection and acceptance, one of the most prominent problems is the control of information and resistance to the control of information (Xueguang Zhou, 2017: 221). Compared with the information-filtering mechanism from the subcontractor to the contractor in the administrative subcontracting system, the grading information monitoring system forms an inverted triangle. The education and examinations authority, the grading institution, and grading teachers are in the same spatial dimension. A sequence of numbered computers isolates the grading teachers, each focused on the screen in front of them. The small group leader makes irregular inspections in the space in which the group is distributed. The inspectors of the education and examinations authority and the leader of the big group tour circulate and control the overall situation, which constitutes a spatial structure of the panopticon in the sense of Foucault (1997, 1999). Under this spatial structure, the principal party and the management party share the information advantage and have the maximum information authority. The responsible dean, inspectors from the education and examinations authority, and inspectors from the educational affairs department of University A have the maximum authority to monitor the grading information. The special monitoring computer not only records quantitative indicators such as achievement ratio, amount of unfinished exams, average score, average speed, and estimated time, which are used to monitor the overall working state of the grading groups, but also records quantitative indicators such as effective degrees, average speed, inefficiency ratio, and standard deviation, which are used to monitor every grading teacher’s working state. Each small group leader only has the power to supervise information on their group, while the grading teachers can only see their own grading indexes and cannot even check the index data of other members of the group. Therefore, the spatial and information hierarchy structure of the grading environment endows the education and examinations authority with the dynamic monitoring ability of the grading process equal to the grading institution. Both of them can monitor the information dynamics of primary grading groups at any time, and issue instructions directly to grading teachers with the help of information advantages, forming an operational model in which the principal party and the management party jointly control the agency. The key link of this model is to divide the right of inspection and acceptance into the right of inspection and the right of acceptance.
Because the grading progress is time consuming and labor intensive, all three parties must avoid the big error of rework. For the education and examinations authority, it is necessary to monitor the dynamic indexes of the grading institution and grading teachers. Therefore, the inspector of the education and examinations authority shares the right to inspect the grading process with the grading institution. When the monitoring information shows that the average score of a test question is too low, the inspector will communicate with the small group leaders and suggest relaxing the grading criteria to raise the average score. However, the relaxation of grading criteria will bring about the problem of inconsistent understanding of grading criteria by grading teachers, leading to the decrease of average speed and a surge in the “three-grading rate”. As the principal party, the education and examinations authority is always responsible for the grading results, and the big group leader is not willing to openly disagree with the inspector. However, administrative personnel are usually loyal to their own organization, rather than the rules and principles of the administration as a whole. They represent their own organization but not the overall governance system (Zhang, 2014: 91). After the intervention of the inspector, the big group leaders often communicate with the small group leaders privately, and negotiate an alternative method that neither violates the opinions of the inspector nor slows down the grading speed, instructions which are then passed on to the grading teachers.
The right of incentive distribution led by professional technical authority
According to Xueguang Zhou (2017: 106), the transfer of incentive-distribution right between the principal party and the management party is the direct reason for the transformation of the highly related governance model to other types of governance model. In the grading process, the total grading expenditure, the total remuneration of the grading institution and grading teachers is allocated to the grading institution in a lump sum after the education and examinations authority has fixed the quota. The above budget plan has been formulated and submitted for approval by the education and examinations authority before the start of the annual grading task. This kind of quota-based funding and allocation method means that the grading institution has no residual claim or strong incentive, and the education and examinations authority cannot monitor the grading institution’s work performance by sharing with the grading institution or paying the “balance payment”. Instead of adopting the piecework system, which links remuneration to the total quantity of examinations graded by the grading teachers, the grading institution decides the basic remuneration of the grading teachers according to the budget quota and the number of grading teachers. At the same time, the grading teachers are given a performance salary of several hundred yuan according to the relative difference in their respective workloads. Teachers assigned from the history department of University A are paid the same as the average remuneration of grading teachers. This incentive distribution scheme obviously cannot arouse the enthusiasm of the grading institution and grading teachers. Compared with the strong incentive of intra-administrative outsourcing and administrative outsourcing, both the education and examinations authority and the grading institution seem to have given up the right of incentive distribution.
