The Move to Quality Assurance in Chinese Higher Education: Tensions between Policy and Practices.

This paper investigates quality assurance in Chinese higher education by looking at the Evaluation of the University Baccalaureate Programmes Project (EUBPP) conducted in China. It analyses factors that bring about tensions between policy and practice in this area through a study of two contrasting tertiary institutions. The results show that the aims of the EUBPP have been partly achieved through the interaction of policy-making and implementation. In the process, contextual factors connected to institutional status and stages of development have led to varying degrees of tension between policy and practices.


Introduction
In recent decades, policy and practices in higher education around the world have been strongly influenced by neoliberal ideology. Neoliberalism envisages opening up world markets and developing the role of the private sector in political and economic affairs, including sectors such as education (Adamson et al., 2012). Varghese (2012) suggests that there are a number of drivers behind these changes in higher education: the desire to enhance national economic competitiveness; the need to expand higher education to allow access for a greater proportion of the population; the impacts of geopolitical changes and the forces of globalization on student mobility; the emergence of higher education as a marketable commodity; and the imperative for higher education institutions to find funding from sources other than the state. The notion of higher education as a business has resulted in the importation of various practices related to efficiency, accountability, and quality assurance.
The world trend shows that tertiary institutions can no longer take their values and privileges for granted (Massy, 2003). Under the significant pressure of the 'convergent effect of financial restrictions, rising expectations and social demand, mandates of new economy, and weakening symbolic capital' (Amaral et al., 2003: 131), higher education is experiencing 'the erosion of trust' in terms of the quality of education being provided (Massy, 2003: 3). To enhance the quality of tertiary institutions -and to respond to the pressures of global competition, the requirements of economic growth, and the calls for inspection from diverse stakeholders -many countries have introduced various types of quality-assurance mechanisms (Lewis, 2009;Vidovich, 2004;Westerheijden, 2003). Quality assurance has been emphasized not only in industrialized countries but also in less developed nations, such as India, Thailand, and Indonesia (Bray, 1996;UNESCO, 2006;Yavaprabhas, 2008).
In many countries, quality-assurance systems have a systematic and far-reaching approach to ensure that universities have efficient mechanisms for assessment and improvement (Harman, 1998b;Harvey and Williams, 2010). One important approach is the external evaluation of institutional processes and performance, in which the control of central government over higher education is strengthened. Several studies have explored the critical issue of tension between managerialism and the academic profession (e.g. Trow, 1994;Massy, 2003). In addition, Perellon (2007) indicated that understanding policy involves two dimensions: the ideational and the material. The ideational dimension refers to the set of cognitive values and norms that underpin the production of policies. The material dimension is the implementation of the policy that includes the methods and tools used, the procedures and approaches adopted, and other concrete actions.
Since the end of the last century, China has promulgated a series of policies to stimulate the development of higher education. In this process, provincial government and universities gained much more freedom to make their own decisions in order to cater more efficiently to the needs of a highly competitive market. China has experienced a dramatic expansion of higher education since 1999 -enrolment increased from 9.8 per cent in 1998 to 26.5 per cent in 2010 (Ministry of Education, 2011) -and faces the same challenges as other countries in maintaining or improving the quality of higher education with limited resources. Additionally, systematic criteria to be followed by administrations within the higher education system had, until recently, gone largely unenforced (Li et al., 2008: 13). In order to strengthen governmental control over quality assurance, the Chinese Ministry of Education (MOE) initiated the Evaluation of University Baccalaureate Programmes Project (EUBPP) 1 in 2002. Prior to the unveiling of the EUBPP, some evaluation-related work had already been carried out. For example, in 1985, the government conducted an evaluation of guidelines within the schools of engineering as part of a trial run. This was followed by the 'Draft Regulation of Higher Education Institution Evaluation' (MOE, 1990). In 1994, China started three evaluation programmes: the University Evaluation Standards Project, Exemplary Evaluation, and Random Evaluation (HEEC, 2004 1. To enhance the level of macro-administration and direction on higher education given by the national government. 2. To encourage education departments at all levels (for example, provincial level and municipal level) to support tertiary institutions in various aspects of their teaching. 3. To provide clarity regarding the way forward for institutions so that they may: improve managerial conditions by increasing fiscal investments; strengthen educational infrastructure and deepen education reforms; and stimulate development by clarifying development plans, improving teaching conditions, strengthening the quality of teaching staff, developing management skills, and gradually establishing internal quality assurance mechanisms.
Central government designed criteria to evaluate each institution based on uniform standards. The criteria include eight basic aspects: institution mission, quality of staff, facility and utilization, subject construction and curriculum reform, teaching administration, style of study, teaching effect, and special items. These 8 aspects are further extended to 19 sub-standards and 44 observation points, covering nearly every facet of teaching and instruction. Some of these aspects are easier to measure with accuracy than others. Taking the observation point 'the numbers and construction of frontline teachers' as an example, the university would be graded A if teachers holding postgraduate degrees occupied over 50 per cent of all frontline teaching posts, and C if the number were between 30 and 40 per cent. Quantification and grading here is straightforward. However some observation points, such as 'institution mission' and 'style of study', are not so easily quantifiable and the grading mainly depends on the judgment of government-appointed on-site experts.
The EUBPP encompasses three stages: the first, 'self-review', stresses the institutions' need to scrutinize themselves according to the evaluation criteria and to submit self-evaluation reports to the on-site experts. Following these submissions, on-site experts inspect the selfreports according to the standard criteria. The second stage is an on-site visit. About a dozen experts go to the university and listen to the university leaders' presentation of their self-review, and then inspect various aspects of instruction, utilizing a range of different approaches. This process lasts for four or five days, after which the inspectors will give a brief report to the institution, including initial evaluation results and some suggestions on future development. After that, the 'central expert committee' -the on-site experts selected by government to make the final decision on evaluation grade -considers some related impact factors, such as university history, geographical factors, and the type of tertiary institution, and publishes a final evaluation grade. The third stage is aimed at improvement. The university makes improvement plans based on the evaluation report, and then moves to implement those plans in a logical and efficient fashion. An implementation report of improvement must then be submitted about one year after the evaluation.
Although some researchers have studied the effects of the first round of EUBPP (e.g. Li et al., 2008;Zou et al., 2012), the multiple perspectives involved in this evaluation -especially the perspectives of institutional administrators and frontline teachers, as well as their complicated relationship during this evaluation -have not yet been fully explored. Spillane and his colleagues (2002) emphasize that understanding the personal cognitive dimension can contribute to our understanding of the evolving process of educational policy and its concomitant challenges and issues. They also argue that the key aspect of how agents make sense of reforms has seldom been explored in western research. Honig (2009: 333) further states that the outcomes of a policy 'depend on interactions between that policy, people who matter to its implementation, and conditions in the place in which people operate'. To fill this gap, this paper focuses on the implementation and adaptation of the external evaluation mechanism in higher institutions, as well as the multiple perspectives and complex reactions of diverse stakeholders in China.

