I first presented this paper at the conference on “The Century of the School: Continuity and Innovation During the First Half of the 20th Century,” Monte Verità, Ascona, Switzerland, September, 2007. A revised version was published in a book edited by Claudia Crotti and Fritz Osterwalder, Das Jahrhundert der Schulreformen: Internationale und Nationale Perspektiven, 1900-1950, where it promptly disappeared. Here’s a link to that version. A shorter and substantially revised version appeared in 2010 as chapter three of my book, Someone Has to Fail: The Zero-Sum Game of Public Schooling.
Over the course of my career, I’ve frequently felt compelled to rescue orphaned papers that appeared only flittingly in print and I thought deserved a better fate. So I published them as chapters in one of my own books. Waste not, want not. This is one of those papers.
Here’s an overview of the argument:
At least in the United States, the primary impact of educational reform efforts on elementary and secondary schooling has been at the periphery rather than the core of this institution. These pressures from the social context have been able to exert a major impact on the rhetoric of education, and the most effective reform efforts have even been able to shape the formal structure of schooling; but they typically have had little impact on how teachers teach and even less on what students learn. In this paper, I present a four-level model for understanding the relative inability of school reform in the U.S. to reach the core of teaching and learning in classrooms. Then I examine the movement for progressive education in the U.S. in the first half of the 20th century as a case for understanding the limitations of reform. Looking separately at the two main branches of this reform movement, I argue that pedagogical progressivism (John Dewey’s version) shaped the rhetoric of education but never penetrated the institutional practice of schooling, whereas administrative progressivism (the version pursued by David Snedden and Elwood Cubberley) succeeding in changing rhetoric and in reshaping the formal structure of schooling, especially at the secondary level, but failed to extend much beyond this to shape the practice of teaching and learning in classrooms. In the final section of the paper, I explore the reasons that U.S. reform movements in general have had more success changing the form of schooling rather than its instructional substance (loose coupling and weak administrative control), and I also explore some facets of the practice of teaching that make reforms at the classroom level potentially difficult anywhere in the world.
See what you think.
Limits on the Impact of Educational Reform:
The Case of Progressivism and U.S. Schools, 1900-1950
David F. Labaree
Educational reform is difficult in part because any large scale effort to transform social institutions is difficult. Peter Rossi, who has evaluated a number of such programs, has formalized these difficulties in a series of “laws,” starting with “The Iron Law of Evaluation: The expected value of any net impact assessment of any social program is zero.” One corollary of this law is that “The better designed the impact assessment of a social program, the more likely is the resulting estimate of net impact to be zero.” Another is that “The more social programs are designed to change individuals, the more likely the net impact of the program will be zero” (Rossi, 1987, pp. 4-5). He argues that this track record of failure and minimal impact is the result of the complexity of social institutions, which in turn makes it difficult to discern how they work, formulate valid plans for intervention, and successfully implement these plans.
What is true for social reforms in general is particularly true for education, where social complexity is high, where measuring improvement is a challenge, and in particular where changing people is the central task. Education, after all, is an exercise in changing students in valued directions; and educational reform is an even more demanding practice, which seeks to change the attitudes and behaviors of a vast array of administrators, teachers, and students in the direction intended by reformers.
In his book on educational reform, The New Meaning of Educational Change, Michael Fullan summarizes the problem this way:
Educational change is technically simple and socially complex…. A large part of the problem of educational change may be less a question of dogmatic resistance and bad intentions (although there is certainly some of both) and more a question of the difficulties related to planning and coordinating a multilevel social process involving thousands of people. (Fullan, 2001, p. 69)
He devotes separate chapters to the major actors who populate and complicate the system of schooling: students, teachers, principals, district administrators, consultants, parents, communities, and governments. Trying to mobilize all these actors to move in the same direction is enormously difficult, and Fullan devotes the latter part of his book to a discussion of how to accomplish this, both by building capacity for the actors to pursue change and by developing forms of accountability that will maintain pressure toward particular reform goals.
Larry Cuban (1988) zeroes in on a key component of the problem of school reform, arguing that the prospects for educational change vary according to the degree that reform efforts aim to transform the central functions of schooling. At the core of schooling, he says, are the academic programs for educating the ordinary child; at the periphery are the programs for students who don’t fit into the core curriculum and instead find themselves in areas such as vocational education, special education, and gifted education. He notes that over the years reformers have been much more likely to effect substantial change at the periphery of schooling than the core.
Richard Elmore and Milbrey McLaughlin (1988) see the problem of school reform in organizational terms, arguing that the key variable in shaping the success or failure of reforms is the ability of intended change to move across the levels of the school hierarchy. In particular they point to three levels in the educational system: policymakers, school administrators, and practitioners. Each of these sets of actors occupies a critical part in the reform process. Ideally, the first frames and initiates the reform effort, the second organizes and facilitates it, the third adopts and implements it. But each group of actors lives in its own distinct social and cultural world, and communication across boundaries is inherently problematic. These differences make it difficult for change efforts to move from one level to the next, which means that most reforms get stalled at the door of the school administrator or the classroom teacher.
Four Levels of School Reform
In combination, these authors depict a reform process that has difficulty dealing with education’s organizational complexity, overcoming resistance to change in its core functions, and moving reforms down the hierarchy and into the classroom. Building on these insights, I have found it useful to model the complexity of schooling, like Elmore and McLaughlin, as a matter of organizational hierarchy, but extended to four levels, with the Cuban’s core functions of schooling occupying the lowest levels. Each level has its own peculiar set of actors; and each set of actors occupies a distinctive ecological niche with its own language, media of expression, tools, organizational incentives, and problems of practice. From this angle, the challenge for reformers is to try to move reform down to the core levels of the system without letting it get blocked, deflected, or diluted. Cuban provides a succinct description of the way this system frustrates reform: “Organizations, I have discovered, have plans for reformers” (1988, p. 101). Although it may have applications in other countries, the four-level model of school reform is drawn from historical experience in the United States, and I focus my attention on this setting.
At the top is the level of reform rhetoric. This is where most reform efforts begin (and frequently end), with statements of principle, educational visions, rationales for change, frameworks for representing that change, and norms for reconstructed educational practice. The actors are a mixture of policymakers, lawmakers, education professors, judges, and educational leaders; and their primary media are reform reports, speeches, policy papers, journal articles, laws, and court rulings. The rhetorical level is the most open to reform efforts, since the actors are part of the same discourse community and thus are in tune with rhetorical currents running through this community. This is also the level of reform that is best represented in the history of education. With its national scope, reform rhetoric acts as an efficient mechanism for historians to use in representing the essence of a reform movement; the rhetoricians can conveniently serve the role of movement leaders in the stories historians tell about reform; and the reform documents are easy to find in the library. Partly as a result, most histories of reform are largely histories of reform rhetoric, saying little about the changes farther down in the system.
Next is the level of formal structure. This is where reform rhetoric needs to be translated into key components of the organizational structure of schooling at the district level, such as educational policies, organizational units, curriculum frameworks, classroom textbooks, and professional development workshops. The actors are school administrators, school board members, curriculum developers, textbook publishers, and workshop leaders. The level of formal structure is harder for reformers to influence, since such an effort means translating national reform principles into forms that will work within the 14,000 local school districts currently in existence in the U.S., each with its own organizational structure, school politics, and modes of operation. Historians of education traditionally do not focus as much on reform at this level. To do so requires locating documents (if they exist at all) in dispersed locales, trying to interpret them in light of the local situation, and then struggling to relate this complex set of structural changes back to the rhetoric of national reform texts. There are some interesting efforts to trace reform at this level, especially with school curriculum (e.g., Angus & Mirel, 1999; Franklin, 1986), but the numbers of such studies are small in comparison with the studies of reform rhetoric.
Third is the level of classroom practice. This is where reform ideas, if they succeed in passing through the machinery of the school district largely unscathed, need to pass through the door of the self-contained classroom in order to merge into the professional practice of teaching. The actors here are teachers, and the key issue for school reform is the extent to which these practitioners adapt the content and process of their instruction, both to the principles of reform rhetoric and to the local structure of reform implementation. At this level we arrive at the instructional core of the educational enterprise, and reform success depends entirely on the capability and willingness of 3 million American public school teachers in 95,000 schools (using the current U.S. numbers) to take on the reform agenda and put it into practice in their classrooms. I will explore the problem of reform at the classroom level in more detail at the end of this paper; but on the face of it the sheer number, diversity, and geographical dispersion of these teachers and classrooms suggests how difficult it is to effect reform at this level of the educational system. If it is extraordinarily hard for reformers to shape the practice of teaching in individual classrooms, it is equally hard for historians to trace how teachers taught in the past. The available evidence is remarkably spotty, indirect, and difficult to locate. Larry Cuban (1993, 2007) made a heroic effort to determine “how teachers taught” in the 20th century U.S., and Barbara Finkelstein (1989) tried to sketch the outlines of practice in the 19th century; but there are few others who have even tried.
Last of all is the level of student learning. Even if a particular reform effort improbably manages to shape the rhetoric of schooling, alter the structure of some school districts, and penetrate the practice of teachers in some classrooms, it still needs to transform the learning that students take away from their classroom experience if it is going to be declared a success. The key actors here are the students; and for reasons that I will explore at the end of this paper, they constitute a final barrier to reform that may be the most formidable, especially under the peculiar conditions of teaching and learning that characterize schools in the United States. Needless to say, historians have almost nothing to say about how educational reform in the past shaped student learning. Even if they can find some data on the curriculum that teachers were supposed to teach and some hints about whether and how teachers actually taught it, they are unlikely to have any evidence that students really learned what the reformers intended.
The problem of school reform, then, is the question of whether reform efforts can move through all four levels of the system, from the rhetorical periphery to the core of teaching and learning in classrooms. Is the reform able to persuade educational opinion leaders at the rhetorical level, reshape the formal structure of schooling at the district level, remold the practice of teachers in individual classrooms, and reconstruct the learning of students in these classrooms? Even though teaching and learning in classrooms are the only levels that really matter in educational reform, the historiography of U.S. education suggests strongly that reform movements in general have not been successful in reaching down the levels of the system to reconstruct the instructional core. The more successful movements managed to reshape the formal structure of schooling, even if they didn’t extend much beyond this point; the less successful movements never moved past the level of rhetoric.
Let us consider the case of one of the most high-profile and long-lasting educational reform movements in the history of American education, the movement for progressive education in the first half of the 20th century. Of the two strands of this movement, I argue that one strand reshaped the rhetoric of education but had little effect beyond this level, while the other one reshaped the rhetoric and the formal structure of schooling before being largely blocked at the classroom door.
The Case of Progressivism:
A Short History of a School Reform
Initially an outgrowth of the political movement known as progressivism, the movement for progressive education emerged at the end of the 19th century in the U.S., became a major force in the first decade of the 20th century, and lasted for another 50 years. Its major chronicler, Lawrence Cremin, sets its end date at 1955, when the Progressive Education Association finally disbanded. As Cremin noted, progressive education was not one movement but several, since
the movement was marked from the very beginning by a pluralistic, frequently contradictory, character. The reader will search these pages in vain for any capsule definition of progressive education. None exists, and none ever will; for throughout its history progressive education meant different things to different people…. (Cremin, 1961, p. x)
Historians disagree about how to label the various tendencies within the movement. Herbert Kliebard (1986) identifies three, which he called social efficiency, child development, and social reconstruction. Robert Church and Michael Sedlak (1976) see two forms, which they call conservative and liberal. David Tyack (1974) calls these two tendencies administrative and pedagogical. I will use the latter terms, which have become the most common (Rury, 2002).
These two strands of the progressive education movement were strikingly different in their effect on the American system of schooling. Ellen Lagemann put the difference this way:
I have often argued to students, only in part to be perverse, that one cannot understand the history of education in the United States during the twentieth century unless one realizes that Edward L. Thorndike won and John Dewey lost. The statement is too simple, of course, but nevertheless more true than untrue and useful for several reasons. First, it suggests that, even if Thorndike and Dewey both spoke and wrote in the “progressive” idiom, the differences of view that separated them were large and significant. Beyond that, it calls attention to differences in the way each man’s ideas were received. If Dewey has been revered among some educators and his thought has had influence across a greater range of scholarly domains – philosophy, sociology, politics, and social psychology, among them – Thorndike’s thought has been more influential within education. It helped to shape public school practice as well as scholarship about education. (Lagemann, 1989, p. 185)
If Dewey and the pedagogical progressives lost the fight for reshaping the structure of schools, they nonetheless exerted a major and continuing impact on the rhetoric of education. In contrast, Thorndike and the administrative progressives managed to effect enduring changes in both rhetoric and the formal structure of schooling, but had a limited impact on the core of teaching and learning. To explore the difference between the two reform efforts and the limitations on both, let us consider the aims of each of these tendencies in the progressive movement, the impact each had on the educational system, and the reasons for the greater, if still limited, success of the administrative progressives.
Understanding the Two Types of Progressivism
The two tendencies had several orientations in common, which helped justify the common label applied to them and which often put them on the same side in reform efforts. One is that they both shared a grounding in developmentalism. Both sought to establish a system of education that was designed around the needs and capabilities that students had at different stages of their intellectual and social development. But the strongest link between them was their antagonism toward the traditional academic curriculum. They both detested the practice of basing school subjects in academic disciplines instead of tailoring these subjects to the practical needs of modern life. The two strands of progressivism, however, took these two common orientations in very different directions.
Pedagogical progressives saw developmentalism as a rationale for rejecting the traditional curriculum in favor of classroom processes that would harness individual student interests and abilities and would foster engaged, self-directed learning. A broad vein of romanticism runs through this form of progressivism, which saw learning as a natural process that would occur best if artificial mechanisms like schools and curricula would just get out of the way of children’s natural urge to learn. Pedagogical progressives resented the traditional curriculum because they saw it as a conspiracy to impose adult mindsets and adult roles on children who needed the freedom to explore childhood and develop their own learning styles. They were more concerned about having students learn to learn rather than learn the curriculum. In their view, the ideal school would revolve around student initiative and student inquiry; focus on discovering knowledge instead of presenting knowledge; emphasize active and engaged learning over passive recitation of the text; organize study around projects, which drew on student interests and which synthesized knowledge and skill from multiple disciplines; and create a school as a democratic community, modeling values of justice and cooperation for later life. In addition to Dewey, some of the major players in pedagogical progressivism were William Heard Kilpatrick, Harold Rugg, and Boyd Bode.
Administrative progressives eschewed the romanticism of their pedagogical counterparts in favor of a hard-headed utilitarianism. Instead of Dewey’s focus on naturalistic teaching and learning, they tended to focus on school governance, professional administration, and scientifically designed formal curriculum. The two main principles of administrative progressive reforms were social efficiency and differentiation. Social efficiency meant training and empowering professional administrators to manage school systems instead of leaving them at the mercy of school boards filled with amateurs. It meant organizing a system of education that would prepare graduates to play future roles as productive workers and capable members of the community. It meant shifting the curriculum away from academic and toward vocational education. It meant sharply differentiating the curriculum and the whole school experience, so that schools would mirror the differentiated patterns of work and life. And it meant aggressively testing student abilities in order to place them in the correct subjects and at the correct ability level in that subject. In addition to Thorndike, some of the major administrative progressives were David Snedden, Elwood P. Cubberley, Charles H. Judd, and John Franklin Bobbitt.
Assessing the Impact of the Administrative Progressives
In the first half of the 20th century, administrative progressives had a substantial effect at the rhetorical level. In particular, they gave public credibility to the idea that the primary goals of education are the production of human capital and the promotion of social efficiency. The movement’s most prominent rhetorical expression, The Cardinal Principles of Secondary Education (1918), made these points with force and clarity on the opening page of the report:
Within the past few decades changes have taken place in American life profoundly affecting the activities of the individual. As a citizen, he must to a greater extent and in a more direct way cope with problems of community life, State and National Governments, and international relationships. As a worker, he must adjust himself to a more complex economic order. As a relatively independent personality, he has more leisure. The problems arising from these three dominant phases of life are closely interrelated and call for a degree of intelligence and efficiency on the part of every citizen that can not be secured through elementary education alone, or even through secondary education unless the scope of that education is broadened. (p. 1)
The vision of social efficiency carried the day during the progressive era and continued into the latter part of the century as a central theme in the politics of American education. It stood in sharp contrast to the rhetoric of the common school movement, which established the American public school system in the second quarter of the 19th century and set the rhetorical tone for American education until the progressives arrived on the scene. The common school leaders provided a purely political rationale for public schooling, portraying it as a mechanism for promoting republican community and creating capable citizens. Its most forceful leader, Horace Mann (1841), could barely conceal his contempt for the proposition that these schools might also provide vocationally useful skills and promote economic prosperity. But in the early 20th century the administrative progressives managed to shift the rhetoric about the purposes of American education from an emphasis on democratic equality to an emphasis on social efficiency. Diane Ravitch in Left Back, her study of administrative progressivism, argues that the movement’s ideas reshaped educational discourse in a way that diminished the intellectual aims of education in favor of a focus on “job training, social planning, political reform, social sorting, personality adjustment, and social efficiency” (Ravitch, 2000, p. 459).
Administrative progressive reform reached past the rhetorical level and exerted a substantial impact at the formal-structural level of the school system, with particular emphasis on the organization of school systems and the structure of the formal curriculum. Especially in urban school systems, they succeeded in professionalizing school administration and consolidating governance in the hands of small elite school boards. In rural areas they launched a remarkable process of consolidation, which during the 1930s alone reduced the number of school systems by 10,000 (Bureau of the Census, 1975, Table H 412). Their impact on curriculum was significant if less dramatic. They sought to broaden traditional school subjects in order to shift them from narrow preparation in academic disciplines to broad preparation for work and life. Their greatest success along these lines was the substitution of social studies for history, but they also introduced courses in general math and general science. In addition, they added a set of explicitly vocational courses aimed at preparing students for clerical, technical, and industrial work; and they added courses like home economics and physical education that were aimed at “life adjustment.” They also created a differentiated curriculum that organized courses in curriculum tracks, which were distinguished from each other by the students’ level of tested academic ability and by their projected future occupational role. When Lawrence Cremin summarizes the impact of progressivism on schools in his comprehensive history of the movement, he enumerates changes that represented primarily the aims of the administrative progressives and that affected primarily the formal structure of schooling. He lists such changes as introducing the junior high school, expanding the curriculum into nonacademic areas, expanding the extracurriculum, differentiating courses by vocational trajectory, grouping students by ability, and professionalizing school administration (Cremin, 1961, pp. 306-308).
A systematic study of the 20th century high school curriculum in the U.S. by David Angus and Jeffrey Mirel (1999) shows both the extent and the limits of the impact of administrative progressives. They identify a sharp increase in the differentiation of the curriculum by ability level, career trajectory, gender, and social class. They also find a substantial rise in vocational and life adjustment classes and a declining share of academic courses. But the latter changes were less dramatic than the increase in differentiation. Vocational classes never rose about 20% of the curriculum and academic subjects hardly ever fell below 60%. The largest growth area in nonacademic courses was not vocational classes but physical education, and the largest drop in academic studies was in foreign languages rather than the core academic subjects. By the 1950s, the curriculum was much more differentiated than it had been at the start of the century, but the bulk of schooling still focused on English, math, science, and social studies, and most of the latter was still devoted to history. High school teachers still identified themselves by their disciplinary specialty rather than by the new vocational and life-adjustment labels (Angus & Mirel, 1999, tables A.2, A.5, and B.11).
The evidence is strong that the administrative progressives had a major impact on American schooling at the rhetorical and structural levels, but the evidence suggests a much weaker impact at the levels of teaching and learning in classrooms. That they were successful in shaping the formal structure of schooling, however, did have some necessary consequences for how teachers taught and what students learned. By defining the outlines of the curriculum and shaping the focus and content of textbooks, they were able to put constraints on what was likely to happen in the classroom. The reformed administrative structure of the school system was designed to facilitate implementation of the new curriculum. Teachers were assigned classes to teach that were defined by the curriculum, and the content they introduced to their classes was influenced by the content in the available textbooks and other curriculum materials. Students, in turn, were more likely to be exposed to some kinds of content over others; and this differential access to knowledge increased the likelihood of some kinds of learning over others.
But these constraints are not as restrictive and as consequential for practice as they might seem in formal terms. As I will explain more fully in the last section of this paper, over the years American school administrators have had a strong hold on the administrative apparatus of the school but have been lacking in the necessary tools required to influence the way teachers taught behind the doors of their classrooms. And American teachers, in turn, have been in a weak position to motivate students to learn the content they have been trying to teach. Reshaping curriculum and administration is not as strong an intervention in the core instructional work of schools as it might initially appear. I defer until later a more detailed discussion of these limitations on administrative control over classroom discussion. But one aspect of the approach to reform taken by the administrative progressives served to reinforce their generic weakness in shaping teaching and learning in classrooms: Their primary focus was elsewhere. This was a movement aimed at the formal structure of schooling and not at the instructional core. Many of its leaders were school superintendents and its primary reform target was other superintendents. Its focus was on changing the way school systems were administered and the way curriculum was structured, and they succeeded in both goals, in part because they limited their aspirations and they targeted the elements of schooling that administrators had most directly under their control. A close reading of The Cardinal Principles report and other major documents in the administrative progressive canon reveals almost no mention of teachers. There is an assumption that teachers will do what the administrators and curricula require, but there is no real effort to address the problem of how to bring teaching in line with administrative expectations.
Student learning was also something that was assumed by the administrative progressives instead of being actively facilitated as a central part of the reform process. This strand of the progressive movement was grounded in Thorndike’s theory of learning, which argued that learning is largely not transferable. This means that the curriculum should not focus on training core faculties through exercises in arbitrary substantive domains. The latter approach was mandated by the once-dominant learning theory known as faculty psychology, which saw the study of classical languages as a mechanism for training the mind for future learning in any area. In contrast to the faculty psychology approach, Thorndike argued that students’ learning is heavily dependent on the content to which they are exposed. This meant that schools needed a scientifically differentiated curriculum, which corresponded with all the domains of practical knowledge that students would need as workers and citizens and parents. The beauty of this theory was that it put curriculum in the driver’s seat. Once you had set up a carefully designed curriculum to cover all the knowledge students needed, learning naturally followed. Teachers were just there to deliver the curriculum; it was the curriculum that taught the students.
The key was to match up the student with the right curriculum track, and this was where testing came in. The administrative progressives were intensely interested in testing and were aggressive in introducing standardized tests into schools in a massive way. But the testing they emphasized was intelligence testing, which allowed them to place students in the correct track according to each person’s ability level. They did not, however, focus on measuring how well students actually learned the curriculum to which they were exposed. Tests to measure learning outcomes were an invention of a reform movement that became prominent in the U.S. at the end of the 20th century, the standards movement, whose leaders have learned from the failure of the administrative progressives and deliberately use testing to insure that curriculum mandates actually lead to the desired student learning.
Assessing the Impact of the Pedagogical Progressives
In the first half of the 20th century, pedagogical progressives had almost no success in shaping schooling at the level of formal structure. The best evidence for this is the success of the administrative progressives at the same level. For the primary accomplishments of the latter were anathema to the pedagogues; the success of the one denoted the failure of the other. The focus on curriculum by administrative progressive reforms in this era was itself a major problem, since it ran against the grain of the effort by pedagogical progressives to tailor instruction to the interests and initiative of the child. The social efficiency curriculum was designed as a deliberate imposition on children, channeling them in the direction the reformers thought they should go, for the social and economic good, whereas the child-centered form of instruction was supposed to subordinate curriculum to the student. And the form assumed by the reformed curriculum was particularly abhorrent to them. The emphasis on creating an elaborate differentiated structure of curriculum tracks and then assigning students to experience schooling solely within the confines of these tracks introduced a rigidity to schooling that made it impossible to achieve the kind of naturalistic student-directed pursuit of learning they envisioned.
The theory of learning behind the social efficiency curriculum also ran in the opposite direction from that of the pedagogues. Whereas the administrative progressives saw learning as confined to the specific domain being studied, the Deweyan progressives were proponents of the transferability of learning. For the former, the curriculum was everything, the object of learning; for the latter, curriculum was a medium through which learning took place, a means to an end, and a variety of means would do. In fact, for the pedagogues it was dysfunctional for learning to be trapped in the kind of particularized curricular cul-de-sac that administrative progressive reform mandated for the scientifically tracked student. This was exactly the kind of curriculum that Dewey (1902/1990, p. 205) warned against: “externally presented material, conceived and generated in standpoints and attitudes remote from the child, and developed in motives alien to him.”
The pedagogical progressives also had little luck in introducing their reforms at the levels of classroom practice and student learning. But unlike the administrators, this was not for want of trying. The administrative progressives focused on administrative structure and formal curriculum, and they assumed that appropriate teaching and learning would naturally follow; the pedagogical progressives, however, ignored the structural level and focused entirely on the processes of teaching and learning within individual classrooms. The success of the administrators at the structural level is in part because this is where they concentrated their energies and where they engaged their primary constituency of other administrators. The failure of the pedagogues at the structural level is in part because they discounted the importance of this level and focused on the levels below it. Yet just because they concentrated their energies on teaching and learning didn’t keep them from failing there as well. This is what gives their failure at these core levels the tenor of tragedy. Proponents of the Deweyan credo have been railing against their fate ever since the second quarter of the 20th century, when the dimension of their failure started becoming apparent.
Arthur Zilversmit (1993) and Larry Cuban (1993) are two historians who have examined the impact of pedagogical progressives on the practice of teaching, and both concluded that this impact was modest and transitory. Ravitch (2000, p. 527, fn 6) agrees with their assessment. Cuban looked at a variety of sources (such as photographs, district studies, accounts by visitors, classroom layouts) to determine the extent of child-centered instruction that was actually taking place in classrooms in New York, Washington DC, and Denver during the second quarter of the 20th century. He classified classrooms as pedagogically progressive if they demonstrated progressive characteristics in at least some of five domains: class arrangements (classes organized in student clusters rather than rows), grouping (students working in small groups), talk (students talking a lot and initiating talk in the class), activities (projects, small groups, independent study), and movement (students free to leave their desks). Even using these generous proxies for child-centered instruction (any one which could be adopted without really carrying out the spirit of Deweyan progressive teaching), he found that the evidence for it was slight. In New York City from 1920 to 1940, for example, he found that “No more than an estimated one of four elementary teachers, and an even smaller fraction of high school teachers, adopted progressive teaching practices, broadly defined, and used them to varying degrees in the classroom” (Cuban, 1993, p. 75). The most likely elements of the canon to appear in the classroom were activity and movement.
Zilversmit examined the evidence of pedagogically progressive teaching in classrooms in the Chicago area in the 1930s and 40s. His conclusions about this period:
Despite the impassioned discussions of progressive education in the 1920s and 1930s, despite the marked progressivism of a few school districts and the increasing importance of progressive ideas in state education departments and teachers’ colleges, it is clear that by 1940 progressive education had not significantly altered the broad pattern of American education. The call for a child-centered school had, for the most part, been ignored. (Zilversmit, 1993, p. 34)
His explanation for this outcome is this “The ultimate failure was that so much of progressivism’s apparent success was rhetorical. While some schools and individual teachers had heeded Dewey’s call for a more child-centered school, most had given only lip service to these ideas while continuing older practices” (Zilversmit, 1993, p. 168). The pattern that these and other historians have noted is that child-centered progressivism was largely confined to a few independent schools, whereas social-efficiency progressivism dominated in the public school systems (Semel, 2006, p. 13).
If the pedagogical progressives largely failed in their primary aim, to reshape teaching and learning in public school classrooms around the principles of child-centered instruction, they did succeed in effecting reform at one level of the educational system. As Zilversmit points out, that was at the level of rhetoric. This was not a minor accomplishment. As Cremin points out in his summary of the state of the progressive project at the middle of the 20th century, American educators had by then all come to talk about education in the language of pedagogical progressivism:
There is a “conventional wisdom,” to borrow from John Kenneth Galbraith, in education as well as economics, and by the end of World War II progressivism had come to be that conventional wisdom. Discussions of educational policy were liberally spiced with phrases like “recognizing individual differences,” “personality development,” “the whole child,” “social and emotional growth,” “creative self-expression,” “the needs of learners,” “intrinsic motivation,” “persistent life situations,” “bridging the gap between home and school,” “teaching children, not subjects,” “adjusting the school to the child,” “real life experiences,” “teacher-pupil relationships,” and “staff planning.” Such phrases were a cant, to be sure, the peculiar jargon of the pedagogues. But they were more than that, for they signified that Dewey’s forecast of a day when progressive education would eventually be accepted as good education had now finally come to pass. (Cremin, 1961, p. 328; emphasis in original)
In the last sentence, Cremin tries to stretch this rhetorical accomplishment into something more substantive. But the weight of this paragraph – in conjunction with the weight of the historical evidence about the impact of pedagogical progressivism on teaching and learning – suggests strongly that the impact was primarily limited to the way American educators talked about schooling rather than the way they practiced it. As a result, we had come to use the child-centered language of discovery, engagement, and inquiry to talk about a structure of schooling, a practice of teaching, and a process of learning that are deeply grounded in the social efficiency principles of differentiation and human capital production.
How can we best understand the failure of the pedagogical progressives to extend reform beyond the rhetorical level at the same time that the administrative progressives managed to reach the structural level? As I have pointed out earlier, one reason was that administrative progressives focused their attention at a level that they, as administrators, could realistically affect, while the pedagogical progressives tried to transform the levels of teaching and learning at the heart of the educational system, where change is particularly difficult and where they had no special leverage, especially with the formal structure and the curriculum in the hands of their opponents. Another reason is more basic, that a romantic appeal for educational reform is likely to be less effective than a utilitarian appeal. Education is an enormous enterprise, which requires massive public investment in order to bring about social outcomes of great consequence. Under these circumstances, the investment is easier to justify in support of practical goals like promoting economic growth and providing needed job skills than in support of romantic goals like promoting pleasurable schooling and engaged learning. The administrators offered the schooling we needed, whereas the pedagogues offered the schooling we might have liked.
Why the Deeper Levels of the School System
Have Been Harder to Reform
The case of progressive reform in the first half of the 20th century suggests that, at least in the United States, educational reform had trouble bringing about fundamental change in the system of schooling. Effecting change at the rhetorical level is relatively easy for reformers who have access to the educational discourse community and who have a message that resonates with community members. Making change at the structural level is more difficult, but it too is possible for reformers who have credibility with and access to the administrators charged with developing curriculum and running school systems. But changing how teachers teach in their individual classrooms is much more challenging; and changing how and what students learn in those classrooms is the most daunting of all reform aims.
In this last section of the paper, I explore some of the factors that make reform so difficult at the core levels of the school system where teaching and learning take place. I argue that there are elements of teaching and learning that are intrinsically difficult to shape from the outside in any educational setting and that these elements are particularly resistant to change in the distinctive setting of American schools. Let me start with two elements of the school system that are particularly characteristic of the American setting: loose coupling and weak administrative control.
As Karl Weick (1976) and Charles Bidwell (1965) have spelled out, the organization of American schooling is a loosely coupled system. In a tightly coupled system, like a nuclear power plant or a petroleum processing facility, changes or actions in one part of the system quickly travel to other parts. But this is not the way things work in American schools, where the parts of the system operate as semi-autonomous segments rather than integrated components of a single entity (Perrow, 1999). Historically in the U.S., state systems of schooling have operated quite independently of each other, and they are only tangentially connected with the federal government. Likewise, school districts have had their own governance structure, funding sources, political constituencies, hiring authority, and organizational cultures, largely buffered from intrusions by state authorities and quite separate from other districts. Within districts, individual schools have a similar degree of independence from each other, as parallel but not closely interrelated self-contained segments of schooling, and they are also protected from much intrusion from the district administration by their physical isolation from the district office and by their ability to deliver schooling to a particular community more or less on their own. Within schools, individual classrooms act as separate instructional modules, which are independent of each other and which are cut off from the school principal by their distinctive function (instruction, in contrast with the principal’s focus on administration) and their physical location (behind the walls of the self-contained classroom). And within classrooms, individual students act as separate units of teaching and learning, each bringing distinctive abilities and motivation to the learning process and thus offering distinctive challenges to the teacher independent of the other students in the class. The relative independence of states from the federal government, districts from the state government, schools from the district, classrooms from the principal’s office, and students from the teacher provides functional semi-autonomy for both teachers and students in relation to all of the layers above them in the organizational structure of schooling.
Each of the six layers of the organization of schooling has a similar structure, with a series of parallel (and in many ways interchangeable) organizational segments (students, classrooms, schools, school districts, state systems), operating quasi-independently in delivering an educational service and nested loosely within a higher level unit (classrooms, schools, districts, states, and federal system) that has the same segmented structure. The basic building blocks of this nested structure are individual students, which accumulate in stages into classrooms, schools, districts, state systems, and finally a national system. Thus when we talk about the American system of schooling, we’re really referring to a series of nested segments that ultimately rest on the teaching and learning of 50 million students that goes on in approximately 3 million public school classrooms, which in turn are nested in 95,000 schools, 14,000 districts, and 50 states. (In 1938, the first year the U.S. starting collecting data on school districts, the numbers were about 25 million students, 900,000 classrooms, 250,000 schools, 120,000 school districts, and 48 states.) Under these conditions, it should not be surprising to find that school reform efforts are hard pressed to penetrate vertically all the way down through these layers in the system, and to spread horizontally across all the segments in each layer, to reach – at long last – the classroom and the student. Having significant impact on the district level, as the administrative progressives did, was in this view quite an accomplishment.
The standards movement, which arose at the end of the 20th century in the U.S., has been in part a major effort to learn from the limitations that this system imposed on reformers by deliberately seeking to create a more tightly coupled system of schooling. This started with federal rhetorical calls for a common set of curriculum standards (in the 1983 report, A Nation at Risk and updated in the 2001 law, No Child Left Behind); which led to the setting of state curriculum standards reinforced by high stakes tests; which brought local consequences for districts, schools, teachers, and students who failed to meet performance benchmarks; which in turn helped coerce school districts, schools, and classrooms to align both the formal curriculum at the district level and the curriculum-in-use by teachers at the classroom level with state and federal guidelines. Another element that has helped reinforce this move toward tighter coupling during the standards movement was the trend during the same period to shift the funding for schooling away from dependence on local property taxes (which tend to reinforce district autonomy) and toward the state budget.
Weak Administrative Control Over Instruction
Another historically distinctive characteristic of American schooling is the relatively weak control that school administrators exert over instruction (Lortie, 1969). In part, of course, this administrative weakness is a function of loose coupling, which makes it difficult for administrators to reach across the boundaries separating school districts from the state, schools from the district office, teachers from the principal’s office, and students from the teacher; and which makes it difficult for changes brought about in one school district to spread to other districts, from one school to others in the same district, from one classroom to others in the same school, or from one student to others in the same classroom. But there is an additional component that weakens the impact of administrators over teaching. The structure of teaching-as-work in the U.S. is such that school administrators have traditionally been lacking the basic levers of power that enable employers in most occupational settings to motivate employee compliance with their boss’s wishes. In most jobs, the employer can manipulate some combination of two core mechanisms to make sure that employees do as they are told. One is fear, the other greed.
Fear works at an elemental level: Do as I say or I’ll fire you. In its more sophisticated forms, it presents the employee with a variety of possible punishments short of termination for noncompliance with the employer’s expectations: demotion, reprimand, transfer to a less desirable posting, relegation to the margins of the work group, reduced perks, and barriers in the path to promotion. Principals in American schools, however, traditionally have held a weak hand in deploying the fear factor against teachers. Consider the current pattern in the U.S., which has been in place since teacher unions became strong after World War II. The contract with the teacher union makes it extraordinarily difficult to fire a teacher after she has completed an initial probationary period of three years or so. American teachers enjoy a form of tenure guarantee that, though somewhat weaker than the tenure rights of American college professors, is much stronger than in nearly any other occupational setting. To fire a teacher after being tenured is so onerous in its requirements for documentation, due process, and battles with the union that most principals don’t even try. So this leaves only lesser forms of punishment that are not as effective in getting the teacher’s attention. The principal can provide a negative evaluation, deny special requests, block access to resources, push a teacher to transfer to another school, or assign a teacher to less desirable classes. But in the absence of the big gun of termination, these forms of punishment are a weak weapon in the ongoing effort to get teachers to teach the mandated curriculum in the desired manner.
If fear provides the stick for the employer, greed provides the carrot. It is equally simple in its operation: Do as I say and I’ll reward you. Jobs in most complex organizations offer a finely tuned array of possible rewards for the employees who perform well in the eyes of their employers, but the two primary modes are pay and promotion. Pay increases and promotions are doled out or denied based on the degree to which the employee’s work meets management expectations. Pay, bonuses, and other forms of financial incentives offer an infinitely flexible mechanism for rewarding employee behavior. And promotions supply a similarly rich series of gradations for offering possible benefits for the compliant employee: a higher grade, a better title, a bigger office, a window, more generous benefits, or a company car. In American public schools, however, administrators since World War II have had virtually no discretion in allocating either pay or promotion. Pay levels are defined by union contracts, and traditionally they are based on only two criteria: the number of years the teacher has served and the number of graduate credits and degrees. Reformers have repeatedly proposed a shift toward merit pay for teachers, and a few districts and their unions have experimented with such plans. But with these few exceptions, school administrators in general have no ability to hand out pay increases to teachers based on their performance in meeting particular educational goals. Promotion is also largely unavailable to administrators. Teaching is a horizontal profession, lacking the stratified career path of lawyers in a law firm or professors in a university. Except for high schools, which have the position of department chair, there is no higher level teaching position to lure the aspirations of teachers. Newcomers to the profession take on the role of teacher in their first year on the job and hold that same role until they retire. Promotion normally comes only by leaving teaching for administration. Again, there have been proposals to create higher tiers of teaching (lead teacher, career professional, board certified teacher), but these have not yet become incorporated into the structure of the profession (Carnegie Task Force, 1986; Holmes Group, 1986).
That is the current pattern of teacher tenure, pay, and promotion in the U.S. But what were conditions like during the first half of the 20th century, when the progressive education movement was in full swing? Teacher unions started exerting a major influence in the largest American cities during the first two decades of the 20th century, affiliating with the American Federation of Teachers (founded in 1912). These efforts lost steam in the 1920s and then redoubled in the 30s and 40s (Urban, 1989). Formal teacher contracts had been the norm in public schools since the second half of the 19th century. But initially these contracts enumerated a variety of reasons that would warrant termination, although these primarily related to sexual behavior (consorting with a man without a chaperone, marrying) rather than instructional behavior. A teacher shortage in the early 1920s, however, compelled school districts around the country to offer teachers job tenure in order to attract and retain the teachers they needed (Sedlak, 1989, p. 272). In addition to tenure, teacher contracts by the 20s also began to change the terms of pay and promotion. The older contracts had set up a pay scale that offered higher salaries for men; they also offered higher pay for teachers in the upper grades, which created a de facto promotion system for teachers based seniority, encouraging teachers over time to move up the scale of school grade levels and teacher pay. But in the 20s, teacher shortages and union pressure forced districts to raise pay and also to yield to demands for a single standard pay scale, which equalized pay between men and women and across grade levels. As a result, the old promotion ladder quickly disappeared while the pattern of rewarding seniority remained. This left a structure of teacher tenure, pay by seniority, and nonpromotion, which approximated the current pattern in the U.S.
During most of the 20th century, therefore, school administrators in the U.S. largely lacked the ability to fire, promote, or regulate the pay of teachers, which left them with few of the resources that administrators in most other workplaces could use to keep employees working in line with the aims of management. This helps explain why it has been particularly difficult for reformers to reach past the second level of the American school system. Even when reformers have succeeded in reshaping the formal structure of the school district, as the administrative progressives did, they have still found themselves blocked at the classroom door. And it is this third level of the system, classroom teaching, that constitutes the instructional core of the entire system.
The Peculiar Nature of Teaching as a Practice
We have been examining factors that limit the ability of educational reform to reach the lower levels of the educational system; and up until this point, we have been focusing on factors that are particularly salient to the American case: loose coupling and weak administrative control over teaching. But there are some other factors that are more generally applicable to school systems in other parts of the world, and these arise from the distinctive characteristics of teaching as a professional practice. In exploring these, I draw on seminal work by Willard Waller (1932/1965), Dan Lortie (1975), David Cohen (1988), and Richard Elmore and Milbrey McLaughlin (1988).
The primary problem of teaching as a practice is that it won’t work unless the teacher is effective at establishing a special kind of personal relationship with the individual students in the class. Without this kind of relationship, students will not learn what schools want them to learn. And teachers can only establish these kinds of pedagogically effective relationships if they are allowed the discretionary space to do so. They need the latitude to figure out a way of doing things that works best for the individual students in the class and for the special situations of time and place.
Thus a central dilemma of school reform: If reformers are going to have an impact on the instructional core of schooling instead of limiting themselves to the rhetorical or structural levels of the system, they need to change how teachers teach in classrooms; but intruding on teachers in this way threatens to undermine the degree of teacher discretion that is necessary to foster effective learning (Elmore & McLaughlin, 1988). Teachers fit the occupational category that Michael Lipsky (1980) calls “street-level bureaucrats.” These are public service workers whose clients are nonvoluntary, who function under conditions of crushing demand and inadequate resources, where goals are ambiguous or conflicted, and where performance in relation to goals is hard to measure. In cases like these (police officers, social workers, teachers), the bureaucracy has no choice but to allow substantial discretion to the front-line agent to decide how to apply general policies to the myriad peculiarities of the cases at hand. From this perspective, then, school reform at the classroom level may not only be difficult, it may be counterproductive. And a key reason that teachers often resist reform efforts may be that they are trying to preserve a form of teaching and learning that seems to work and to fend off an alternative approach that might not. As Michael Fullan has noted, it is just as helpful to schooling to block a harmful reform as it is to implement a beneficial reform (Fullan,1991, p. 18). Let me briefly explain some of the key elements in this pedagogical relationship between teachers and students that create barriers to reform at the classroom level.
Complexity: One factor that promotes teacher autonomy is the sheer complexity of the task of teaching a class full of students. Even when students are sorted into grades by age and into classes by ability, they still bring an enormous range of salient characteristics to the learning process. They have differences in reading, writing, speaking, and language ability; in cognitive style and preferred mode of learning; in class, ethnicity, and gender; in psychological state and family relationships; in social, economic, and human capital; in prior learning of individual subject matters; and so on. Teachers have to adjust their instructional approach to these needs and capacities of individual children if they are going to optimize the prospects for learning. And in turn they need to adjust these instructional adaptations for individual students to a series of broader variables that affect the whole class, such as school culture and community norms; grade level and subject; time of year, time of day, and even weather. These factors are impossible to generalize across different classrooms and schools, and efforts to structure instruction around such generalizations – which is what school reform is all about – may reduce the effectiveness of teaching.
Client Cooperation: Another major factor reinforcing teacher autonomy is that teachers’ success as professionals is entirely dependent on the cooperation of their students. Teachers don’t succeed unless students learn, and students only learn if they choose to. Cohen (1988) notes that teaching is a “client-changing profession,” since teachers are charged with the task of transforming the skills, knowledge, values, and behaviors of students. The client’s active cooperation is a necessity for meeting these goals, whereas other professionals, like doctors and lawyers, often practice their craft on behalf of clients who are anesthetized or mute. Teachers, therefore have to focus on finding a way to motivate students to learn the curriculum; because, in the absence of this motivated compliance, teachers’ professional practice will be a failure.
Involuntary Client: This need to gain cooperation is made more difficult because the student is a conscript. Unlike most professionals, whose clients voluntarily seek their help, teachers are working with students who are there in the classroom because of coercion: from compulsory attendance laws, from parental pressure, from a labor market that requires a good education to get a good job. With an involuntary clientele, teachers have to struggle just to maintain control of a classroom filled with students who would rather be somewhere else, much less to use this control to facilitate learning with students who would rather do something else (Waller, 1932/1965). So in contrast with professionals for whom client cooperation is the norm, teachers have to manage the classroom in a way that motivates compliance with academic learning by clients who don’t want to be there or do that.
Motivating the Involuntary Client to Learn: The need to motivate the active cooperation of conscripted clients means that teachers cannot afford to establish the kind of professional distance from the client that doctors, lawyers, and accountants seek to maintain, since this kind of emotionally neutral relationship only works with clients who are passive and voluntary (Fenstermacher, 1990). The student must want to learn what the teacher is teaching, so the teacher’s first job is to find a way to instill this desire. One way to do so is to try to make the curriculum intrinsically interesting and engaging: Learn this because it’s fun. Another is to try to make it relevant and useful: Learn this because it will help you when you grow up. Yet another is to punish students for the failure to learn: Learn this or I’ll reprimand you, give you a bad grade, of send a note home to your parents. Yet there are significant limitations to all these approaches. Much of the curriculum is not intrinsically enjoyable; much of its usefulness is hard to discern, or so long deferred into the future that it is a poor motivator for the present; and punishment tends to negate student interest, at best promoting only the most minimal compliance. So teachers need a fourth motivator, which is to try to establish a close personal relationship with students and then use that to leverage learning: If you like me, then you’ll want to please me by learning what I teach. This kind of relationship is highly functional in motivating learning. It helps make learning more pleasurable, it enhances the teacher’s claims about the usefulness of learning, and it makes punishment less necessary and more effective.
Teacher Persona: The teacher’s imperative to establish a professionally functional personal relationship with students puts teaching in an occupational category that Arlie Hochschild calls “emotion management” work. In her book, The Managed Heart, she defines these kinds of jobs as face-to-face work roles that “require the worker to produce an emotional state in another person – gratitude or fear, for example” (Hochschild, 1983, p. 147). To do this well requires a form of acting.
We all do a certain amount of acting. But we may act in two ways. In the first way, we try to change how we outwardly appear. As it is for the people observed by Erving Goffman, the action is in the body language, the put-on sneer, the posed shrug, the controlled sigh. This is surface acting. The other way is deep acting. Here, display is a natural result of working on feeling; the actor does not try to seem happy or sad but rather expresses spontaneously, as the Russian director Constantin Stanislavski urged, a real feeling that has been self-induced. (Hochschild, 1983, p. 35)
Teachers need to engage in this kind of deep acting, by forming a teaching persona that is useful in inducing affection from the student and channeling this relationship into learning. To work well, this persona has to be an extension of the teacher’s self and must come across to the students as natural and authentic, even though it is in part constructed for pedagogical purposes.
Developing such a persona is a central task of becoming a teacher. Samuel Freedman shows how this was done by one New York high school English teacher named Jessica Siegel.
She wants to draw in the students, to thrill them a little. The bulletin board is part of the strategy, and so is her penchant for bright, funky attire. Today she wears four earrings and five rings, two silver on her left hand and three gold on her right, and a dress from Pakistan, bone-white cotton printed with blue designs that are as cryptic and angular as cuneiform…. A student once asked, “Miss Siegel, do you water that dress?”
Even as Jessica tries to captivate her students, she wants to control them – not to dictate or deaden, but not to abdicate authority either…. It took her years to develop a classroom presence that felt organic, for she was naturally a listener, a backbencher, a person who began countless sentences, “I don’t know a lot about this, but….”
Gradually, she created from pieces of herself a persona that might best be called The Tough Cookie. She stands this morning with right hand on the hip, head cocked slightly, eyebrows arched in mock disbelief; every so often, she shoots a phrase Jersey City style out of a gully at the corner of her mouth. “Gimme a break,” she says to a lying latecomer. Her students will hear her say the same thing a hundred times before the term is over, hear her bite down hard on “Gimme” and stretch “break” into an aria of annoyance. (Freedman, 1990, pp. 29-30)
The teacher persona – an engaging combination of the likeable and the tough – is not something that a teacher puts on lightly or sheds with ease. Developing this persona is hard work, as it was for Jessica Siegel, carried out in isolation from professional colleagues behind the walls of the self-contained classroom, where the teacher is thrown on her own resources to deal with an unruly crowd of unwilling students. It is a form of method acting that lasts not merely for the duration of a play but for the course of an entire career. And it is not just a way of practicing a profession but a way of being. As a result, we should not be surprised to find that teachers in general tend to resist efforts by reformers to change the way they teach. To do so is to change who they are.
Once a teacher has succeeded in developing a teacher persona and has begun using it effectively to motivate learning among the conscripts in the classroom, she is unlikely to give it up just because a new wave of reform rhetoric says she should and because school administrators issue a revised curriculum. In a study of efforts to transform the way teachers teach math in the U.S. in the 1990s, Stephen Mattson (2004) concludes that changing pedagogy is like changing religious beliefs; it calls for a tumultuous personal transformation that is akin to a conversion experience. This helps explain why such transformations as the result of reform efforts are not very common, and why the changes that teachers do make in their pedagogy in the wake of reform movements are often formalistic and incremental. Recall what Larry Cuban shows in his analysis of the marginal impact of pedagogical progressivism on American teaching in the 20th century – that teachers were more likely to adopt a few of the movement’s forms (like gathering students in groups or letting them move around the classroom) than to adopt the substance of child-centered instruction; more likely to add a few elements to their core of continuing practice than to transform the way they teach.
One way to untangle fruitless debates about whether a particular educational reform was a success or a failure is to examine how deeply that reform managed to reach down through the levels of the educational system to its instructional core. I argue that reformers have had the most success in shaping the system at the rhetorical level, by changing the way we talk about school. And with the right people and the right proposals they have even been able to remake elements of the system at the level of formal structure, by changing curriculum guidelines and school district administration. But reformers have had a much more difficult time effecting significant change at the levels of classroom teaching and student learning.
The movement for progressive education in the early 20th century U.S. has provided an instructive case in point. One strand of the movement, the administrative progressives, succeeding in changing rhetoric and reforming the structure of school districts, while the other strand, the pedagogical progressives, exerted an impact that was limited to the rhetorical level. But both strands of the movement were largely unable to change classroom teaching and student learning. Progressivism was the dominant reform movement in American education during this period, and it drew support from an extraordinary array of educational administrators, professors, theorists, and policymakers; yet its impact on the core of schooling was marginal.
In part this relative failure can be attributed to characteristics that are specific to the American system of schooling: a loosely coupled structure of organization, which buffered schools and classrooms from changes coming down from above; and a structure of teaching that denied school administrators control over teacher tenure, pay, and promotion, thus protecting teachers from administrative efforts to change the way they taught. But in part the difficulty that reformers faced in transforming the instructional core of the school system arose from characteristics of teaching practice that extend well beyond the borders of the United States. Teachers in general face a common problem of practice. Their professional success depends on their ability to motivate an involuntary group of students to learn what the teacher is teaching. In an effort to accomplish this, teachers invest heavily in developing a teaching persona that enables them to establish a relationship with students and lure them to learn. Once they have worked out a personal approach for managing the instruction of students within the walls of their classroom, they are likely to resist vigorously any effort by reformers or administrators or any other intruders to transform their approach to teaching. Teacher resistance to fundamental instructional reform is grounded in a deep personal investment in the way they teach, a sense that tinkering with this approach could threaten their very ability to manage a class (much less teach a particular curriculum effectively), and a realization that changing how they teach is akin to changing who they are.
So, as Cohen (1988) and Lortie (1975) and Britzman (1986) point out, when it comes to matters of practice, teaching is an extraordinarily conservative profession. Pedagogical change offers the teacher little apparent benefit and great apparent risk. But this aversion to change is not simply a matter of self preservation or personal preference; it is also potentially a matter of pedagogical effectiveness. If they are going to succeed in promoting learning among their students, teachers may well need the discretion they so zealously protect. As the street level bureaucrats of the educational system, they may need the space to work out their own arrangements for handling the complex learning environment of the classroom, where general rules simply don’t apply, even if they come from reformers with the highest educational aims.