Should the United States have national standards for our education system? Americans have been debating this question for the last quarter of a century. If you are trying to make up your own mind and think the decision should be simple, you're likely to be surprised.
In a system in which localities and states pay about 93 percent of the cost of schooling and the federal government has no constitutional role in education, the nation has flirted with the idea of injecting broader national control into education policy. The United States has seen historic episodes of federal or national initiatives in education. These episodes include efforts to bolster math and science capability in response to Sputnik, the quest for racial and ethnic equity through the courts, efforts to equalize resources in public schools, and an outright seizure of control in crucial areas of policy and practice through No Child Left Behind (NCLB).
In parallel efforts since the mid-1980s, policymakers have made sporadic attempts to raise achievement levels by creating both a national definition of what students should be taught and a national test to see whether schools were successfully bolstering achievement. This effort was touched off by the 1983 report of the National Commission on Educational Excellence, which found the nation "at risk," although its corrective recommendations were directed at local and state governments.
By 1989, concern about the U.S. education system's worldwide status had risen so high that President George H.W. Bush gathered the nation's governors for a summit to establish goals for the United States to achieve by 2000. Other calls for national intervention included the National Council on Education Standards and Tests, the National Education Standards and Assessment Council, the Clinton administration's Goals 2000 legislation, and that administration's partial development of a voluntary national test, which Congress soon abandoned.
This condensed history conveys the continuing desire for action at the national level, the failure to bring that desire to fruition, and a lack of agreement and enthusiasm for this level of federal intervention. America has clearly been of two minds about bringing change to our locally based public education system.
One reason we lack agreement is that people place different degrees of value on schools' traditional local control and diversity, show different degrees of willingness to take the risks involved, and come to different conclusions about whether accompanying challenges can be overcome. Perhaps most basic, people have different mind-sets about what "national standards" means.
What Are "National Standards"?
When people say they are for or against national standards, they often harbor quite different views of what they want to create—or protest. Some people have in mind setting standards for the content of what is taught in any grade or subject in school. For some, establishing content standards means increasing the rigor of instruction; for others, it means "standardizing" what knowledge is taught. These two goals are related but different. An extreme example of the latter goal would be the uniformity in France, where at almost any hour, all students are studying the same thing.
Some supporters of standards envision a single national test that all students would be required to take. Others want both prescribed course content and a national test. There are those who claim that standards should be voluntary and those who believe that we should subject all students to them. Goals vary: Some want standards to give teachers information that will improve instruction, whereas others want a national test to measure performance in an NCLB-like accountability system.
Differences also exist on the "how." Some people clearly want standards to be national but to be created and enforced by some agency other than the federal government (although possibly with government funding); others want the federal government to develop and prescribe standards. In another version, both standards and tests would emanate from outside the federal government, but they would be used to enforce a federal accountability system—a scenario that I believe would effectively federalize such a test.
Before we can fruitfully discuss issues involved in setting national standards, everyone will have to agree on what we are after. Although the options that individuals or groups might propose would still entail differing challenges, issues, and values, all parties would at least then be on the same page.
Surprising Variation
One complication to developing standards is the extensive variation in our education landscape. A 2009 report I wrote for the Educational Testing Service reveals huge variations in content of instruction, states' performance standards, and student achievement levels across the nation. We could look at this high rate of variation in two ways: as a sign that the obstacles to standardization are insurmountable or as evidence for why a single set of standards is needed.
The amount and degree of variation throughout our education system may surprise many readers. Let's consider three major areas of variation: (1) the content of instruction from place to place, (2) the performance standards states have set, and (3) student achievement.
Variation in Content
Throughout U.S. history, certain forces have worked to produce uniformity in curriculum, and others have pushed toward increasing variation. Historian Daniel Boorstein has pointed out that early U.S. schoolmarms taught a standard English; therefore, unlike British citizens, Americans could understand one another wherever they went in the country. McGuffey readers were widespread in schools throughout the 19th century, and standardized tests have long been in use.
