If I told you, "I favor school choice," what might I actually mean? Perhaps I'm a libertarian, so I favor choice because I place an extreme value on individual freedom. Or perhaps I believe in the invisible hand of the free market or I buy into certain assumptions about the benefits of school competition. Maybe I want a school that reinforces my religious beliefs. Or perhaps I see school choice as a way to create more diverse schools, uncoupled from neighborhood segregation.

But in truth, favoring school choice means about the same thing as favoring computers -- the statement means very little unless and until the ideas take on a specific form. After all, a choice plan intended to mitigate neighborhood segregation would be designed with very different rules than a choice plan envisioned as a way to increase competition.

So if I say, "I favor school choice," your next question should be, "How do you favor designing andoperating school choice?" That is, "What does your school choice policy actually look like in practice?"

School choice isn't a single thing. In addition to charter schools and conventional vouchers, choice policies include homeschooling, cyberschools, magnet schools and other types of within-district "open enrollment" policies, choice across school districts, and tuition tax credit plans that provide a public subsidy for private school tuition (policies that I've called "neovouchers").

In addition, there are many variations on each of these approaches. Are vouchers provided to everyone or just to lower-income families? Do schools enrolling voucher students have to comply with the accountability policies that apply to public schools? Similarly, must they comply with anti-discrimination laws? Must teachers at cyberschools or charter schools be fully trained and certified? Should any constraints be placed on the choice decisions of parents and schools in order to avoid segregation and stratification by wealth, race, special needs, or other attributes?

There is a widespread misconception about school choice: that it's a complete policy in itself. In reality, it's a broad policy tool that can be included as part of a complete education policy. Like most tools, school choice can be used in beneficial as well as damaging ways. It's not a policy in itself any more than bell schedules are complete policies. At the most basic level, school choice is simply an approach for assigning students to schools. Once we appreciate this distinction, we return to our basic question: "In what ways do you favor designing and operating school choice?"

Unfortunately, our lawmakers have generally failed to ask such sensible questions. As a result, they wrongly look at school choice as a political end in itself, not as a tool. And this misunderstanding blinds them to all the ways that school choice tools can be used.

Along with three of my colleagues at the National Education Policy Center -- Gary Miron, Patricia Hinchey and Bill Mathis -- I recently edited a book called Exploring the School Choice Universe. It examines not just what school choice policies have done but also what those policies can do, offering reasons to believe that choice policies can further some educational goals but also suggesting many reasons for caution. If choice policies are to be evidence-based, a re-examination is in order. We hope that the information, insights and recommendations in the book will facilitate a more nuanced understanding of school choice and provide the basis for designing sensible school choice reforms that can pursue a range of desirable outcomes.

If we don't understand our options within the school choice framework, or if we're not even aware of them, we are very unlikely to carefully deliberate about them. We end up with choice-based policies that have a set of rules arrived at through default or through political pressure, rather than through careful -- and evidence-based -- consideration. Simply put, if we don't ask useful questions, we don't tend to get useful answers.

There can be a true value in parental choice -- matching, for example, a child's interests with the focus of a school. But in making policy we shouldn't assume school choice has some magical power. And in designing choice policies we should remain aware of potential unintended consequences: it can become a system for the rich to get richer, the poor to get poorer, and the entire society to balkanize further. What the authors of this book tell us, however, is that it needn't have such adverse consequences. We can design choice systems that advance the democratic nature of society and educational equality. But to make this work, we must rethink school choice, designing our policies with goals -- not tools -- at the center of our thinking.