The Politics of Education Reform

The recently released Third International Mathematics and Science Study (TIMSS), which made international comparisons of math and science performance among fourth- and eighth-grade students, strengthened the case of those who are calling for ambitious reform of U.S. education. U.S. fourth graders did relatively well in science and about average in math; eighth graders did slightly better than average in science and slightly below average in math. These findings are consistent with other assessments of U.S. student performance.

The TIMSS study also provided new and valuable information about the relationship between instructional practice and student performance. The message to U.S. educators was clear: science and math education needs to be better focused and more rigorous. Although one can still hear arguments that international comparisons are not fair, that the diversity of the U.S. population or the pluralistic nature of its political culture makes it impossible to replicate the coherence found in other countries’ schools, or that U.S. schools are already improving at an acceptable pace, the reality is that the majority of the public, of elected officials, and of educators believe that change is needed. The task is to determine what changes are necessary to make a real difference to students and how reform can be achieved in the U.S. political culture.

The evolution of U.S. education reform

U.S. elementary and secondary education is a vast and extraordinarily complex enterprise that seems to defy simple generalizations. However, the two central imperatives of U.S. educational governance are dispersed control and political pluralism. I have chosen my words carefully here. I use “dispersed” control rather than the more conventional “decentralized” control because I do not think that control of education is actually decentralized in the United States. The notion of local control of schools is, I think. largely inaccurate and outmoded, especially in light of the direction education reform has taken in the past decade. The idea of political pluralism is more straightforward. It captures a fundamental principle of U.S. politics-that political decisions and actions are the result of competing groups with different resources and capacities vying for influence in a constitutional system that encourages conflict as an antidote to the concentration of power.

The story of U.S. education reform since the early 1980s is worthy of either a Gilbert and Sullivan operetta or theater of the absurd, depending on your tastes. In 1983, the National Commission on Excellence in Education releases A Nation at Risk, focusing public attention on a “crisis” of low expectations, mediocre instructional practice, and menacing foreign competition; thereby legitimating a nascent education reform movement that has already begun in a handful of states. From the beginning, it is fairly clear that there is little the federal government can actually do to fix this crisis, because the ideological climate is running against a strong federal role. By the mid-1980s, with many states gearing up to take on the issue, the National Governors Association, under the leadership of a politically ambitious Governor Clinton of Arkansas, promotes the idea of a “horse trade”-greater flexibility and less regulation for schools and school systems in return for more tangible evidence of results, reckoned mostly in terms of student achievement. This is followed by another spate of state and local reforms aimed at deregulation, government restructuring, and tighter state monitoring of student achievement.

In 1989, an extraordinary event occurs: President Bush and 50 governors meet in Charlottesville, Virginia, to draft national goals for education. This Education Summit inaugurates an all-too-brief period in which there appears to be broad bipartisan support for some sort of national movement to support explicit state and local goals and standards. This consensus results in Goals 2000, a Clinton administration initiative with striking similarities to a prior Bush administration proposal. There then ensues a complicated and largely unsuccessful attempt to translate the apparent national consensus on goal-setting into an institutional apparatus that puts the federal government in the role of enabling state and local action. Beginning in the Bush administration, the federal government also gets into the business of lending financial and political support to professional associations to draft national content standards in subject matter areas.

Then “whammo,” with the congressional election of 1994, a seeming ideological reversal occurs on anything vaguely resembling federal action on goals and standards, followed by an unraveling of the earlier bipartisan consensus, some ungraceful wrangling over the funding and implementation of Goals 2000, and some stunningly adept pirouettes by right-leaning previous advocates of standards who overnight become critics of standards and advocates of local control. During this period, the education profession gets an introduction to bloody-nose politics. The carefully crafted history standards are shot down in debate in the Senate and their drafters are sent back to try again. The drafters of the English/language arts standards are held up to public ridicule for their inclusion of deconstruction theory. As if to show the final absurdity of the standards debate, Governor Pete Wilson of California vetoes funding for the state’s ambitious new student assessment system after a blistering debate about its content (too multicultural) and its feasibility (too little data on how individual students are doing). From 1992 onward, the standards movement has been declared officially dead at least once a week.

The national debate on educational standards has not been pretty to watch, but it has embodied a faithful enactment of the principles of dispersed control and political pluralism. The temporary bipartisan consensus on goals and standards after the Charlottesville summit concealed deep suspicion of anything national or federal in matters of curriculum and student learning. It did, however, demonstrate that a coalition of national decisionmakers could, however temporarily, presume to make authoritative judgments about the purposes of schooling. Control of education, it turns out, is only local when schools and school systems appear to be doing the right thing; when they’re not, they are fair game for elected officials, at whatever level of government, with a political interest in their performance. Likewise, the unseemly tussles over the California assessment system and the history and English/language arts standards demonstrate that issues of professional practice are vulnerable to the most basic form of pluralist politics-groups mobilize against proposals they regard as invidious to their interests, without regard for the professional or political authority those proposals carry.

Moving toward standards

What’s most interesting is not that standards inevitably provoked partisan and pluralist debate but that despite this debate, professional organizations, states, and localities continue to plod ahead with the development of standards in a tremendously varied way that fits remarkably well with the principle of dispersed control. Many states and localities are developing and implementing content and student performance standards despite, and often in response to, partisan criticisms. In addition, as standards have become a more prominent part of political discourse in states and localities, the range of actors involved in their development and revision has expanded to include many groups that were not involved in their early formation. States have also been engaged in a broad effort to develop and implement statewide testing programs (many of which antedate the current standards debate) that deliver, with increasing precision, data on student performance in individual schools. Meanwhile, NAEP has become more and more visible in its periodic statements of what U.S. students know about core academic subjects. NAEP now provides state-to-state comparisons, which would have been unimaginable 20 years ago. Finally, despite the highly visible partisan debate over the history and English/language arts standards, other efforts to develop content and performance standards have been much more successful. The math standards developed by the National Council of Teachers of Mathematics and the science standards developed by the National Academy of Sciences are viewed with increasing respect by educators and politicians. The New Standards project, a private nonprofit organization financed by private philanthropies, has recently released a comprehensive set of content and performance standards in mathematics, English/language arts, and applied learning that are increasingly seen as benchmarks for state and local standards-development activities.

Some fundamental changes have occurred in education policymaking at state and local levels over the past decade or so. A decade ago, only a few states and a relatively small proportion of localities collected and reported data on student test performance at the school level. Now, virtually all states and localities have the capacity to collect and report school-level student performance data, and in most states these data are now reported publicly once a year. Thus state and local policymakers and the public at large now have routine access to some sort of data on how individual schools are performing.

A decade ago, most states did not have formal policies that set expectations for measured student performance nor did they have policies that dealt directly with the content of academic instruction. Now, more than half the states are in the process of developing explicit policies about acceptable student performance levels on statewide tests as well as curriculum guidance about what should be taught.

A decade ago, most states viewed their role as setting broad, minimum, largely procedural requirements for local districts to follow in delivering education and dispersing state revenue to local school districts. No state, as far as I can tell, intervened directly in the affairs of individual schools, except in extraordinary cases of incompetence, or challenged the authority of local school districts to serve the students in their communities. Now, many states have adopted a much different posture. Some states have instituted inter-district choice programs that allow students to move across district boundaries, often taking state money with them. Many states have the authority to declare schools or entire districts deficient and assume temporary control over them. And many states directly authorize the creation of publicly supported “charter schools” operating outside the ambit of many state and local regulations.

A decade ago, it was virtually impossible to compare states in terms of useful measures of student performance. Now, with the development of state-level results by NAEP and the disaggregation of international student performance data to the state level, it is possible not only to compare states against each other but against other countries. As one might expect, these comparisons are met with much criticism and gnashing of teeth regarding how states differ in their student populations, but once the data are available, it is impossible to prevent comparisons.

A decade ago, a teacher, principal, or district curriculum specialist looking for the cutting edge of curriculum and instructional practice in a given content area would probably have consulted a teachers’ magazine or the curriculum collection in the neighboring education school library. In a few rare instances, such as the network of practitioners that formed around high-school advanced placement courses, teachers would be exposed to a fully developed curriculum and a group of colleagues trying to learn how to teach it. Now, many educational practitioners are exposed to a virtual blizzard of leading-edge advice on curriculum and pedagogy, such as national content standards sponsored by professional associations, state curriculum frameworks, and staff development consultants purveying what they consider to be the latest ideas about instruction. It is true that the penetration of these ideas and materials into the classroom is often superficial, that most schools probably still exist as isolated islands of practice, and that most curriculum and staff development materials that are available to most teachers are still of a decidedly mediocre sort. But the important shift from a decade ago is the current existence of a relatively well-organized, extensive professional community, with strong incentives for self-promotion, producing explicit instructional guidance on the leading edge of practice. What’s most remarkable about this growing industry is that it is setting standards of practice that are calculatedly beyond what the average teacher can do, calibrated instead to what students ought to learn.

The principle of dispersed control leads me to predict that states will continue to push toward state-to-school accountability measures until they can muster evidence on student performance that allows them to make a persuasive argument that they are discharging their political and fiscal responsibilities. States and localities vary widely in their capacities and in their political incentives to engage in standard-setting, and therefore the result of this dispersed activity will, at least in the short term, be a high degree of variability in standards from one place to another and (ironically) less standardization of policy and practice from a national perspective. Local districts and the federal government will increasingly become spectators in this state-to-school struggle unless they can find some way to participate in it productively.

The principle of political pluralism leads me to predict that political debate about the content of standards will probably continue, especially in highly contentious areas such as history and literacy, because content is such an attractive target for organized interests. But this debate will increasingly become a sideshow in the larger standards game. Schools, as they are subjected to increasing pressure for accountability, will reach for content and performance standards in order to simplify their task and reduce uncertainty and will find ways to submerge and deflect debate over the content of standards so they can get on with the task of satisfying state and local accountability pressures. The principle of political pluralism also leads me to predict that professional communities and commercial and nonprofit enterprises will become increasingly prominent in supplying advice on curriculum and pedagogy in response to pressures on schools for increased accountability for student performance, further fueling the press for standards.

Notice that there is no necessary coherence at the national or state level in this scenario, at least in the short term. It doesn’t even suppose that there will be coherent goals, standards, and instructional guidance from states to schools, although it will be extremely difficult for states to maintain pressure over the long run on schools if they can’t provide some sort of coherence in their expectations. It does suggest, however, that schools will be subjected to constant pressure for the foreseeable future to focus on demonstrable student learning and to seek external guidance from states, professional communities, and commercial enterprises about how to solve the difficult problems of what to teach and how to teach it. Without countervailing forces from the national (notice I didn’t say federal) level, variable capacities and incentives at the state and local level will probably produce more variability in policy and practice and could produce more variability in student performance.

The TIMSS findings and reform

The authors of TIMSS are careful to point out that their findings largely antedate most recent activity in states and localities concerning standards-based reform. Hence the findings are, in effect, baseline data on the state of instructional practice and student performance in math and science. The picture they give is, I think, exactly what one would expect from a system of dispersed control and political pluralism running on autopilot. In the absence of explicit external standards for content and student performance, teachers give great weight to the way content is portrayed in textbooks, which is the “default mode” for instructional guidance. Commercial publishers have little or no incentive to focus content; their incentives are to produce materials that are marketable to the broadest possible cross-section of customers and to gear content to the largely content-free nature of existing standardized tests. Administrators at both the school and system levels have little or no political incentive to engage in explicit instructional guidance for teachers; their main job is to orchestrate, deflect, and buffer the multiplicity of organized interests that try to influence schools.

Hence the knowledge that is enacted in curriculum and pedagogy becomes a byproduct of the political incentives that operate on teachers-discrete bits of information, emphasis on coverage rather than depth, diffuse and hard-to-understand expectations for student learning, little convergence between the hard day-to-day decisions about what to teach and the largely content-free tests used to assess student performance, and a view of pedagogy as a function of the personal tastes and aptitudes of teachers rather than as a function of external professional norms. Students who do well in such a system recognize that they are being judged largely on their command of the rules of the game, which reward aptitude rather than sustained effort in the pursuit of clear expectations. All systems have a code; the job of the student is to break it. Some do, some don’t.

It seems plausible that if one were to impose content and student performance standards, we well as assessments geared to those standards, on this system, school and system-level administrators would focus more on instructional guidance, the variability of instructional practice among teachers would decline, students would receive clearer expectations about what they are supposed to learn, and presto! student performance would both improve and become less variable. This is the underlying theory of standards-based reform. There are two major problems with it. One is that standards-based reform doesn’t displace or override the principles of dispersed control and political pluralism; it blends with them in ways that we understand only imperfectly. The result of this blending might well be, at least in the short term, increased variability in both instructional practice and student performance as states, localities, and schools struggle to adapt to standards under conditions of variable capacity and political incentives.

The second problem with the theory is that instructional change is difficult, demanding, and unfamiliar work for teachers, students, and administrators. Standards reform requires fundamental changes in the way education is practiced and governed. It requires teachers, students, and administrators (not to mention parents) to accept explicit external standards for what constitutes acceptable content and performance. It requires teachers not just to teach to these standards but to learn how to teach in ways most of them have never done before. It requires administrators to redesign their jobs and their organizations to focus on continuous improvement of instruction in classrooms and schools rather than primarily on managing the political environment of schools. It requires teachers and administrators to make hard judgments about whether their colleagues are meeting performance expectations and whether, if they fail to do so, they should be given additional assistance or encouraged to find work elsewhere. It requires teachers and administrators to deal with the inevitable frustration and anger that will come from parents and the public when students and schools are found not to be meeting the requirements of external standards. It requires school governance authorities such as school boards and state legislatures to maintain a commitment to standards over time and to allocate the resources and authority to teachers and administrators that are necessary to sustain this commitment in the face of the inevitable dissent and conflict that will accompany explicit judgments about the performance of students and schools. And it requires education practitioners and school governance authorities to manage and adapt performance standards on the basis of hard evidence about whether they are working to promote student learning. This is just the simplest list of what standards-based reform entails.

Connecting policy and practice

In their profound analysis of the history of U.S. education reform, David Tyack and Larry Cuban note the persistent gap between what they call “policy talk” and the world of daily decisions about what to teach, how to teach, and how to organize schools. Most reforms, they argue, exist mainly in the realm of policy talk-visionary and authoritative statements about how schools should be different, carried on among experts, policymakers, professional reformers, and policy entrepreneurs, usually involving harsh judgments about students, teachers, and school administrators. Policy talk is influential in shaping public perceptions of the quality of schooling and what should be done about it. But policy talk hardly ever influences the deep-seated and enduring structures and practices of schooling, which I have called the “instructional core” of school.

The Tyack and Cuban analysis, I think, accurately captures the way Americans have historically dealt with education reform. The debate on standards-based reform, though, opens up the possibility of dealing with reform in a way that establishes a more direct connection between policy and practice. Never before in U.S. history has there been such a broadly based conversation about these matters. But although the conversation is about important issues related to the core of schooling, it is still largely policy talk because it has yet to address at least two important questions related to sustained improvement of the instructional core: What new knowledge and skills do educational practitioners need to teach to ambitious standards of student learning? And what incentives do practitioners have to engage in the hard work of acquiring and using this knowledge and skill?

“What do I teach on Monday morning?” is the persistent question confronting teachers. Because they are inclined to ask such questions, teachers are often accused by researchers, reformers, and policymakers of being narrow and overly practical in their responses to the big ideas of education reform. Given the state of the current debate on standards-based reform, though, I think the Monday morning question is exactly the right one, and it should be firmly placed in the minds of everyone who purports to engage in that reform.

Consider the following practical issues. Most statements of content and performance standards coming from professionals and policymakers take no account whatsoever of such basic facts as the amount of time teachers and students have in which to cover content. They are merely complex wish lists. In order to be useful in answering the Monday morning question, they have to be drastically pared, simplified, and made operational in the form of lesson plans, materials, and practical ideas about teaching practice. Furthermore, most standards fail to take account of the drastic differences among schools in the type and level of schooling, student populations, resource levels, and the makeup of the teaching force. The problems of implementing standards are also vastly different at the elementary and secondary levels. Elementary teachers usually teach everything. Secondary teachers tend to specialize in particular content areas. Using standards to inform instructional practice at the elementary level can seem impossibly complex to most teachers. I have often visualized this problem by imagining a pile of standards documents, cutting across math, science, English/language arts, and social studies, on the desk of the typical elementary-school student, with the teacher trying to figure out how to reduce the pile to a manageable size and engage the student in some productive work. Most sane people would not stand for this task, yet we expect teachers to willingly engage in something like it. The problem of secondary-school teachers is somewhat different. Although the pile of standards is still impossibly large for a given teacher and collection of students in a given classroom, teachers face the additional problem that we expect them to override years of experience in teaching with a collection of external prescriptions about how they ought to teach. Most sane people would not stand for this task either. They would expect, at the very least, to have an intelligent conversation with someone about why these prescriptions should be useful to them and how they should get from where they are in their current understanding of content to where the standards say they should be.

On this subject, it seems to me that the TIMSS findings carry an important and powerful message: the need for parsimony. The findings paint a picture of scattered and largely shallow coverage of content. Getting from standards as they currently exist to practice focused on a deep understanding of key ideas requires reducing external prescriptions to the minimum possible level and focusing them on the most important aspects of instruction. This idea sounds appealing when stated as an abstract principle, but someone has to engage in the hard and controversial job of throwing out vast amounts of repetitive, overly prescriptive, and distracting bits of content; and someone has to deal with the problem of political pluralism whereby anyone with an idea, no matter how half-baked, about what should be taught can organize a political movement to get it written into the official curriculum. The hard reality is that U.S. schools have no processes in place for making these difficult political judgments, much less making them binding. So getting standards pared down to a manageable level of complexity requires the development of a new way of making curriculum decisions in states and localities-one that holds feasibility in high regard while respecting the enormous pressures of political pluralism. Imagine a world in which state legislators, school board members, and local superintendents play an active role in making the hard judgments that enable teachers to answer the Monday morning question. This is much different from the one in which we presently live.

Suppose we were somehow able to answer the Monday morning question and create a system of standards that values parsimony over undisciplined pluralism. Here the brute facts are even more daunting. The work day of most teachers allows them virtually no time to engage in any sustained learning about how to do their work differently. Their time is fully scheduled during the school day, with the exception of a few brief and scattered preparation periods. The time available in the summer is time when students are typically not in school, so the learning that occurs then is largely done in isolation from actual practice in classrooms. Organized professional development, if it occurs at all, takes place in most local school systems during a few days scattered throughout the school year or in after-school sessions, again in isolation from actual practice. Most professionals learn new practices by working with other professionals in close proximity to the details of practice and by making their clients pay for the surplus time required to retool and renew themselves. However, we expect teachers not to learn new practices as part of their daily work life and to sandwich time for learning into spaces in the day and year when students are not there. Every minute of time for professional learning that comes at public expense is begrudgingly granted.

Time is money in the educational enterprise, as in all others. Creating more time in the space of the school day for teacher learning might mean hiring more people to cover classes when teachers are engaged in learning. Organizing professional development experiences around actual instruction in the classroom with real students means hiring people to consult with teachers or freeing up other teachers to work with their colleagues. Creating manageable instructional materials for teachers to use and adapt in their classrooms requires time on someone’s part to sift through the multiple competing packages to find the best and most appropriate. Some of this time and money can be extracted from the organizational slack that exists in most school systems. But some of it will have to come from additional resources, wisely invested. Who is going to make the difficult adjustments necessary to pry loose existing resources and find new ones? Again, this looks like a job for state legislators, school boards, and local administrators, but it is one for which they are currently ill-prepared.

Now imagine that we somehow solve both the parsimony problem and the problem of how to organize learning for teachers. We are left with the still more daunting problem of how to adapt general guidance for instruction and learning of new practices to the realities of diversity among schools. Would we expect instructional improvement to look the same in a school that serves students from several language groups, many of whom come from homes where there are no computers and few books, as in a school where all the students are fluent in English and come from homes that have computers and are packed with books? Would we expect the same kinds of materials to work for teachers faced with children struggling to understand what school is about as for teachers in schools where students come steeped in knowledge about why they are in school? Would we expect the same instructional content and pedagogy to be successful in a classroom where all but a few students are reading at the same grade level as in a classroom in which the range of reading ability spans several grade levels? To answer the Monday morning question credibly in all schools and classrooms, we have to find a way to make standards responsive to such variations. Again, this sounds like a job requiring a much different kind of leadership from policymakers and administrators. But it also requires the sustained engagement of teachers in understanding how to adapt general prescriptions to the specific conditions of their classrooms, without shortchanging students who are being asked to compete with other students who are not like them.

The knowledge and skill problems presented by standards-based reform are deep and difficult. They cannot be solved by engaging in more sophisticated policy talk, but by sustained engagement between policymakers and practitioners in difficult discussions about resources, expectations, and the realities of diversity in schools and classrooms. If either side of this discussion pulls away from the other, standards-based reforms will go the way of earlier reforms: They will exist largely in the realm of policy talk and, more ominously, will devolve into blaming teachers and students for the failures of everyone.

The incentive problem is equally difficult. The work described above is not only hard and demanding, it is different from the work we have asked teachers and administrators to do in the past. These new definitions of educators’ work may seem self-evident to critics and reformers, but they are far from self-evident to people who work in schools. Educators, like everyone else, do what they are rewarded and reinforced for in their daily life. If we want educators to do their work differently, we have to reward them for doing the right things. Acknowledging that we know almost nothing about how to do this is the first and most important step in understanding how to do it well.

The existing array of standards-based reforms now in place in states and localities contains the first attempts to solve this incentive problem. The idea that schools should be evaluated and rewarded on the basis of student performance is now creeping into policy talk, but it has yet to work its way into educational practice. One doesn’t have to think very hard to understand the destructiveness of this idea. Imagine a world in which, overnight, all schools were rewarded financially, and were ultimately permitted to live or die, on the basis of some measure of gains in student performance based on clear standards. The clearer the standards and the more direct the rewards and sanctions, the worse the consequences. Race, social class, and home environment are the strongest predictors of educational performance for students. Rewarding and punishing schools on the basis of their performance under these circumstances means, in effect, rewarding and punishing them for the students they serve. Worse yet, adjusting rewards and punishments for student background probably means that certain schools will be allowed to continue to have lower expectations for their students than other schools, thus defeating the main purpose of standards-based reform, which is to promote high-quality learning for all students. Under these circumstances, the more pressure we apply in the form of external standards of student performance, the more variability we are likely to create in the very areas where we are trying to reduce it. Still more troubling, the more pressure we apply, the more we are encouraging schools to recruit “good” students and push away “poor” students, and the more we are encouraging schools to blame students and their families for the schools’ failures. If this sounds like a dismal prospect, I want it to. It is a horrendously difficult problem and it needs to be faced squarely by everyone who supports standards-based reform.

I propose a new principle of standards-based reform, which I call “reciprocity of capacity and accountability.” The principle goes something like this: Every increase in pressure on schools for accountability for student performance should be accompanied by an equal investment in increasing the knowledge and skills of teachers, administrators, students, and their families for learning about how to meet these new expectations. In its simplest form, this principle means that no school is judged to be failing until policymakers are satisfied that investments in learning new ways to teach, new ways to manage instructional improvement, and new ways of understanding student and family responsibilities have been implemented and paid for. In its more complex form, this principle means that everyone who occupies a position of formal authority in the educational system should judge their actions against the criterion of value added to instructional improvement.

If schools are to be held accountable for student learning, then the people who run them should be judged by the extent to which they add value to the quality of classroom instruction. So the first diagnosis of school failure should not be directed at teachers and students but at the way policymakers and administrators have organized resources to promote new knowledge and skills in schools. For example, a failing school in which teachers have not had sustained and effective professional development, organized in a way that is directly connected to standards for student performance, is not a failing school. It is a school managed by failing policymakers and administrators. In a system governed by the principle of reciprocity of capacity and accountability, everyone would rewrite their job description in terms of the value they add to the improvement of classroom instructional practice. If you can’t rewrite your job description in this way, look for work elsewhere.

The answer to the question of how we give educators the incentives to do the kind of work required by standards-based reform is that we provide them with the necessary knowledge and skills and we reward them for participating in such activities, in tandem with the implementation of external accountability measures that are designed to reward and penalize schools for student performance. No external accountability measure should be implemented without a specific investment in knowledge and skill designed to improve the capacity of educators to meet that measure. Furthermore, when external accountability measures are found to reward the wrong things, such as rewarding schools for shifting students around rather than educating the students they have, then the measures should be changed. I cannot stress enough that we know little or nothing right now about how to engage in this delicate balancing of capacity and accountability. Until we learn how to do it better, we should be modest in our demands on schools for external accountability measures and ambitious in our attempts to solve the capacity problem.