However, this representation only applies to a single grading process. The delegation of repeated grading tasks, and even long-term delegations of grading tasks, is the core to understanding the grading governance pattern. The process of issuing a grading task to a given grading institution, designated by the education and examinations authority, contains a competition mechanism, but this competition is not based on the market competition mechanism of public bidding, but an internal competition controlled by the education and examinations authority, and the competitors are limited to a few colleges and universities in Province X. The education and examinations authority allocates the corresponding grading subjects in accordance with the expertise areas of these institutions. Resource endowment, social networking, and leader preference are all important factors in determining which institution becomes the grading institution. Before the launch of the pilot program on gaokao reform, the grading task in history, for both the high school academic-level examination and the gaokao, in Province X was undertaken by University B, another comprehensive university in the province. As a normal university specializing in the humanities and located in the provincial capital city, University A intends to develop its right to influence the basic subjects of literature and history for secondary schools in the province. After 2015, however, University A has been assigned as the grading institution for the two basic subjects of Chinese literature and history. During a grading mobilization meeting, one leader of University A said bluntly that “we snatched the right to grade these two basic subjects from University B”. In his opinion, “snatching” the grading right was crucial to University A’s status in the province. “Our teachers have already been able to participate in setting these subjects’ examination test questions every year. Now that we have obtained the right to grade, we will have a comprehensive grasp of what to teach in these subjects, what to test, and what are the grading criteria. It is beneficial for us to improve our ability to control the teaching in high schools in the province. Our teachers can be of guiding significance to the provincial high schools in terms of curriculum and teaching methodology. Our graduates will also be valued more by provincial high schools, especially key ones, and their career prospects will be greater, which has an important impact on the quality of our university.”
For the grading institution, only by obtaining the grading qualification can they improve their discourse power and resources in relevant subjects. The more times the grading task is done, the greater the space and capacity for improvement will be. Although a single grading task has neither internal incentive based on personnel promotion nor external incentive based on market mechanism, the reason why the grading institution is willing to strive for the right of grading is based on the incentive of maintaining the claim with the education and examinations authority for subsequent grading tasks. The education and examinations authority has the professional and technical authority to organize educational examinations. This authority gives it the ability to choose the grading institution, and this is the incentive distribution power held by the education and examinations authority as the principal party. By virtue of this power, the education and examinations authority has a relatively stable formal authority and business guidance relationship with the grading institution, and generally controls the grading task. Since University A can “snatch” the right of grading from University B, other universities can also “snatch” the right from University A. As the principal party, the education and examinations authority has the final right to decide which university will be assigned grading tasks.
In the administrative subcontracting system, the subcontractor and the contractor are strongly constrained by the organizational institution and the code of conduct. It is easy for them to form a highly stable and even normalized behavior model and interactive relationship. Although public services such as education and examinations and public health are intermittent, abrupt, and short term, they usually have a specific cycle and frequency. Although the education and examinations authority has the incentive and distribution right with regards to assigning the grading institutions, it is not easy to use this right considering the regularity and stability of education and examinations tasks. As a result, there are often lots of instances of collaboration between the education and examinations authority and grading institutions. These cooperation experiences are a double-edged sword, which either enhance the stability of cooperation and strengthen the willingness of the education and examinations authority to continue to entrust the grading task to a given institution, or may cause differences and contradictions between the two sides, leading the education and examinations authority to use its trump card: changing the grading institution.
It can be seen that in the grading progress there is a difference between the education and examinations authority and the grading institution in terms of target setting, which is caused by the consideration of the two parties based on their respective interests. The education and examinations authority controls the right of target setting by sharing the inspection right with the grading institution, and obtains the expected grading results. The right of incentive distribution is also in the hands of the education and examinations authority. However, the target of this right is not the grading teacher, acting as the implementation agency, but the grading institution, acting as the management party. Because the education and examinations authority can assign the grading teacher rosters to the provincial high schools through the channels of the department of education in the form of administrative instructions, it does not need to consider the cooperation of the assigned schools. However, the education and examinations authority must maintain incentives in order for the grading institution to perform the grading as expected by the education and examinations authority. So, in the dimension of the distribution of control rights, although the principal party, management party, and the implementation agency do not belong to the same bureaucratic organization system and differences of target setting exist, in the principal party’s control of implicit incentives for the management party and in its sharing with the management party the right of examination, the three parties are closer to the high correlation governance model, rather than the subcontracting system or the loosely related model.
