I am en route to DC to help with planning the next National Assessment in Education Progress (NAEP) for Civics. I was also on the design team for the 2011 test, which yielded newspaper articles like “Failing Grades on Civics Exam Called a ‘Crisis’” (The New York Times) and a Washington Post editorial on “civic illiteracy.”

I have a somewhat different view of the data. I think the NAEP shows that students know quite a bit about civics, although less about some topics than others. For example, 61 percent of American twelfth graders can interpret a passage from the majority opinion in Schenck v. The United States (1919), although only those at the NAEP’s “advanced” level could compare the citizenship requirements of the United States to those in other countries. (Other research has found that American students compare well to international peers when it comes to understanding their own political system, but are very naïve about foreign governments.)

There is no evidence of decline. The mean scores at eighth and twelfth grade are flat, and the fourth grade scores have risen. But surely there is a “crisis” when only 24 percent of twelfth graders score at “Proficient” on the NAEP Civics Assessment? Not necessarily: the cutoffs for “basic,” “proficient,” and “advanced” are essentially arbitrary, set originally by a small committee of teachers and subject-matter experts. A different committee could easily have set the standards differently and concluded that most American students are proficient. Compared to 14-year-olds in a sample of 28 other countries, American students perform substantially above average. The value of the proficiency levels is not for deciding how much our students know but for comparing scores from year to year or one group of students to another. The most recent NAEP shows improvements, but only at the fourth grade level, and troubling gaps in proficiency by race, ethnicity, and parents’ education.

Education Secretary Arne Duncan has deplored the fact that “a staggering number of Americans do not know much of the basic history and traditions of our nation.” He cited information that many adults fail to know, ranging from the three branches of government to the names of today’s Supreme Court justices. But the structure of the federal government is included in state standards, taught in mandatory courses, and included on standardized tests. Every state except Iowa (with its strong tradition of local control) has civics standards. Ninety-seven percent of high school seniors reported in 2011 that they have taken classes in civics or American government

In other words, education policy is already designed to ensure that young people know the three branches of government, and 69 percent of high school seniors can correctly answer a question about Marbury v. Madison (relevant to relations between two of the branches), even though the NAEP has no stakes and no one studies for it.The main problem is probably that adults tend not to remember such concepts even though they knew them when they were teenagers.

On the other hand, the names of individuals who serve today on the Supreme Court and hold other public offices are virtually never included on standardized tests. News and current events are not common areas of emphasis in social studies classes.

How we assess policies for civic education depends essentially on what we think is most important for young people to learn. In my view, the constitutional structure of the United States government, its origins, underlying premises, and most debated aspects, are worthy topics of study. I teach them myself, even at the graduate level. I wish that everyone understood them. I also think that Kantian, utilitarian, and Aristotelian ethics, doctrinal differences among world religions, and the nature of modernism and postmodernism are worthy topics that people should study. We teach the US constitutional structure but we don’t teach, for example, Kant (who is a major topic in other countries and of much greater global significance than any American founder). Our choices are defensible but not necessary. The Republic would survive if fewer people understood Marbury v. Madison. What really worries me is not measured on the NAEP, and that is the proportion of people who have the skills and values they need to participate effectively in our civil society.

Underlying my position here are some doubts about “American exceptionalism” and “Constitutional piety” combined with severe concerns about the state of our civil society.