Jay P. Greene the 21st Century distinguished professor of education policy and chair of the Department of Education Reform at the University of Arkansas. Greene’s work has been published in journals from a diverse set of disciplines, including education, sociology, public policy, psychology, political science and economics.

Assuming there are at least some legitimate differences of opinion about what constitutes character or virtue (or which aspects of character should be prioritized), shouldn’t we decentralize this area of education and allow districts and schools decide what to do? That way different communities can choose for themselves how their students should be formed. Or is there a role for something more standardized at the state level?

The simple answer is that we must decentralize education if we hope to have any meaningful type of character education that is not highly oppressive. Good character is the set of qualities that people believe their children should possess to pursue a good life. Since we have significant and legitimate diversity of views on what constitutes a good life and what qualities are necessary to accomplish that, we need to allow local communities and families the flexibility to pursue the type and content of character education that best suits their differing perspectives.

It’s one thing to measure reading or math proficiency—standardized tests can do much of that. But, how do we measure (and therefore hold teachers, schools or school networks accountable for transmitting) character and virtue? Or, said another way, how can we tell if we are successfully developing students of character?

The answer to this question hinges on what is meant by “we.” Distant policymakers, scholars and advocates really have no way to assess character education. The attempts that have been made to develop standardized assessments of character education or social-emotional learning outcomes have largely failed. The measures they propose are too vague, too gameable and insufficiently predictive to be of much use. But if “we” means families and communities, then it is fairly easy to tell if character education is producing success because those constituents can directly observe students and assess whether they are developing desired qualities.

A critic might say that “character” is nothing more than the norms of the dominant group. If that’s the case, families who don’t identify as part of that group—whether because of income, race, religion, heritage or something else—might always take issue with aspects of democratically developed character education. What should we make of that?

This is a fair criticism of efforts at centrally controlled character education. The solution, as discussed above, is to decentralize control over character education so that families and communities can design the manner and content of the effort in a way that is consistent with their legitimately diverse set of preferences.

Is there a meaningful difference in your mind between character education and today’s more popular “social-emotional learning”? For example, does your preferred vision of character education add something that SEL lacks—or maybe you believe SEL helps us avoid problems caused by character education?

I do not believe there is any real difference between SEL and character education. In fact, the main argument of a recent piece I wrote that SEL is inextricably connected to character education, which, historically, is part of religious education. SEL seems like a marketing effort to give character education faux-scientific authority, move character education to a level of generality that will meet less political opposition but also likely be less useful, and to disguise its moral and religious roots.

Public school systems can be wary of engaging in anything that seems like instruction about morality because morality brushes up against faith traditions, the First Amendment, different conceptions of the “good life,” and so on. How can a school have a robust approach to forming student character and avoid the charge that it is veering toward religious instruction?

Unfortunately, there is no avoiding the fact that character education is connected to moral and religious traditions. As I noted in my piece cited above, the main components of SEL identified by CASEL, the main national organization promoting this effort, correspond almost identically to the Cardinal Virtues, which come to us from Plato through Christian moral teaching. This is why education in general, and character education in particular, cannot fully avoid religion. For almost all of U.S. history and in most places in the country outside of elite, costal centers, schools have continued to be religious in practice even if they are officially secular. Public schools used to be very strongly Protestant, and they remain that way in diluted form in much of the country. If we had private school choice, we could avoid church-state controversies by allowing families to find the character education embedded within the religious or moral traditions they prefer.

(Read the full list of questions here. Find the other “Character Matters” Q&As as they publish here.)