The study
The description above shows that the EUBPP is a large-scale, comprehensive evaluation for tertiary institutions in mainland China. To explore the authentic opinions and feelings of the participants thoroughly, this paper adopts complementary qualitative methods such as document analysis and semi-structured interviews. The data were collected from 2009 to 2010 in a developed coastal province in eastern China. Partly because of its location and traditional culture, the province ranks highly in national statistics for both economic development and education (Wu and Li, 2002). The province has a number of prestigious, top-tier universities with a long history, in addition to a range of other types of tertiary institutions, such as technical colleges. Finding one or two examples that can typically represent each type is unrealistic. Therefore, two strongly contrasting universities in terms of student academic achievement (higher and lower), history (longer and shorter), and features (comprehensive university -one offering social sciences and natural sciences -and professional institution) were selected within the province.
University A, as one of China's key comprehensive universities, has a long history. During its development, University A has received financial support from both the government and the general public. University A is renowned for providing good facilities for both teaching and research work. Currently, the student body in University A numbers over 43,000. As a whole, University A is representative of top-level tertiary institutions in China.
University B is a key provincial university co-established by the Chinese Ministry of Industry and Information Technology (MIIT) and the Provincial People's Government. The first undergraduates were enrolled in the early 1990s. University B developed rapidly with engineering, information technology, and telecommunications being the main subjects offered. Statistics in October 2010 showed that University B had fewer than 20,000 students. Generally, in terms of quality, University B belongs to the middle-to-bottom level of tertiary institutions in China.
Formal research interviews were conducted in these two universities with staff members that included school-level administrators, department-level administrators, and frontline teachers. The interviewees were employed in various departments and had a wide range of academic responsibilities. Information about each interviewee is summarized in Table 1. Given that the faculty is the basic unit for evaluation in EUBPP, most informants in this study were administrators at the university and faculty levels. Each interview was conducted in a private space within the respective university, lasted at least one hour, and focused on eliciting interviewees' opinions and feelings about the external evaluation. Constrained by restricted research time, the size and scale of data collection were limited. However, the selection of various informants with disparate backgrounds facilitated a multi-layered understanding of decision-making and quality-assurance adaptation. In addition, frontline teachers and staff were also selected and interviewed for data triangulation. The major interview questions were as follows: • What did you/your department/your university do in the EUBPP?
• What were your feelings about the EUBPP? • How do you evaluate the impacts of the EUBPP? Why? The interview data were transcribed and analysed inductively. Following the study by Perellon (2007), the categories and issues were identified during data analysis. In 2007, Perellon posed five fundamental questions, as follows: 1. Objectives: What should be the aims and objectives of the quality-assurance policy? 2. Control: Who should control the process of quality assurance? 3. Areas: What are the domains covered by the quality-assurance procedures? 4. Procedures: How are the quality-assurance procedures established? 5. Uses: How will the collected information be used?
Although these five questions were designed to compare quality-assurance policies, they also assist our understanding of EUBPP implementation. Based on Perellon's theory, we divided the EUBPP process into five aspects. In the first round of coding, following the stages of familiarization, the interview data were initially divided into categories on the basis of these five key questions. In the second round of coding, repetitive key phrases and issues within each aspect were underlined. The data were further analysed by highlighting the similarities and differences between University A and University B. By triangulating related data, such as self-reports or interview data, pertinent quotations were selected, translated into English, and double-checked by researchers.