The mechanism of variation in grading criteria
Who controls the grading criteria?
Under the highly correlated governance model, the principal party has the right of target setting, inspection and acceptance, and incentive distribution. Accordingly, the answers to the grading examination paper formulated by the education and examinations authority should obviously be the only criteria which should be obeyed closely by the grading institution and the grading teachers. However, in the pre-grading mobilization meeting, both the education and examinations authority and the big group leader all emphasized that the grading criteria of the examination must be in the hands of the small group leaders rather than the grading teachers. The inspectors stressed that the answers issued by the education and examinations authority are not “standard answers” but “reference answers”, and the small group leaders in charge of judging specific questions have the actual decision-making right and discretion. It can be seen, therefore, that although the education and examinations authority has the professional and technical authority to organize the examination, it does not adopt absolute centralization in the grading process, but rather gives the grading institution autonomy to implement the policy. This can be regarded as the transfer of the principal party’s expertise and technical authority to the management party. We have stressed from the very beginning that the grading criteria for examination questions must be in the hands of the heads of the small groups, and cannot be controlled by the grading teachers. Because these teachers come from different high schools, their teaching methods and understanding of the test questions are not exactly the same. Every teacher wants to be able to influence the criteria so that they are as close to their own understanding as possible. If how they themselves teach is consistent with the grading criteria, it would undoubtedly help improve the scores of the students in their school. In this situation, it is unfair to other high school teachers and students. (Interview subject S5, inspector of the education and examinations authority of Province X, April 22, 2016) Too many mouths talk when negotiating grading criteria. Once we announce the criteria, sometimes before we finish our sentences, they start shouting loudly. They usually say, “how do we usually teach, how is this test question in the textbook” and so on. In the end, either the small group leader’s voice is drowned out in the noisy discussion or the small group leader shouts down them, forces the criteria on them, and does not allow discussion. But, even so, some grading teachers don’t listen at all, and grades according to their own understanding of the criteria whatever you say. (Interview subject S9, small group leader, November 9, 2016)
“Grading roughly” and “grading precisely”: Two behavioral preferences of grading teachers
At the pre-grading mobilization meeting, the responsible dean and the small group leaders analyze the wording and expression of the grading criteria and example answers of each test question formulated by the education and examinations authority word by word, extracting the core knowledge points corresponding to each score value, and clarifying the grading criteria. However, in the process of grading, the examinees’ various answers lead to the emergence of two kinds of behavioral preferences as criteria: “grading roughly” and “grading precisely”.
“Grading roughly” refers to those whose answers are similar to the basic points in the reference answers. This criteria requires that, according to the reference answers issued by the education and examinations authority, the keywords and similar expressions with the meaning of the answer are separated out. However, the examinees’ answers are so varied that it is impossible to cover all the possible answers to each question in the “trial” stage, which is only two hours long. Therefore, after the “formal” stage begins, it is often encountered that the content of an examinee’s answer goes beyond the scope of the grading criteria discussion, which causes discussion among the grading teachers about whether they should give scores. Although these grading criteria are more humanized and conducive to improving the average score, it also increases the ambiguity and uncertainty of the judging scale, which leads to the inconsistent understanding of the “similar” degree of the grading teachers, causing the deviation of the first and second grades to exceed the threshold range, and increasing the “third grading rate”. For some test questions, the reference answer is “oppose colonialism”. Can the student’s answer be “oppose imperialism”? If the answer is no problem, how about “opposition to hegemony”, “opposition to power doctrine”, or “opposition to colonial aggression”? The answers are varied. It’s hard to know which is right and which is not, and how rough it is. On this scale, every grading teacher has a different understanding. And that’s just one of the little points in the standard answer. For each of the other points, there are countless similar, vague statements in the students’ answer sheets. Some of them are even unexpected. If the grading teacher is left to understand according to their own speculation, it is difficult to ensure that everyone is judging the same way. (Interview subject S15, grading teacher, November 10, 2016) Some of the answers are people and places. It is easy for students to write wrong words. For example, the answer to a question is Zhu Yuanzhang [朱元璋], I have seen, there are several answer papers to write the character “zhang” incorrectly. Some are written as 章, some as 彰, all kinds of mistakes. Do you give a point or not? The students obviously knew what to write, but it just wasn’t accurate enough. Moreover, if the student writes Ming Taizu [明太祖], do you give them a point? Can you show that Ming Taizu and Zhu Yuanzhang were not the same person? If you gave a point, but the other teachers in the group did not give, then the two grades are inconsistent and third grading will be needed. (Interview subject S26, grading teacher, November 15, 2016)
