This paper is entitled School Choice: Three Essential Elements and Several Policy Options and I think readers are due an explanation as to why, after more than a decade of research and policy action on school choice, I am still discussing its essential elements.

The idea of school choice comes from economics. It is not merely that an economist, Milton Friedman, is the author of the modern idea of school choice. Rather, the entire logic of school choice is based on economics, and the answer to every question on school choice draws on some part of economics – whether it be human capital theory, public finance, fiscal federalism, mechanism design or general equilibrium analysis.

Yet, it is a struggle to keep economics in the discussion of school choice. When I first started researching school choice more than a decade ago, the reason it was a struggle to keep economics in the discussion was that hardly any economists were participating. Books, papers and conferences were dominated by political scientists and sociologists and, even though some of them were admirers rather than detractors of economics, none drew upon economic logic with any real consistency. They would describe programmes as ‘school choice’ initiatives even when those programmes lacked all of the essentials that give school choice its economic logic.

At the same time, it was difficult to get researchers to recognise this as a problem because they were focused on the political and social aspects of these so-called school choice programmes, thereby neglecting incentives, constraints and other properties that are crucial to a programme’s economic logic. Today, there are many economists working on school choice, so we have far more opportunity not only to participate in, but to direct, the discussion.

Nevertheless, even with numerous economists at work, the struggle to keep economics in the discussion of school choice has re-emerged. Why is this? Some empirical economists have become absorbed in doing almost pure programme evaluation, so that they evaluate programmes with the word ‘choice’ in them as though they were school choice programmes, without themselves noting that the policies have little or nothing to do with the economic idea of school choice. How can we economists fall into such error?