Comment

Julian R. Betts and Anne Danenberg have done an admirable job both in explaining the complex California accountability system and in analyzing its initial effects. They offer sound evidence to support two strong conclusions: that student performance, as measured by the state test, improved substantially, particularly for the lowest performing schools; and that teacher quality, as measured by a number of dimensions, declined, particularly in those same schools.

At one level these two findings pose a paradox: How can student performance increase in schools with less-qualified and less-experienced teachers? There are several possible explanations. Maybe teacher quality does not matter as much as other research suggests it might.³⁴ Maybe the gains in achievement are inflated, as Daniel Koretz suggests may have been the case in other states.³⁵ Betts and Danenberg pose a third possibility--that the incentives created by the accountability system were sufficient to encourage low-performing schools to overcome the shortcomings in teacher preparation and raise performance anyway.

I will suggest a fourth possibility: the lack of coherence in California's education policy. To be sure, coherence in education policy is much easier to wish for than achieve in a system of widely dispersed authority, and California is far from alone in falling short of the ideal. But standards-based reform was aimed, at least by its architects, at imposing some coherence by placing standards for student performance at the core of policy around curriculum, teacher development, and testing.

California is not there yet. In part, this is because, more so than in most other states, many actors have hands in setting education policy in the state--the governor, the elected state superintendent of public instruction, the state Board of Education, the legislature, and the voters, through the widely used initiative process. And that is only at the state level; districts set their own policies as well. Achieving coherence is difficult under any circumstances with that kind of governance structure. Even acting with the best of intentions, these actors can adopt policies that do not support, or may conflict with, other policies.

For example, in 1996, as the state was beginning to embark on its standards-based strategy by developing content standards and implementing a statewide test, the legislature, with the strong support of Governor Pete Wilson, mandated a substantial reduction in class sizes in the early grades. This initiative in principle supported the notion of standards-based instruction by enabling teachers to interact with young children. But in practice the law produced a scramble to hire new teachers to staff the smaller classes, and many schools--particularly those serving low-income and minority children, which tended to be low performing--were forced to hire underqualified teachers. This outcome, which undermines the vision of standards-based reform of teachers prepared to teach the content standards, could help explain the lower levels of teacher qualifications in low-performing schools that Betts and Danenberg found--at least in the early grades.

Are Betts and Danenberg correct that the accountability system was powerful enough to overcome this trend? Unfortunately, no one knows, and California's lack of coherence makes it difficult to make judgments. In a more coherent system, someone might be able to do so. The theory of action underlying the reform suggests that the standards communicate the expectations for instruction and learning, assessments measure the extent to which schools and students are meeting those expectations, and accountability creates incentives for schools to improve performance against the standards and directs resources to schools that are falling behind. This theory rests on assumptions that may not have proven true, such as the availability of high-quality professional development. But the theory suggests that the incentives encourage schools to focus on the standards and that improvements represent progress toward those standards.³⁶

California's system, at least so far, does not match that ideal. The state has all the elements, but they do not work together as the architects of standards-based...