Objectives of the EUBPP
The interviewees from the two universities basically agreed that the government's purpose in initiating the EUBPP was formative and the evaluation was mainly designed for further improvement and development: There are too many problems that need to be solved in higher education. That is why MOE introduced this programme. They want to stimulate the improvement mechanisms and encourage tertiary institutions to face their problems as soon as possible. (A-1) To tell the truth, we didn't know how to manage an institution before … I think that is why central government conducted the EUBPP. To some extent, they want to give us criteria to facilitate our management. (B-1) With regard to the acceptance of the EUBPP, the reactions of the interviewees differed from one university to the other. The Director of the Teaching Affairs Office and frontline teachers at University A told us that, to some extent, they felt that they were forced to conduct the EUBPP: To some extent, this round of evaluation requires you to do it even though you don't want to. We all share the pressure stemming from being obliged to be evaluated externally and get an outcome of excellent performance. (A-1) Frankly speaking, our university doesn't need to have that evaluation. We are at the top level in China! It is ridiculous that the institutions like ours need to undergo the evaluation. (A-3) Conversely, University B's reaction was more enthusiastic. As a developing institution, University B wanted to use the EUBPP as a means to strengthen its management system and expand its campus: Our institution has developed very fast recently so that our management systems couldn't cope with the demands of the increasing development. The EUBPP gives us an opportunity to reflect, design, and implement a more systematic and efficient quality-assurance mechanism. (B-1) Without the EUBPP, we couldn't expand our campus so efficiently. Because of the criteria of the EUBPP, we had to speed up the construction of our new campus to a year for the inspection. It was amazing! From this point of view, our previous goals for campus extension were successfully completed under the pressure of the EUBPP. (B-2) University A felt that the criteria and requirements of the EUBPP did not match their developed status. The external evaluation brought nothing positive to the institution; on the contrary, it had a negative impact on their plans for innovation. Therefore, the motivation of University A to participate in the evaluation was mainly to finish the task. In addition, as one of the most prestigious universities in China, it would be unacceptable if the university failed to achieve a ranking of anything but 'excellent'. Thus University A's aim in undergoing the EUBPP was summative. In comparison, the objectives of University B integrated formative and summative aims, as improvements in management systems and campus facilities could be embedded in preparations for the EUBPP. This process fitted the government's (stated) formative purposes very well. In addition, University B saw a summative dimension in that they had been classified as 'unacceptable' in the previous evaluation, so undergoing the EUBPP gave them an opportunity to demonstrate their improved quality and development.
Although the teaching function of universities was highlighted in government documents, the promotion of teaching innovations and enhancement of teaching quality proved comparatively difficult to measure. Hence, this aspect was not emphasized by middle managers or frontline teachers in their reflections on the EUBPP. Perellon (2007) stated that two main bodies could be identified from previous evaluation experiences: political authorities on the one hand, and higher education institutions on the other. Although these two might in theory be seen as polar opposites, to a substantial extent they have to interact and work together in the implementation of any quality assurance scheme.