However, is “grading roughly” really the root cause of lowering the average speed and raising the third grading rate, as the grading teachers claim? Actually, no. In past grading tasks, the overall average speed of the grading teachers is usually stable at between seven and nine seconds per test question. Those who exceed ten seconds are the “laggards” (luohou fenzi) who need to be “knocked”, and those who are below five seconds are the “model workers” (laodong mofan) worthy of praise. An interesting phenomenon is that many of the “model workers” in the grading progress mostly adopt the “grading roughly” criteria. One teacher author interviewed, whose average speed was under four seconds, explained why: If you glance at it intuitively, give scores regardless of whether it’s vague or marginal. In this way, you don’t need to judge and think, and you won’t be slow. If every grading teacher graded according to this method, the third grading rate would definitely go down, because for the same test question, all teachers must give high scores, so there will not be much difference. What really slowed down the reading speed and increased the third grading rate was that the teachers were out of step. Some teachers are quick to give high scores, while others are thinking about whether this sentence is similar and should be given scores in each test paper. A test question, you carelessly look at the past, no problem. If you look carefully, it’s easy to find problems. Therefore, thinking and deliberation can easily lead to a lower grading. On the same test question, you used five seconds to give a score of five in the first grading, and he used ten seconds to think and deliberate in the second grading, and finally gave a score of three. In this way, the average speed naturally went down, and the third grading rate was also high. (Interview subject S16, grading teacher, November 10, 2016)
Compared with the fuzziness and uncertainty of “grading roughly” on the operating level, the criterion of “grading precisely” is much clearer. “Grading precisely” normally means that if the examinee’s answer is not exactly the same as the reference answer, it is incorrect; even a single incorrect character is forbidden. Although this clear and straightforward grading criterion can lead to a decline in average scores in its rigid adherence to the reference answers, in the eyes of the grading teachers, it minimizes disagreement and controversy among the grading teachers, thus reducing the third grading rate and increasing the average speed. At a glance, it is clear whether the content of the answer sheet is consistent with the reference answer. You don’t have to spend time thinking about things that are close to each other, you just give scores according to reference answer, and the average speed won’t be slow. If grading teachers take the same criteria to the same test question, first grading and second grading both give the same scores, there won’t produce third grading, which is useless. (Interview subject S24, grading teacher, November 13, 2016) Teachers from key high schools often carefully look at the answer paper content, pick test questions, and assign points in strict accordance with the reference answer. This is their vocational training. They have a detailed grasp of the teaching knowledge points in the usual teaching and testing. As a result, their grading speed will naturally not increase, and the average score will not be high. Teachers from ordinary high schools are not as detailed as key school teachers in their everyday teaching and testing. When they are grading, they also tend to grasp the criteria more broadly and apply them less carefully. Any answer that is even remotely relevant will be given points. In this way, their speed won’t be slow and their average score won’t be low. (Interview subject S12, the responsible dean of the department of history of University A, April 22, 2016)
From polarizing to one-sided
At the initial stage of grading, it is difficult for the education and examinations authority, the grading institution, and grading teachers to completely agree on the understanding of the subject to be graded and the reference answers, and there is coordination and conflict among the three parties. During this period, multiple supervisors such as the small group leaders, the big group leaders, and the inspectors of the education and examinations authority often give instructions to the grading teachers. These instructions often contradict each other. The small group leaders have limited information authority, which makes it impossible to effectively grasp the progress difference between the groups. When receiving an instruction from the big group leader to speed up, the small group leader usually requires the members of the group to adopt the criterion of “grading precisely”. However, this criterion will not only improve the average speed and reduce the “third grading rate”, but also lead to the decline of the average score. At this time, the inspector from the education and examinations authority will most likely come forward and give instructions to the small group leader to raise the average score, so the head of the small group will announce the new criterion of “grading roughly” to the members of the group, thus causing the “torturous” phenomenon in which the two grading criteria are used alternately in the early stage of grading.
