Brainstorming ways to improve school choice markets has consumed much of my thinking of late. Regulation plays a key role in marketplaces, and thus improving how we regulate school choice markets is central to improving school choice writ large.
Over the long weekend (and between House of Cards bingeing) I sat down to read what many have called the Bible of regulatory reform, Stephen Breyer’s 1982 tome Regulation and its Reforms.
It was a treasure-trove, but two specific passages from the chapter on “standard setting” ring particularly true for the current debate over how school choice markets should be regulated.
Ultimately, regulating school choice markets is about standards-setting. It is about creating either design or performance expectations for schools and allowing only those schools that meet those expectations to participate in the program. Throughout his chapter on that topic, Breyer uses the National Highway Traffic Safety Administration’s forays into regulating automobiles as a case study in standard setting regulation. The first insight regards the process by which standards are created:
It is tempting but misleading to think of the standard-setting process as it might exist in the idealized world of the rational policy planner… This description is misleading, because it cannot account for elements that even a cursory description of NHTSA experience suggests are typical of the standard-setting process. The process in practice appears to rely heavily upon precedent for the content of its standards, it is characterized by continued negotiation, modification of proposals, and long delays. Yet its standards, once in place, prove surprisingly resistant to change, even when, as in the case of head restraints, experiences suggests at they are ineffective. (pg. 99)
As I wrote last week, we are only now becoming able to quantify just how multi-dimensional school quality is. A push-back to my argument was that we don’t know how to measure these other constructs, so until we do, we need to stick to what we can measure. While that makes a great deal of practical sense, Breyer’s cautions regarding the durability of those metrics should worry us. Metrics are slow to change and adapt to changing circumstances and can continue to require products to have redundant or ineffective features in order to be in compliance.
The second insight regards the type of regulations that are used to judge firms and products, whether they are “performance” or “design” standards. In education, almost all reformers proposing regulations for school choice markets (present company included) push for performance, not design, regulations. We care how schools affect student knowledge and well being, not how they look. But Breyer’s words should give us pause:
In practice, the notions of ‘performance’ and ‘design’ tend to converge. Congress, for example, insisted that NHTSA write only performance standards. Yet it was not difficult for the agency to write performance specifications that could be met only by a machine of certain design. When NHTSA initially set passive restraint standards, it insisted that manufactures satisfy performance tests that effectively required them to use airbags. Similarly, it set bumper standards that could not be met by metal bumpers, thereby requiring the use of ‘soft-face’ bumpers…Despite the agency’s theoretical ability to transpose performance and design standards, the underlying tension remains. There is a fundamental tension between flexibility, embodied in a statement of goals that allows maximum freedom to the industry to meet them, and enforceability, which points in the direction of specific, usable, recognizable detail. (pgs. 105-106)
Gulp. Regulating performance is hard. Regulating design is easy. Regulators have every incentive to draw clear, defensible rules that will stand up to legal challenge and are straightforward to explain to policymakers and operators. This should trouble us when we think about schools.
The past half decade has seen an explosion in new models for educating kids. Hybrid schools, online schools, schools with innovative staffing, course-choice programs, and new models that we don’t even know about yet have the potential to fundamentally disrupt how we educate children in America. In order for them to exist, any regulations will need to be performance-based, because regulation is simply too slow to keep up with design. But we have to avoid performance measures that can only be satisfied by particular school models. This is tricky.
The key questions to ask when evaluating new school regulations: do the metrics we’ve selected measure the things we really want out of schools, or just what is easiest for us to measure? Do they privilege a particular organization of schooling? Do they rule out other models wholesale?
Without careful thought, regulators have every incentive to fall back onto design regulations, and that shuts the door to an incredible opportunity.
Follow AEIdeas on Twitter at @AEIdeas.