One would expect textbooks to exert pressure toward uniformity. Schools in all 50 states generally choose from only a handful of textbooks for any particular subject or grade. However, there is great variation in the levels of student achievement in any grade across the country, and there are large differences among district and state prescriptions of what should be taught. So if textbook publishers are to market themselves well, they must span a wide diversity of instructional objectives. A detailed review of prescribed content for 4th grade instruction in 10 U.S. states, for instance, revealed 108 possible learning outcomes, only four of which were common to all 10 states (Reys, as cited in Beatty, 2008). Publishers are pushed in the direction of covering a vast number of objectives. If we hope to set common standards, we must address the forces that produce this tremendous variability.
Variation in Performance Standards
The U.S. Secretary of Education recently said that we don't need 50 different goal-posts, a reference to NCLB's provision that each state can set its own performance standards. So how much variation is there? A lot. Each test's performance standard is simply a cut point on the state test that represents the level at which a test taker is deemed proficient. The stringency of standards is difficult to compare because each state has its own tests.
However, the cut points for each state test have been "mapped" onto the national score scale for the National Assessment of Educational Progress (NAEP), and the Department of Education has published tables showing where each state's cut point falls on that scale. The percentage of students meeting or exceeding the score that NAEP has set as representing proficiency ranges from 18 percent in Hawaii and Mississippi to 38 percent in New Jersey (Barton, 2009, p. 18).
States differ greatly, of course, in average income, the richness of students' experiences before they start school, and the size of the state's tax base. Do these differences account for some of the variation in how high they set their requirements for student achievement? One way to examine this would be to map the cut points on each state's test onto that state's own NAEP scale rather than the national NAEP scale to see how states compare in reaching their own cut points for their student populations. I did this mapping and found that the disparity on this basis was greater, not less: The percentage of students reaching or exceeding the cut point ranges from 30 percent in South Carolina to 88 percent in North Carolina—two adjoining states with identical average NAEP scores (see Barton, 2009, p. 18). So the dynamics are complicated. A lot of carpentry will be required to make all goalposts the same height.
Variation in Achievement
Broad variation exists in student achievement for any particular grade level and school subject. The degree of variation makes a huge difference in the effort it will take to raise all U.S. students to the same level. Figure 1 (p. 28) provides a panorama of student scores on the reading portion of the 1990 and 2004 NAEP assessments at ages 9, 13, and 17 for all racial and ethnic subgroups. If you place a ruler across the chart at the point showing 9-year-olds scoring in the 90th percentile to see what percentile it reaches for the scores of 17-year-olds, you'll see that the bottom fourth of 17-year-olds read no better than the top tenth of 9-year-olds.
Figure 1. Percentile Distributions of NAEP Reading Scores by Age and Racial/Ethnic Group, 1990 and 2004
Source: Data from the National Assessment of Educational Progress analyzed by Educational Testing Service.
To gauge the magnitude of this variation, consider that the spread of scores within any one grade level is as great as, or greater than, the difference in the average scores from the 4th grade to the 12th. This is a lot of variation to deal with.
Commonality Versus Tolerance for Diversity
The United States is a diverse society created by people from many cultures, religions, and parts of the world. We live in communities and neighborhoods with extremes in wealth and income. These extreme differences have been well tolerated by Americans because we are a land of opportunity. And, because we believe that opportunity depends a lot on education, we want to reduce inequality in our school system.
The American experience has always accommodated a great deal of disagreement about what to teach, how to teach it, and when to teach it. As late as March 2009, a panel appointed by President George W. Bush tried—again—to put to rest what the New York Times called "the long, heated debate over math teaching methods" (Lewin, 2008). Debates over how to teach reading are still divisive. Controversies over teaching evolution and other topics still flare up around the country—at times even involving literal textbook fires, such as in West Virginia in the 1970s.