When the grading progress enters the middle and late stage, the education and examinations authority, the grading institution, and grading teachers have gone through the conflict period, and their understanding of the test questions and the reference answers is gradually converging. During this period, the behavioral preferences of grading teachers will have an impact on the target setting of both the grading institution and the education and examinations authority. The latter adjusts the grading teachers through positive and negative feedback according to the fluctuation of these quantitative indicators. The “average score” orientation of the education and examinations authority and the “average speed” orientation of the grading institution make the grading criteria shift from the polarization of the initial stage of grading to the one side of “grading roughly”, resulting in the polarization of “laggards” and “model workers” among the grading teachers. Many grading teachers in key high schools are labeled as “laggards” and are disciplined and urged by the grading institution, thus putting them in an inferior position. However, the grading teachers coming from ordinary high schools are often deemed “model workers” and praised and rewarded by the grading institution, so they are in a dominant position. There are two reasons for this.
First, in the grading group’s discussion of scoring criteria, ordinary high school teachers have a stronger voice than key high school teachers by virtue of their numbers advantage. As the implementers of the curriculum on which the exam is based, the first-line grading teachers are mainly from key high schools and ordinary high schools, but the numbers of grading teachers selected from them is not evenly distributed. The Education and Examinations Authority of Province X (2016) has a clear stipulation on the criteria for the selection and transfer of high school teachers: “In-service high school teachers with three years of teaching experience or above, who have the title of first-grade high school teacher or above”. This threshold not only keeps out lower-quality high school teachers, but also effectively filters out teachers who are teaching in key or ordinary high schools but whose professional ability and professional title are not up to standard. Therefore, high school teachers who participate in grading are all above average regardless of their personal professional level or the units they work for, and the probability of “bad schools” and “bad teachers” participating in the grading is reduced to the lowest level. At the same time, the education and examinations authority hopes to mobilize the enthusiasm of its subordinate high schools as much as possible, and “the selection of grading teachers should take into account balance among high schools” (Education and Examinations Authority of Province X, 2016). The base group of teachers in ordinary high schools is much larger than that in key high schools. Therefore, the number of teachers in ordinary high schools participating in grading is also significantly greater than that in key high schools. It is not difficult to imagine that, as such, the voice and influence of ordinary high school teachers, with the number advantage, will obviously be greater than that of key high school teachers.
The second and most important reason is that the “grading roughly” criterion adopted by the grading teachers from ordinary high schools caters to the target setting of the education and examinations authority and the grading institution, while the “grading precisely” criterion adopted by the grading teachers from key high schools fails to meet the target setting of the education and examinations authority and the grading institution. This was a “common practice” phenomenon in the previous grading processes that the author has participated in. Faced with contradictory targets—the “average speed” of the grading institution and the “average score” of the education and examinations authority—the grading teachers from ordinary high schools are not only relied upon by the grading institution because of the “more, faster, better, and less” approach to the grading task, but also get the acquiescence and encouragement of the education and examinations authority because of the higher average scores they give. In the grading process, they are constantly praised orally by supervisors, and stand out from the group and were in a dominant position in the overall discourse. Their grading criteria and grading methods will be used by the small group leaders as a model for improving the team’s work efficiency and will be publicized throughout the whole group. In order to reduce the “third grading rate” and improve the speed of their own grading, other members of the group are also willing to take the initiative to get closer to the “model workers” criterion of “grading roughly”. Even if the “model workers” occasionally joke with the rest of the group and do something unrelated to grading, management is more likely to take a blind eye. In contrast, grading teachers coming from key high schools often get verbal warnings and even criticism from the big group leaders and the education and examinations authority because of their adherence to the criterion of “grading precisely”, which fails to fit the target setting of the education and examinations authority and the grading institution on the two indicators of average speed and average score. As the implementation agency, they become “laggards” in the eyes of the principal party and the management party, and as such they are at a disadvantage in the overall discourse on the grading process. Their preferred grading criterion is not copied by other grading teachers. If they take a break from grading, or do anything other than grading, they will receive an immediate verbal warning from management. Most of the “outstanding” assessment meted out at the end of the grading process is not given to them.