Central government
On paper, the central government designed the EUBPP as a formative measure to facilitate innovation and improvement in higher education, and the evaluation is not the ultimate goal in itself. However, poor evaluation results can bring negative consequences for the future development of an institution. Normally, institutions that undergo the EUBPP review are classified under one of four categories: excellent, good, acceptable, and unacceptable. The MOE (2007) stipulated that the intake of new students should be limited or suspended for those institutions that received a rating of 'unacceptable'. In addition, these institutions could not establish new degree programmes, could not apply for the certification to authorize Masters or Doctoral degrees, and could not apply for permits to construct vital laboratory facilities. For these reasons, the MOE can, to some extent, be said also to have adopted a summative approach with the introduction of the EUBPP, thereby placing it in a dominant role in the relationship with the institutions under review.

Input of experts
The on-site experts are scholars in various fields, including university presidents as well as academic experts. They are invited and assigned to be on-site experts by central government and have an intermediate role between the political and the academic agencies. However, they were not always regarded by university staff as fit for the role. One administrator from University A was sceptical about their capabilities -a view shared by other administrators in the university. I personally think that you cannot say these experts have a problem about their level of quality. However, they have no professional training and they are not suitable for undertaking such a job. The Ministry of Education is aware of this issue and so only the team leaders receive training. Still, whether the training really works is another question because, when they [the team leaders] go to Beijing, they only listen to some speeches from the Ministers and the secretary experts, and read the documents on the evaluation system. (A-1) Such scepticism makes it difficult for the experts' recommendations to be considered deeply or implemented after the inspection. From this point of view, the control powers of these experts were more formal than substantial in the evaluation of University A. On the other hand, the suggestions of external experts were welcomed and accepted in University B. A director from University B said: 'These experts understand what we really need. They give us many pertinent suggestions which have been integrated into our development plan for the next ten years'. It can be seen that the extent of acceptance of experts' opinions depends on the perception of the external inspectors' understanding of the institution and their related knowledge and capabilities.

Tertiary institutions
In both universities, the people in positions of responsibility took the EUBPP seriously because of the implications of the evaluation results for future development opportunities. In both institutions, a special office for preparing for the EUBPP was established, chaired by the university's president or vice-president and filled with administrators charged with organizing the preparations. At the department level, corresponding evaluation offices were also established to work in collaboration with the university-level office.
University A treated the audit as a one-off event. The new Evaluation Office was staffed by personnel from the Teaching Affairs Office and the Quality Control Office. After the EUBPP, the Evaluation Office was disbanded, as its function was solely to cater to the requirements of the audit. The situation was different in University B. Before the EUBPP, there was no specialized office in charge of quality control. In order to prepare for the evaluation, some middle-level managers and staff were drafted to form an evaluation office chaired by the university president. They adapted the university's development orientation, enhanced the teaching facilities, constructed an evaluation system, and improved quality assurance with regard to the requirements of the EUBPP. After the inspection, the evaluation office remained in place to deal with issues of quality control. An interviewee stated that: We want to grasp this opportunity. There are some problems in our management. If we want to get better and better, we need to consider how to resolve these problems. Preparing for the EUBPP was a priority for all of us. This was not only because of the external pressure, but also because of the internal requirements. Now, we find that our university really has changed a lot after the EUBPP. The evaluation engineers our development in a more stable and faster way. (B-1) From this point of view, University B embraced the developmental opportunities afforded by the EUBPP, while University A eschewed them.

Inspection areas of the EUBPP
Although the EUBPP was intended to be quite comprehensive in scope, some aspects such as class visits, staff-to-student ratio, proposals for student training, and student assessment were specifically highlighted as problematic by on-site experts and the universities' local administrators. Two indicators elicited an especially high number of concerns. First, experts stressed the difficulties created by standardization of documents (including examination papers and graduate theses). Many managers and teachers worked very hard to meet the demands of this standardization. As a result, the inspection of such documents was frequently criticized by local administrators as well as frontline teachers.
The winter vacation in 2008 was used to make up the documents. We went to every department and checked the examination scripts. We discovered the errors and corrected teachers' mistakes in marking and asked both the graduate students and department officers to look at the scripts again. After re-marking the scripts, we asked the teachers concerned to change the marks and sign off on them accordingly. (A-2) For marking, this question carries ten marks. A student gets 8 marks and has 2 marks deducted. On the left, it is required to write +8 marks with a circle and -2 marks with a circle on the right … If there is a change in the marks, you have to sign on the paper and change the overall marks and sign again. This is too formalized. (A-1) Although I know that the aim of the evaluation is to help us construct an effective managerial system, compiling documents and students' test papers is difficult and time-consuming. We have spent a long time preparing. We bought all these bookcases and folders to sort out the paper work. (B-3) Second, due to their easy-to-quantify nature, the construction of an aesthetically pleasing environment and the purchase of research facilities tended to become the focal point of review.
I dare to say that if the MOE had not conducted this evaluation, many universities would not have been able to build new school campuses. The MOE's evaluation has its own requirements and standards and so, for universities to comply with these standards, they had to borrow money to build student hostels. (B-2)

Evaluation procedures of the EUBPP
From the empirical data, it appears that some institutions added extra stages to the threestage model envisaged by the MOE's Higher Education Evaluation Centre (HEEC). First, most institutions conducted a pre-evaluation to make sure that they could achieve a satisfactory grade in the formal evaluation. The pre-evaluation, which was conducted by the provincial education bureau or by the university itself, involved a number of external experts who had been involved in the inspection of other institutions. To ensure that the pre-evaluation stage was as authentic as possible, the duration and procedure of the trial inspection were exactly the same as the formal inspection.
Preparing for the pre-evaluation is more nerve-wracking than the actual evaluation. It is because you employ experts to find out your own weaknesses. These experts are involved in teaching and they easily detect and focus on the problem areas. In fact after the pre-evaluation, the actual evaluation is more relaxed. (B-4) In practice, due to the limited time available, evaluation experts and local administrators paid more attention to objective products (such as hardware, facilities, paper documents, and so on) than the more time-consuming and difficult-to-quantify formative processes.
They (evaluation experts) put too much attention on the details, especially on paper document inspection. Let me give you an example: it was a statutory requirement for teachers to write the students' scores in a specific location when marking exam papers. It was unacceptable if teachers wrote the score in the wrong location. Furthermore, teachers needed to draw a circle around the score. You know, teachers always have their own style when they grade students' papers. We had to amend them all one by one. Although the workload was huge for all of us, we had to do it painstakingly. (A-2) The on-site visit lasted no more than six days, making it impossible -in the view of university administrators -for the evaluation experts to gain sufficient information about the instruction and management in a university.

Uses of the EUBPP data
It was noted above that, although EUBPP was designed to be formative, it also serves a summative function. In University A, the formative function was questioned by administrators.
Sometimes I think: is the evaluation a scientific one? What will such an evaluation bring to the university? Well, the panel will comment on issues which the university will follow up … If they take five days to give advice on the university's plans for the next ten or thirty years, is it possible for them to do a good job? ... The panel may feel that we should do this and we should not do that … The university has to make revisions in accordance with their advice. Can they really give a comprehensive review? It is problematic. (A-1) University B proactively integrated the external evaluation with their own desire for improvement and innovation. The suggestions advanced by on-site experts were taken seriously. From this point of view, the use of inspection results in University B was consistent with the policy intentions. Furthermore, the impact of the evaluation outcomes went beyond governmental expectation as they were closely associated with personal career development.
To tell the truth, the EUBPP also helps us select middle-level managers. Have you talked with teacher C? He was a teacher in the Department of Finance and Banking. Because he did a good job during the preparations for the EUBPP, he was promoted to Vice-Director of the university's Teaching Affairs Office. Personal capabilities were put under the microscope during this process. After the EUBPP, many personnel changes took place. (B-4) The implementation of EUBPP thus brought about some unintended outcomes in University B. Harvey (2004) states that accreditation in higher education leads to a fundamental shift of power from educators to managers and bureaucrats. In a reflection on various countries and districts, Harvey and Williams (2010) state that tension clearly exists between quality assurance as a bureaucratic and administrative challenge and the improvement of academic quality. As discussed, in this round of evaluation, the Chinese government intended to promote both organizational quality and educational quality. To achieve the first goal, the evaluation was designed to improve the administration of tertiary institutions at the national level as well as the administrative mechanism at the school level. To achieve the second goal, the evaluation was conducted to continue enhancing academic endeavours. Even though these two goals are intertwined, the tensions between them still exist (Zou et al., 2012). For universities such as University A, which already had a mature mechanism for ensuring educational quality, the external evaluation was perceived as an administrative burden rather than an impetus for educational improvement. Moreover, several flaws in this round of evaluation, such as the overemphasis on standardized papers and on the provision of a pleasing environment, may have disturbed the original process of improving academic quality in University A. There had been a number of large-scale discussions on educational philosophy and comprehensive development at University A since the 1990s. The developmental orientation and university culture were gradually transformed and strengthened during this process. In 2006, University A established conceptual foundations for instructional reform in the succeeding stage. However, the implementation of the reform was deferred because of the ongoing preparation for the EUBPP. After the external inspection, the previous reform plan was found to emphasize fundamental changes of curriculum orientation and curriculum plan reframing, while the transformation from teacher-centred to student-centred pedagogy had largely been carried out. The reform subsequently became widely recognized, gained acceptance, and obtained fruitful outcomes at University A.

Discussion
For University B, which is a newly built university with an incomplete administrative system, the external evaluation undoubtedly served as a stimulus that strengthened organizational quality. After the EUBPP, the evaluation office, as an administrative unit, was retained and assigned full responsibility for assessing curriculum and instruction at University B. Furthermore, in 2011, University B extended its evaluation mechanism to two levels, namely, institutional and faculty dimensions, with a view to precisely and effectively controlling instructional quality. However, on the basis of the research data, it is difficult to conclude that the external evaluation substantially improved the quality of curriculum and instruction in University B. Thus, the self-proclaimed goal of the EUBPP to improve both organizational quality and educational quality is questionable.
A worldwide trend towards decentralization in governance has been identified (Rondinelli et al., 1983), and the Chinese government attempted to integrate this approach into this round of evaluation. The HEEC claimed that informing institutions about their own developmental orientations and improving the quality of management and education would be a key feature of the process. However, as Levin (1997: 260) noted, 'decentralization practices appear very little different than the supposedly more centralized approaches they have supplanted', and many western studies have echoed this statement (e.g., Harman 1998a;Hoecht 2006;Mollis and Marginson 2002). Similar results have also been observed in this study. The central government actually strengthened the control of tertiary institutions by predetermining the controlling agency, inspection focus, and evaluation methods. Institutions were expected to respond to the broad guidelines of the EUBPP. In addition, compared with other countries and districts, the Chinese government adopted a relatively 'hard' approach with regard to the implementation of the EUBPP. One comparative example is informative. Most western countries do not grade institutions after a critical evaluation (Lewis 2009). In China, the central government not only gives final grades to universities, but also makes the grades known to the public. As in other countries, this external audit approach is influenced by the ideology of accountability and neoliberalism, but it may also be closely connected to the history of higher education in China. From the time of the Sui Dynasty (581-618 CE) right up to the Revolution of 1911, higher education was headed by the Guozijian: the top institution in higher education as well as the agency that managed the education system throughout the empire. The Guozijian was a typical representative of the mechanism of 'academic centralization or nationalization' (Zha, 2011: 461) and it served to strengthen the linkage between state management and higher education. In modern times, the acquisition and application of knowledge in higher education has become a key part of national efforts to achieve China's goals for economic development. Consequently, the government views higher education as a critical driver of the country's development and pays close attention to its performance (MOE, 2011. Thus, 'state control seems to have always had a place, and nowadays it is still evident under the guise of something that we might identify as academic centralization or nationalization' (Zha, 2011: 462).
Local agencies have also had a considerable influence on the design of the EUBPP. This might be explained with reference to Confucianism in China. In the west, academic freedom and the search for impartial knowledge is part of the intellectual tradition of higher education. However, Chinese scholars, influenced by Confucianism, have, historically, attempted to realize their ideals through action, and have had direct responsibility for managing the state (Zha, 2011). In China, 'knowledge was less a matter of understanding the world than of changing it and scholars were expected to cultivate the self, regulate the family, govern the state, and make the whole kingdom tranquil and happy' (Zha, 2011: 459). Consequently, in terms of the EUBPP, the opinions of local agencies or scholars received respectful attention and some were adopted by central government. For example, in the early days of the EUBPP, the fact that different types of tertiary institutes were evaluated with the same fixed criteria came in for criticism, and the government responded by adapting the evaluation criteria to make the EUBPP more flexible (MOE, 2006). Marginson and Rhoades (2002) point out that policy analysis should attend to policy implementation at various levels, and there is a need to take the local contexts and practices into account. Marsh and Huberman (1984) distinguish between 'superordinate' (central decisionmaking) and 'subordinate' (local implementing) agencies in policy implementation. Lai Auyeung (1994) adds a further layer, 'intermediate' agencies, responsible for providing expert advice and also for channelling ideas from the subordinate agencies to the superordinate agencies, to inform the decision-making process. The HEEC and Secretariat for Expert Commission of the EUBPP could be regarded as superordinate agencies as they were responsible for organization and possessed decision-making powers. In contrast, the presidents of tertiary institutions could be seen as subordinate agencies as they were expected to implement the governmental policy decisions of the EUBPP. Meanwhile, the evaluation experts could be viewed as intermediate agents in that they influenced central government's final evaluation. Their role was powerful in that their focus on such features as standardization of documentation and an aesthetically pleasing environment meant that these features were over-emphasized by tertiary institutions at the expense of the whole set of criteria. However, the relationship between the agencies did not always involve the top-down transmission of a central decision. From the study, it can be seen that some presidents, such as those at University B, took this opportunity to improve the management system and engineer further development, while those at University A gave a perfunctory response to this policy. From this perspective, the scope, depth, and duration of the EUBPP's impact were profoundly influenced by the respective presidents' conceptions of the EUBPP and their coping strategies.
In terms of contextual factors, the academic level and the stage of development of the particular institution are important in explaining the differences in response to the implementation of the EUBPP. If the university were at the top level and had developed its own mature system for management and quality control, the EUBPP offered little incentive to effect change or innovation. In fact, the EUBPP could be viewed as counter-productive, as preparing for the inspection might hamper the original pace of innovation. This was the case with University A, which had planned to introduce some changes to the student learning process, but the plan had to be postponed because of the EUBPP. However, if a particular university had previously been less academically successful and failed to implement necessary changes, as was the case with University B, the EUBPP, with its comprehensive criteria and strict grading system, could compel the institution to make improvements.

Conclusions
In China, tertiary institutions can be classified into categories such as comprehensive, teacher education, and specialist (such as agricultural) tertiary institutions. These institutions can also be categorized into four types: research-oriented, profession-oriented, application-oriented, and technique-oriented. Given the patently different features of these institutions, their kaleidoscopic responses to the EUBPP may be considered logical and reasonable. For example, while all the institutions have to be evaluated according to the same indicators regulated by the central government, a multitude of universities, especially non-comprehensive ones, complained about the inappropriateness of being evaluated under the same criteria (Lee et al., 2012). Meanwhile, this study has noted that several common issues faced by various institutions -the proficiency of on-site experts, the central control of the EUBPP, the preferences of inspection -have evolved during the process, leading to tensions between policy-making and implementation. The outcomes of this study show that there is a considerable gap between the policy intention of the EUBPP on the one hand and the actual objectives -controlling agencies, inspecting areas, evaluation procedures, uses of the evaluation data -on the other. Within the Chinese higher education system, it is clear that some key agencies at all three levels (superordinate, intermediate, and subordinate), such as the responsible organizations, evaluation experts, and institutional presidents, combined with contextual factors, considerably influence the implementation of the EUBPP.
In 2010, the Chinese government issued the Development Plan Outline for Medium-andLong-Term Education Innovation andDevelopment (2010-2020) in which it announced that instructional quality and teaching reform in higher education will continue to be implemented (MOE, 2010). The MOE initiated the second round of evaluations in December 2013 and a large-scale audit of higher education in 2014. Important lessons for policy-makers can be learnt from the first round, such as discovering how one can maintain the delicate balance between external governance and internal autonomy regarding quality control, or finding the relationship between other countries' successful experiences and their own educational and cultural tradition.

Future research
As with any research, this study has limitations. The first is the choice of province, which is at the top level of education in China. Considering the tremendous variation among Chinese higher education, scholars may aim to explore more cases in different locations for an increasingly holistic picture. The second limitation is the short research time, which restricted the size and scale of the sample selection. Future research can endeavour to examine the effects of the EUBPP from two approaches. The first approach would be a longitudinal study to obtain a more in-depth understanding, while the second approach could involve a broader selection of tertiary institutions that may depict a more comprehensive picture of the relationship between the EUBPP and the development of higher education in China. Such endeavours would be worthwhile, as the introduction of quality-assurance mechanisms within higher education systems remains a challenge.