All this diversity and disagreement is like a coil-spring mattress; the weight pressing on one area of the mattress is tolerated without affecting the rest of the mattress. When we try to settle basic differences about education at a national level, however, we elevate the stakes. We open the door for national-level organizations to bring pressures and counter-pressures to bear. As U.S. citizens balance a desire for commonality against tolerance for differences, we face the question of how strongly we want to raise such issues to a court of national settlement.
Up the Hill Again?
Beginning in March 2009, signs emerged that supporters of national standards were making another effort to storm the hill under a new banner: Common Standards. If the word federal has become tarnished in connection with standards, so has the word national; both words suggest requirements emanating from on high, even if not from the federal government. The National Governors Association, the National Council of Chief State School Officers, Achieve, the College Board, and ACT have organized into a strong coalition. That coalition moved quickly to get most states to sign an agreement to unite around a set of common standards. At this point, this means common content standards for the subject matter to be taught at each grade level.
The starting point has been to write standards for what U.S. students should know when they leave high school to be ready for college and career. As this article went to press, the coalition had drafted standards in math and reading and was putting those drafts through a process of "validation" by a committee that the coalition appointed. The next step will be to write standards for each grade level. Then the question will be whether the coalition moves on to create a common test for each grade in math and reading and advances to other subject areas.
At this stage, there is substantial momentum. As of August 2009, 49 states had signed on. Although this "sign on" has been from the chief state school officials and governors, many state legislatures have been heavily involved in the standards-setting movement. It remains to be seen whether they will go along or balk. In 2007, the National Conference of State Legislatures adopted a measure against national standards that favored higher standards developed for—and by—each state.
Another element in the mix is the fact that at the end of January, states applied for a share of the $48.6 billion in economic stimulus funds to be distributed under the American Recovery and Reimbursement Act. The selection process, underway as this article goes to press, takes into account whether each applying state is engaged in collaborative efforts with other states to develop common standards—a considerable incentive to get with the program.
Learning From Experience
Years of experience with "standards-based reform" and its transformation into test-based accountability tells us that there is more trouble in Accountability City than just lack of commonality. A consideration of our current accountability landscape and our past attempts raises four questions.
How can we develop high-quality content standards and tests?
There are quality problems with current content standards, as evaluations by independent researchers and evaluators have shown. The tests that accompany them are typically "cheapies" that do not get far enough below the surface in terms of content instruction. Content standards, created through compromises among committee members, are far broader than what can be covered in a nine-month school year. And when tests are constructed that merely sample this broad content—as most do—such tests are not sensitive to the actual instruction teachers deliver. Nor do they pick up changes in achievement that follow improvements in instruction.
How can schools serve the highest and lowest achievers?
The selection of a single cut point provides incentives for teachers and schools to overfocus on students near that cut point, to avoid sanctions. Because results are reported in terms of the cut point, teachers know nothing about how students who are far below— or above—that cut point are doing.
Should we contemplate standards across a range of subjects?
Standards now concentrate on math and reading. The incentive this creates for distorting the curriculum has been widely discussed; many people are concerned about cutting back on social studies, music, and art. We must reflect on whether to extend standards to other disciplines to keep the curriculum broad.
How can we measure gains in student achievement?
We now test students at a single point in time to establish the effectiveness of schools. But that test actually measures all a student knows, including learning generated in early childhood, out of school, or even in different schools. To achieve true test-based accountability for schools, we need to measure the effectiveness of what a student learned while in that school in addition to what the student knows in total.
Even if we achieve commonality in curriculum and performance standards, these questions will remain. We should ask ourselves, Do we want a common version of the current operating model, or do we want to work more on the model?
Teachers have seen the locus of power move up the governmental hierarchy over the past few decades. Educators at all levels should stay tuned and use their on-the-ground experience and judgment to determine what is good for students—and for educational equity—as we consider whether national standards are to be or not to be.