Conclusion
By applying control rights theory to analyze the grading process of the history examination of the new gaokao in Province X, we can find that in the governance model of the unstable public organization process—in which there is participation by public institutions belonging to different bureaucratic organization systems—the situation is not akin to the subcontracting system or the loose association system, but closer to the highly interrelated model. The principal party, the management party, and the implementation agency can break away from their own routine work rhythm; close relations and high degrees of cooperation are helpful to the principal party in implementing the top–down policy process. Its professional and technical authority gives the principal party the organizational legitimacy to control the management party and the implementation agency, which helps the principal party to effectively exercise the right of target setting, right of inspection and acceptance, and right of incentive distribution.
Under the highly related mode of stable interaction between government and social organizations, the principal and management parties belong to the same bureaucratic organization system, which is conducive to the former restraining the latter through administrative levels and making the latter implement the targets set by the former. However, in the process of unstable public organization, the principal party, the management party, and the implementation agency neither belong to the same bureaucratic organization system nor common administrative oversight and market incentive mechanisms, which makes the three parties likely to take action strategies based on their own interests, and increases the uncertainty in negotiation and maneuvering. In this case, it is manifested in two aspects: First, the target setting of the education and examinations authority and the grading institution is different. The division of the right of inspection and acceptance and the “right of arbitrary intervention” in the grading process by the education and examinations authority is an action strategy to control the grading process as a whole in response to the difference of the target setting with the grading institution.
Second, and more importantly, it is precisely because of the negotiation space and maneuvering ability of the three parties that the implementation agency, which has been ignored in the former literatures, can have adverse effects on the principal and management parties from the perspective of their own behavioral preferences. In this case, this takes the form of a unilateral shift away from a polarity of two grading criteria and toward the single criterion of “grading roughly”. This fact is reflected in the grading institution being endowed with the right to interpret and consider the reference answers by the education and examinations authority, that is, the management party having the autonomy to implement the policies by transferring its own professional and technical authority, in an attempt to mobilize the enthusiasm of the grading institution. However, the professional cap between the grading institution and the grading teachers has weakened the effect of the exercise of this autonomy implementation, leading to a controversial “fuzzy zone” in the understanding of the reference answer, rendering the subject dean the key link in the grading process difficult to fully exercise autonomy policy implementation, the education and examinations authority is inconvenienced to make unauthorized decisions. This is the weak link in the chain of policy implementation. In this case, as the implementation agency, the two behavioral preferences of the grading teachers themselves get positive and negative feedback from the principal and management, forming a behavioral adjustment mechanism of strengthening and weakening, which makes the three parties’ understanding of policy implementation gradually converge. The “model workers”, who can meet the targets set by both the education and examinations authority and the grading institution, become the objects that the principal party and management party rely on. The grading criteria change in the course of this continuous interaction.
Thus it can be seen that in the governance problems of contemporary China, the process of unstable public organization completed by various social forces such as public institutions and social groups that do not belong to the same bureaucratic organization system urgently needs to be included in the research field. In this process, not only the analytical dimension of control rights theory is worth further splitting, but also the maneuvering and action strategies of various social forces around control rights in the process of organization and implementation should be deeply explored.
Footnotes
Acknowledgements
I would like to thank Professors Chengren Lu, Dalei Miao, Ding Li, Xiulin Sun, Yingying Ji, Jun Qu, Yufeng Li, and anonymous reviewers for their extremely helpful comments and suggestions.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was funded by the College Young Teacher Study Abroad Project (2018) of the China Scholarship Council.