for low-performing schools are a crucial component of a well-designed accountability system. In their absence, a school-based accountability system may become simply a system for assigning blame rather than a system for improving student performance.

Such systems are fully consistent with the goal of altering incentives to make performance count, given their focus on student outcomes rather than on the inputs that typically were the focus of state accountability systems in the past. A second characteristic is that they focus attention on the school as the improvement unit rather than on the school district, on the teachers, or on the students. The focus on the school is designed to encourage all school personnel to work cooperatively toward a common, well-specified goal. Provided schools are given more management authority than in the past, the schools then would be in a position to rearrange the use of resources toward the goal of higher achievement.

One possible disadvantage of the focus on schools as the unit of accountability is the free rider problem: bonuses are typically given to all teachers (and possibly support staff as well) in an effective school regardless of their contribution. A further, potentially more serious problem with the focus on schools is that good teachers in low-performing schools may prefer to leave such schools in favor of schools where their chances of earning a financial bonus are higher.

Use of Tests for High-Stakes Accountability

Central to all the school-based accountability and incentive systems is the measurement of student performance, as typically measured by test scores. In Chapter 4, we commented on the current status of assessment. We turn here to the uses of tests for high-stakes accountability.

Because student test scores are so highly correlated with student background characteristics, it is essential that school-based accountability systems focus on gains in, rather than levels of, student performance. Otherwise the indicators of success would measure the background of the students rather than the contribution of the schools to student learning. Thus, measuring a school's value-added is at the heart of any sophisticated accountability system. Koretz (1996), Meyer (1996), Clotfelter and Ladd (1996), and Ladd and Walsh (1998) describe the technical challenges involved in using assessments for educational accountability and in developing value-added indicators of school performance. In addition to the technical challenges are considerations related to costs. As Meyer and others note, value-added indicator systems may be costly if they involve frequent testing and comprehensive data systems containing information on student test scores and student, family, and community characteristics. In short, they may require a major commitment on the part of school districts and states. Potentially, however, computer-adaptive tests could provide the requisite data without suffering



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement