Case Study I: Standards For Teachers

The Problem of Standards and Quality

Central to the discussions of teacher policies and educational reforms is the issue of standards for teachers. By the time of the 1978 Third Plenum a teacher shortage greater than that experienced before the Cultural Revolution existed quantitatively and in terms of credentials and competence. As figure 7.1 suggests, the rapid expansion of precollegiate education since the 1950s had not seen a commensurate growth in teacher education or trained teachers. The national shortage in 1977 was estimated at 3.45 million elementary and secondary school teachers (Cui 1979, 31). According to leaders, quality was a bigger problem than quan-

Fig. 7.1.
Elementary and Secondary Expansion, 1951–78

NOTE : Figures for teachers represent the total number of teachers (qualified and unqualified) working in elementary, general secondary, and specialized secondary schools. Where data for elementary teachers were missing (1966–71), no estimates of the total teaching force are used. The numbers of teachers in training refer to students enrolled in either teacher-training schools or colleges. These numbers, as well as those for teachers, are plotted in millions. The numbers of students are represented on this graph in tens of millions.

SOURCE : Department of Planning 1984.

tity. The Chinese indicate that quality—as measured in terms of qualifications (teachers' academic credentials)—fell over the years, as is seen in table 7.1.^[7]

If China was to reach its broader educational goals, there had to be a strengthening of standards. Yet given the vigor of the attacks on teacher education and professionalism during the years of the Cultural Revolution, the urgent need to coordinate and regulate the establishment of professional standards for teaching posed a major policy challenge. The study of bureaucratic behavior and the use of authority that were called

[7] The standard formula assumes, for example, that elementary teachers need to be at least graduates of secondary school or a secondary teacher-training school, while junior high school teachers, for example, need at least the equivalent of a zhuanke degree following two to three years of postsecondary training, and senior high school teachers should have completed a four-year college program.

on to meet this policy challenge illustrates three aspects of the process of educational policy formation within the vertical educational policy system: its slow, reactive, and groping character; the power of subordinates; and the limited authority available to education bureaucrats and the resulting strategies they adopt.

The Groping Pace of Bureaucratic Action

Bureaucratic wrestling with the issue of teacher policy has occurred outside any routine temporal cycle. In the case of teacher standards, superiors have placed no deadline. (Contrast this with what an SEdC official described as the one-year deadline the Central Committee gave the SEdC for formulating the draft of the "Education Decision.") Instead, the regulation of teacher standards appears to be a case of an issue looking for a policy, what Cohen, March, and Olsen might describe as "issues and feelings looking for decision situations" (1972, 2). What we see is a practice of management through groping (Paine 1986; Behn 1988).

The policy process in this case involved the simultaneous occurrence of goal setting, discussion, and implementation.^[8] Goal setting began in

[8] My description of this process and its local manifestations is based on field research in 1982–84, 1986, and 1987. As part of that work I conducted extensive interviewing, observation, and library research at Beijing Normal College, Beijing Normal University, Beijing Education Institute, Changchun Teachers College, Changchun Education Institute, Dongbei Normal University, Fujian Normal University, Harbin Teachers University, Huadong Normal University, Huazhong Teachers College, Jilin Education Institute, Jilin Teachers College, Liaoning Education Institute, Liaoning Teachers University, Shanghai Education Institute, Shenyang Education Institute, Shenyang Teachers College, and Siping Teachers College.

1978 with "problem recognition" (Kingdon 1984, 19): the shortage of qualified teachers was identified as a policy problem at the 1978 National Education Work Conference, the 1980 National Teacher Education Work conference, and in numerous articles published during this time. National leaders and high education officials gradually formulated a set of goals and principles for the reform of standards for professional preparation and practice,^[9] yet these statements have been distinguished by their vagueness and breadth. The most frequently cited statement of goals, for example, simply claims that teachers (1) "must study hard and become more erudite; (2) they must seriously study and grasp the science of education and understand educational laws; (3) they must have a noble moral character and a lofty spiritual realm and must be worthy of the title teacher" (FBIS , 30 June 1980, L1).

While national political and education leaders gradually announced policy goals, discussion of teacher standards and teacher education reforms grew. Since 1978 the topic got increasing scholarly and popular attention. Over one hundred articles were published on teacher education between 1978 and 1982 (Tan, 1983), and at least six new teacher-education journals were started between 1982 and 1984. Nonetheless, the discussion did not become markedly more specific. Instead, the discussion was mired in epistemological and fundamental questions, with the most discussion of teachers' professional standards (accounting for 46 percent of the literature) concerning the need for increased attention to be given teacher education and its "special characteristics" (Tan 1983).^[10]

Despite the absence of clearly defined goals, individual institutions, sometimes acting independently and sometimes acting in concert with the MOE or other schools, carried out numerous changes in all major areas of professional teacher preparation. Repeatedly in my interviews with school administrators, faculty, and education officials, respondents referred to this as a process of "groping" (mosuo ), conveying a sense of exploration, trying to find something out, trying to accomplish something. In the Western literature we might prefer the phrase "muddling through" (Lindblom 1959).^[11]

[9] See, for example, Deng Xiaoping 1983; Gao 1980; Gu 1983a, 1983b.

[10] While there grew to be more talk about the need to strengthen professional standards of entry, there was a noticeable absence of discussion of how teachers can best be trained or standards evaluated. A spate of articles in 1983–84 calling for clarification of the mission of teacher education offer a noteworthy contrast to Gao's 1980 assertion that the "basic mission" of teacher education programs is clear (Deng Liqun 1983; Gu 1983a, 1983b).

[11] Behn (1988, 649) distinguishes "groping along" from "muddling through." He argues that Lindblom's concept implies choosing "'among values and among policies at one and the same time,'" whereas "groping along" focuses on public management of various paths one could take to reach the desired goal. In my use of "groping" I agree with Behn that long-range, broad goals may be clear, but I find that these goals, once enacted, take on different meanings in the process of negotiation and the experience of trial and error.

The practice of management-by-groping results in an unevenness of policy activity. For example, during the 1980s bureaucratic and structural changes internal to the system of teacher education and its institutions occurred in a rather swift and uniform way; throughout the country, educators were promoted as leaders in teacher-education institutions, new institutions were established, and the system became more coordinated.^[12] Yet other responses to policy discussion—especially regarding curriculum, admissions, and job allocation—were more idiosyncratic, uneven, and even problematic. Curricular change is a particularly illustrative example of three key features of management-by-groping: local interpretation, mutual adaptation, and policy fluidity.

As early as 1978, schools turned to their curricula to experiment practically with how quality could be assured. Formally, the changes appeared to be carefully controlled by the ministry, with the announcement of precise policy formulations about professional preparation in 1978 and 1980 (through MOE-published jiaoxue jihua , or teaching plans for eleven departments in teacher-training colleges, and jiaoxue dagang , or teaching outlines for 140 courses offered at normal colleges.)^[13] Yet both the plans and the outlines are "reference" (cankaoxing ) documents, which act as guidelines rather than as regulations. Without enforcing power, the guidelines allow for some measure of autonomy for individual institutions of teacher education.

Schools experimented with curriculum reform, and the plans of most schools deviated in some way from the MOE guidelines. Local experimentation varied, but beginning in 1978 it generally tended first toward expansion of and specialization within the academic curriculum and the reduction of course work in politics and education and time spent in student teaching.

Alteration of the central policy subsequently occurred, demonstrating

[12] By the mid-1980s an extensive "teacher education network" of programs and resources was established. Previously uncoordinated institutions now worked in a comprehensive, highly differentiated, and well-articulated system. This included six regional centers coordinating preservice efforts and taking responsibility for research and sabbatical training of provincial and local-level teacher educators; 218 provincial, district, and municipal education institutes (jiaoyu xueyuan ) for retraining secondary school teachers; 2,174 teacher-refresher schools (jiaoshi jinxiu xuexiao ) for elementary teachers; correspondence programs; and by 1986 a national education television station and its "television teacher-training college" (zhongguo dianshi shifan xueyuan ) (Zhongguo jiaoyu nianjian 1986, 104–5; ZGJYB , 22 December 1987, 1; RMRB , 5 August 1986, 3; FBIS , 28 January 1987, K20).

[13] The plans list courses and placement in a student's program, while outlines specify the topics and their sequence for each course.

the interactive quality and mutual adaptation that are at the core of the groping policy process. In the keynote address to the 1980 National Teacher Education Conference, Gao Yi, as a central-level representative, warned against an overly academic curriculum. This served as a response to the experimentation of local units and represented pressure from the Center (Gao 1980). Reorientation of teacher-education programs followed. According to department chairs interviewed, some departments reduced their elective offerings, and others shifted the course content away from theoretically advanced work to "fundamentals." The political-theory core sequence was strengthened, and a new required course in moral education and an extra year of physical education were added. And the MOE, in refusing the request by some normal colleges to expand to a five-year, academically more extensive B.A. program, like those that Beijing Normal University and East China Normal had established, asserted the limits of acceptable reform.

In sum, then, under the guidelines of these vague policy discussions, this reform policy has had fluidity that allowed it to change over time. Typical of the groping process, the current standards represent an evolutionary compromise between the broad objectives of the central bureaucracy (that is, upgrading teacher standards) and specific experiments of local experience. Policy is recast by those carrying it out, somewhat akin to Manion's "policy remakes" (see chapter 8 in this volume), yet different in that here those involved are relatively weak actors with limited power to revise policy. Groping is characterized by responsiveness, as implementation proceeds alongside the process of continual formulation. This policy process is iterative. Thus, the broad goals and principles for strengthening professional standards have not changed, but the boundaries of acceptability shifted after a certain amount of local experience was collected. Reformulation continues.

Actors in the Process: The Power of Subordinates

At the Center the SEdC has moved to regulate and strengthen standards for teachers by improving teacher education—through the promulgation and revision of curricular guidelines and the establishment of an elaborate network of in-service and preservice teacher-education programs—and by introducing a system of professional examinations and certificates.^[14] Yet, we need to be wary of attributing too much leadership to the

[14] See GMRB , 2 October 1986, 1. The system of testing and credentials at the end of the decade remained in an early phase, having produced anxiety in the teachers. The education bureaucracy has had to spend some time responding to their fears and complaints. (See, for example, ZGJYB , 9 July 1987, 2.) When the system was tried out in six provinces, autonomous regions, and municipalities in 1987, only between 22 and 27 percent of the 430,000 secondary school teachers passed (ZGJYB , 10 November 1987, 1). In implementing the formal exam system, the SEdC will design the test for secondary teachers, but each province is determining the test for elementary teachers (ZGJYB , 26 December 1987, 1; FBIS , 29 December 1987, 43).

MOE and the SEdC. A review of participants highlights the limits of the Center and the resultant power of subordinates within the education bureaucracy.

The MOE and the SEdC are constrained horizontally in ways that weaken their efforts: the upgrading of teacher education depends first on attracting students of higher quality, something the MOE could not do on its own. Vertically constrained as well, the Center is limited in the extent to which it can direct and supervise activity. In the early 1980s, for example, the MOE's Teacher Education Bureau had a staff of only six. Shorthanded, it was only able to convene meetings with representatives from normal colleges to "exchange experience" and hold up for emulation the exemplary activities of individual programs. Its staff made occasional school inspections, but regular systematic review of all programs was impossible.^[15]

The result is a kind of autonomy that is rationalized as education intelligence: responding to local needs, encouraging grass-roots initiatives, and so on. While the references in the Education Decision to reforms in higher education grant official legitimacy to this, my 1982–84 interviews at eighteen higher-education institutions indicate that this constrained autonomy predated the Decision.

The autonomy outlined above lends strength to the influence and power of lower-level units. Subordinates have participated in the policy process in four main ways: experimenting, interpreting, and undermining policy, as well as reshaping policy conceptualization.

Experimentation

Local and provincial education authorities exert their power to shape policy by initiating experiments that later are disseminated provincially and nationally. In 1984, for example, a provincial official told me about one province's reforms in job assignment, which included movement toward more contractual arrangements in hiring recent graduates in teacher education. Because the reform was described as an "experiment," it did not require MOE approval.

This situation—played out in provincial, municipal, and institutional practices—illustrates the significance of local experimentation. Many times local and provincial officials and school administrators talked

[15] Instead of close supervision, according to a ministry official in 1984, special topics were chosen or a sampling of regional sites selected for investigation. University people, like MOE officials, were open about the lack of tight coupling. One university administrator in 1983 said that although the MOE "should come to understand our situation," they do not, because of the small staff.

about reforms they had initiated "as experiments"—in admissions to teacher education, curriculum, and professional standards. A commonly told story line runs like this: Some change in practice occurs that will serve the interests of local institutions, enterprises, or the local community's educational needs. An agreement is worked out, later noticed and encouraged by provincial or central officials, and finally widely implemented as a new development of the policy.

The process of experimentation, when it is successful, becomes one form of policy formation. The central bureaucracy, in fact, is organized to support that sort of persuasion-through-successful-experience. The teacher education and the teachers' conferences convened by the SEdC serve this function, as do the regularly featured stories of "successful" experiments of schools and local education authorities described by the education press, particularly Renmin Jiaoyu and Zhongguo Jiaoyubao . The value given to experimentation strengthens the negotiating position of local-level units. The recent distribution of rewards for reforms further encourages this.^[16] Speaking of the power of grass-roots experimentation, one provincial official explained in 1984, "In recent years the big reforms in teacher-training colleges have come mainly from the teacher-training college itself."

Interpretation

The influence of local institutions is also made possible in part by weak connections within the educational bureaucracy. Loose coupling between the Center and local areas has its parallel within individual institutions and schools. A de facto delegation of authority has important consequences for the power of lower-level actors to interpret and redefine policy.

The Dean of Studies office (jiaowuchu ), the unit responsible for overseeing academic work within the normal college, like the SEdC, conducts only limited regular close inspection of its charges. The jiaowuchu typically has only a few staff members in each of four or five sections. Given personnel shortages and weak or nonexistent hardware for management in most of these schools, these offices, like the SEdC, are kept busy simply reviewing department schedules and plans and doing occasional in-depth sampling. Even during the 1980s reforms they could not afford to do regular lengthy studies of program quality, the impact of standards, or the fate of curriculum programs in use. There is no regular feedback mechanism. As a result, the jiaowuchu must delegate much

[16] See, for example, the SEdC and Finance Ministry's award of 200,000 yuan to thirty-seven counties (ZGJYB , 10 September 1988) and the Shandong Provincial Government's decision to give the Pingdu County Education Commission a 100,000 yuan bonus and four individuals involved a 2,000 yuan bonus and a salary raise ("Zhonggong Shandongshengwei Shangdongsheng renmin zhengfu" 1988.)

supervisory authority to leaders of each academic department. What occurs is the regular delegation of responsibility to lower levels—whether from ministry and provincial bureau to the grass-roots institution or from the institution to its academic departments.

Given this delegated authority, departmental-level interpretation of central-level policy is significant. Personalized decision making enters in. Within institutions there were unclear goals, little sense of how they were to be achieved and of who should make decisions. As a result, who participated in the decision in each department was significant, as the arrival of a "problem" allowed participants to match the problem to their pet solutions. The decisions, therefore, varied from unit to unit.

One example of this personalism and the range in policy interpretation comes from the Center's effort to provide more qualified teachers to rural and remote areas. The dingxiang (fixed destination) program of admissions and job assignment was introduced nationally in 1983 as a means of filling rural teaching positions with rural graduates and thereby avoiding urban students' resistance to rural teaching assignments. Although the program officially commenced in 1983, interviews in 1983–84 with eleven department chairs at one teacher-training college revealed a range of interpretation of the policy and consequently of its implementation. Some department chairs described plans to implement the policy, others assured me that they were not planning to do so, and one leader said that he did not know what the policy was.

Undermining Policy

A third pattern of participation is to undermine or contradict policy. A variant perhaps of Manion's concept of "policy remakes" (see chapter 8 in this volume), this autonomy allows grassroots actors to willfully recast policy according to their interests. This policy process can produce reforms that are inconsistent and at times conflicting.^[17]

Reforming professional standards offers a good case of contradictory action. At one institution, one part of the curriculum was undermined by another, as each department interpreted "higher quality teacher educa-

[17] In this way and others the phenomena I describe differ from Manion's examples of remaking policy, for the final outcome in the teacher standards case—uncoordinated policy and policy sabotaged—is not necessarily the desire of any actor or group of actors, but instead is a by-product of relatively weak actors, each exerting limited power to recast policy in their interest. Of course, another significant difference is in the resources and power possessed by individuals involved in the process: Manion's cadres as individuals bring more strength to their negotiations than these departmental leaders, who are not really negotiating but are simply trying to cope in self-interested ways with policies handed to them.

tion" as meaning expanding its own course hours and offerings. In effect, departmental plans that simultaneously called for strengthened preparation in education and more advanced work in the students' major fields produced an increase in specialization at the expense of professional training in education. In the school's 1980 course plan, for example, education courses for all departments constituted only 2.7–5.8 percent of student course hours, in contrast to an earlier 10–20 percent. Struggles within colleges demonstrate ways in which lower-level units deflect decisions in ways they find congenial.

Shaping Conceptualization of Policy

In addition to provincial and university or school actors in the teacher policy process, researchers and consultants play an increasingly important role. Halpern (see chapter 5 in this volume) describes the growing role of research centers and experts as one of "competitive persuasion." In education the influence of researchers and consultants is evident in their ability to shape the conceptualization of policy.

The SEdC conducts its own research within each of the various bureaus and departments and assigns responsibility for policy research to its Office of policy Study and its Educational Development and Research Policy Center. The Policy Center researchers, as described by one educator, write for "leaders" in ways contrasted with the "more open" ability to "expose problems," which the many recently established research centers possess.

The contribution coming from these research institutes and consultants outside the SEdC grew in the late 1980s. The SEdC frequently called on researchers based at university research institutes to serve as consultants on teacher policy. At one university, for example, the SEdC asked one research group to prepare a paper on the academic specializations appropriate to teacher education, another to investigate the qualifications of all teacher educators involved in teacher-training institutes, and a third to analyze teacher standards in other countries. While noting that it is still "authority-driven policy," not "research-based policy," that is developed, researchers claim that the role of research in the policy process has increased since 1978. They find themselves called on more often for a wider range of tasks.

There seemed to be a common formal interpretation of the policy process: the administrative and political center raises an issue, the SEdC responds, and researchers get consulted. Still, this passive or reactive version of the consultant's role does not entirely coincide with the evidence of frequent and informal interaction between key researchers and SEdC officials. In 1987 several different researchers mentioned being

consulted by phone by newly appointed SEdC vice-ministers,^[18] called by an SEdC official to spend a morning talking about teacher education and this researcher's own research, and appointed on short notice to a small SEdC task force to evaluate the success of the reform efforts.

These phone calls, conversations, and last-minute committee assignments all suggest a rather close, informal, and personal network of scholars and officials, one that has grown stronger and perhaps tighter in recent years. Implicit in these anecdotes is the sense that the knowledge of and familiarity with academics is not insignificant in SEdC work. The expansion of the consulting role appears closely related to status—institutional status, personal connections, and status accrued from time overseas. (In teacher policy, researchers from Beijing Normal University and East China Normal appear to be most frequently consulted by SEdC officials.) As these ties grow, researchers also stand to influence policy—particularly the conceptualization of policy issues. As Kingdon found for the United States, these consultants are more likely to affect "alternatives" than national agendas (1984, 58).

Multiple Coping Strategies

The groping process of teacher policy described above rests on the interaction of several levels of actors. Each influences policy, although each is also constrained in its ability to form and carry out policies regarding standards for teachers. Acting with few material resources and both organizational and political constraints, education bureaucrats have pressed for their interests by relying on a range of coping strategies.

Many of the constraints on the education bureaucracy can be explained in organizational terms. In the absence of sufficient personnel and regular feedback mechanisms, weak articulation of central and local levels within the education system reduced the Center's ability to inspect performance at the grass roots. But local-level actors were also constrained, given their reliance on approval by the Center, which sets boundaries for acceptable action. All levels within the system are further constrained because the resources available to education are limited, with a budget representing 11.17 percent of government outlays (Yuan 1988, 24).

External factors also impose significant constraints on bureaucratic options within the educational system. The political, social, and economic climate for schools discourages talented people from entering and staying in teaching, and this situation profoundly limits the ability of the

[18] Recall that although some of the vice-ministers—fuzhuren —come with academic administrative experience, they are not education scholars. As one education researcher in late 1987 pointed out, these new people (the group of vice-commissioners which began in 1986) "don't know the research community yet."

SEdC to make standards for teachers (Paine 1986). A popular phrase sums up the problems about teachers: "You can't get them, can't use them, and can't keep them" (jinbulai, yongbushang, liubuzhu ) (Gu 1983a). Teacher-training colleges, despite reforms, are unable to recruit top students.^[19] Assuring that well-qualified graduates of teacher-preparation programs end up teaching in elementary and secondary schools has likewise proven a formidable task. Despite heightened attention to the issue, there remain many ways in which universities and their students can conspire to escape high school assignments, especially as nonteaching units hire normal-college graduates. The central education bureaucracy's inability to determine and enforce hiring policies (that is, which units employ graduates of teacher-education programs) and the reward structure for teachers simply compounds the difficulties.^[20]

Limited as they may be in resources and by the context in which they work, education bureaucrats use a range of strategies to achieve their policy aims. Three of the most common are linking to other agendas, piggybacking other education reforms, and spreading success stories.

Linking To Other Agendas

Kingdon (1984) argues that changes in the "political stream" and, more important, agendas are significant. Changes in China's national agenda gave force to the reforming of teacher standards. Education actors were keen to link their interests with those of national reform. It is noteworthy, for example, that in an authoritative book on major policy documents of the educational reforms, the 1984 Economic Reform Decision and the 1985 Science and Technology System Reform were included (Jiaoyu gaige 1986). The reform of education is regularly and symbolically connected to other broader changes. Teacher-education respondents frequently related the bureaucratic reforms within their system to changes in economic enterprises or the agricultural responsibility system. For administrators at the grass-roots level there was a keen awareness that the bureaucratic changes associated with reforming teacher standards were part of a

[19] Reforms include early university admissions for teacher education applicants and the continued use of stipends to teacher-training students despite the elimination of stipends in much of the rest of higher education. (See, for example, GMRB, 13 January 1985; FBIS, 28 July 1986, R1; Liu Bin 1986, 6; ZGJYB, 12 July 1986, 1, and 6 June 1987, 1; Xue 1986). The early-admissions program, for example, was described by SEdC, provincial BOE, and university admission officials as effective in attracting a stronger student-to-teacher preparation, yet all noted the limits of this approach. As one university official explained in 1986, the candidates were "not the best, but [academically] upper-middle students."

[20] The SEdC vice-commissioner Liu Bin argued that recently there has actually been a decrease in the proportion of graduates entering precollegiate teaching, with some schools only managing to send 50 percent and others only a few percent (Liu Bin 1986, 6).

larger national agenda. For them, this connection gave persuasive power.

Thus, many of the changes at the local educational institution intended to raise the quality of teachers corresponded with the restructuring of political economic patterns nationally. As part of the process of policy change in teacher standards, for example, autonomy in some areas of educational work increased: using the rhetoric of decentralization, some provinces were able to vary from the national plan to institute early-admissions policies in teacher education that were to increase their chances of attracting strong candidates into teaching. This phenomenon of increased autonomy was justified by the renewed emphasis on expertise and decentralization elsewhere in the system.

Using Momentum of Other Educational Reforms

If grand national changes in the political economy's landscape justify changes in education policy, even more forceful pressure for reform in teacher standards came from other educational reforms. Advocates for higher standards for teachers were able to use the 1985 Education Conference and subsequent Education Decision to increase the pace and specificity of their reforms far beyond what had occurred during the broad discussions in 1978–85. In December 1985 attention focused on teacher examinations, preservice and in-service teacher education, and restrictions on teacher mobility ("Yao zhuajin" 1986, 2–3). Soon after, in March 1986, the SEdC made specific recommendations regarding curriculum, admissions, and job assignments—areas that had previously experienced experimentation but no clear SEdC mandate.^[21] Similarly, policy for teacher standards gained momentum with the April 1986 Compulsory Education Law's discussion of teacher testing and certification; this has encouraged a subsequent increase in both the level of attention and the degree of specificity of policy regarding teacher standards, with the SEdC and provinces moving ahead in developing and administering tests of teacher knowledge.^[22]

Closet Reforms and the Spreading of Success Stories

The groping and incremental quality to the policy change and the constraints felt by individuals at all levels encouraged official conservatism and unofficial experimentation. In the early unfocused stages of reform, the tendency to interpret policy literally and the vulnerability of educational institutions

[21] These recommendations were made in the "Opinion on Strengthening and Developing Teacher Education." For more, see "Guojia jiaowei yaoqiu" 1986, 9; RMRB, 22 April 1986, 1.

[22] See "Zhonghua renmin gongheguo yiwu jiaoyufa" 1986, 2; FBIS, 4 April 1986, K7; "Quanguo gaodeng shifan jiaoyu yanjiuhui" 1987; "Gaodeng shifan jiaoyu ruhe shiying" 1987; ZGJYB, 10 November 1987, 1, and 17 December 1987, 2.

generally encouraged a wait-and-see attitude. Many university officials whom I interviewed expressed this by saying that local reforms depend on reforms elsewhere; they did not feel that their school could go alone in its reform. (It is also clear, however, that higher-status organizations and individuals acted more boldly in interpreting reforms on their own.)

Despite this general conservatism, however, I observed a striking creativity among subordinates as departments or faculty conducted what I call "closet reforms." These reforms often represented significant changes from previous practices, yet their creators seemed to avoid publicizing them until informal support was gathered horizontally (at other institutions) or vertically (in the MOE, the SEdC, or the provincial BOE). An example would be one normal college's early revision of admissions practices before the dingxiang policy was created. The university, faced with graduates unwilling to teach at the secondary level, began to take proportionally more students from the countryside because they had, in the words of one administrator, fewer "conditions" and were more compliant. Within the space of one year the rural share of the entering class jumped from 40 percent to 60 percent. Soon after, the dingxiang program was introduced nationally. The experimentation observed at the local level was generalized as a popular coping strategy for the education sector as a whole.

In sum, the case study of teacher-standards policy represents many common qualities of the educational policy process. This groping process was shown to be interactive and slow and to involve many levels of the educational system. The SEdC, the BOE, institutions, and researchers all have some influence, but each is limited and is therefore forced to rely on multiple strategies for achieving their policy goals. The strategies vary with the unit's location in the structure of the education system's hierarchy, yet generally three kinds of strategies stand out: linking to changes in the political stream, riding the coattails of other educational reforms, or disseminating successful experiments.